commit d838fbf2b422b54f552be06bad1a998337664e16 Author: Russia9 Date: Sat Apr 16 19:15:29 2022 +0300 Init diff --git a/.drone.yml b/.drone.yml new file mode 100644 index 0000000..73fcd08 --- /dev/null +++ b/.drone.yml @@ -0,0 +1,37 @@ +kind: pipeline +type: docker +name: build + +steps: + - name: build + image: plugins/docker + settings: + registry: cr.selcloud.ru + username: + from_secret: docker_username + password: + from_secret: docker_password + repo: cr.selcloud.ru/russia9/${DRONE_REPO_NAME}/${DRONE_COMMIT_BRANCH} + tags: + - latest + - ${DRONE_COMMIT_SHA} + cache_from: + - cr.selcloud.ru/russia9/${DRONE_REPO_NAME}/${DRONE_COMMIT_BRANCH}:latest + + - name: deploy + image: appleboy/drone-ssh + settings: + host: + from_secret: ssh_address + username: + from_secret: ssh_username + key: + from_secret: ssh_key + port: 22 + script_stop: true + script: + - docker-compose -f /srv/chatwars-spy/docker-compose.yml pull + - docker-compose -f /srv/chatwars-spy/docker-compose.yml up -d + when: + branch: + - master diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..f2b36e1 --- /dev/null +++ b/.gitignore @@ -0,0 +1,63 @@ +# Created by .ignore support plugin (hsz.mobi) +### Go template +# Binaries for programs and plugins +*.exe +*.exe~ +*.dll +*.so +*.dylib + +# Test binary, built with `go test -c` +*.test + +# Output of the go coverage tool, specifically when used with LiteIDE +*.out + +# Dependency directories (remove the comment below to include it) +# vendor/ + +### Vim template +# Swap +[._]*.s[a-v][a-z] +!*.svg # comment out if you don't need vector files +[._]*.sw[a-p] +[._]s[a-rt-v][a-z] +[._]ss[a-gi-z] +[._]sw[a-p] + +# Session +Session.vim +Sessionx.vim + +# Temporary +.netrwhist +*~ +# Auto-generated tag files +tags +# Persistent undo +[._]*.un~ + +### JetBrains template +# Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio, WebStorm and Rider +# Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839 + +# User-specific stuff +.idea + +# CMake +cmake-build-*/ + +# File-based project format +*.iws + +# IntelliJ +out/ + +# JIRA plugin +atlassian-ide-plugin.xml + +# Crashlytics plugin (for Android Studio and IntelliJ) +com_crashlytics_export_strings.xml +crashlytics.properties +crashlytics-build.properties +fabric.properties \ No newline at end of file diff --git a/Dockerfile b/Dockerfile new file mode 100644 index 0000000..fa27b2b --- /dev/null +++ b/Dockerfile @@ -0,0 +1,13 @@ +FROM golang:1.18 + +# Set app workdir +WORKDIR /go/src/app + +# Copy application sources +COPY . . + +# Build app +RUN go build -o app gitea.russia9.dev/Russia9/chatwars-spy/cmd/main + +# Run app +CMD ["./app"] diff --git a/LICENSE b/LICENSE new file mode 100644 index 0000000..f288702 --- /dev/null +++ b/LICENSE @@ -0,0 +1,674 @@ + GNU GENERAL PUBLIC LICENSE + Version 3, 29 June 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU General Public License is a free, copyleft license for +software and other kinds of works. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +the GNU General Public License is intended to guarantee your freedom to +share and change all versions of a program--to make sure it remains free +software for all its users. We, the Free Software Foundation, use the +GNU General Public License for most of our software; it applies also to +any other work released this way by its authors. You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +them if you wish), that you receive source code or can get it if you +want it, that you can change the software or use pieces of it in new +free programs, and that you know you can do these things. + + To protect your rights, we need to prevent others from denying you +these rights or asking you to surrender the rights. Therefore, you have +certain responsibilities if you distribute copies of the software, or if +you modify it: responsibilities to respect the freedom of others. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must pass on to the recipients the same +freedoms that you received. You must make sure that they, too, receive +or can get the source code. And you must show them these terms so they +know their rights. + + Developers that use the GNU GPL protect your rights with two steps: +(1) assert copyright on the software, and (2) offer you this License +giving you legal permission to copy, distribute and/or modify it. + + For the developers' and authors' protection, the GPL clearly explains +that there is no warranty for this free software. For both users' and +authors' sake, the GPL requires that modified versions be marked as +changed, so that their problems will not be attributed erroneously to +authors of previous versions. + + Some devices are designed to deny users access to install or run +modified versions of the software inside them, although the manufacturer +can do so. This is fundamentally incompatible with the aim of +protecting users' freedom to change the software. The systematic +pattern of such abuse occurs in the area of products for individuals to +use, which is precisely where it is most unacceptable. Therefore, we +have designed this version of the GPL to prohibit the practice for those +products. If such problems arise substantially in other domains, we +stand ready to extend this provision to those domains in future versions +of the GPL, as needed to protect the freedom of users. + + Finally, every program is threatened constantly by software patents. +States should not allow patents to restrict development and use of +software on general-purpose computers, but in those that do, we wish to +avoid the special danger that patents applied to a free program could +make it effectively proprietary. To prevent this, the GPL assures that +patents cannot be used to render the program non-free. + + The precise terms and conditions for copying, distribution and +modification follow. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "The Program" refers to any copyrightable work licensed under this +License. Each licensee is addressed as "you". "Licensees" and +"recipients" may be individuals or organizations. + + To "modify" a work means to copy from or adapt all or part of the work +in a fashion requiring copyright permission, other than the making of an +exact copy. The resulting work is called a "modified version" of the +earlier work or a work "based on" the earlier work. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + To "propagate" a work means to do anything with it that, without +permission, would make you directly or secondarily liable for +infringement under applicable copyright law, except executing it on a +computer or modifying a private copy. Propagation includes copying, +distribution (with or without modification), making available to the +public, and in some countries other activities as well. + + To "convey" a work means any kind of propagation that enables other +parties to make or receive copies. Mere interaction with a user through +a computer network, with no transfer of a copy, is not conveying. + + An interactive user interface displays "Appropriate Legal Notices" +to the extent that it includes a convenient and prominently visible +feature that (1) displays an appropriate copyright notice, and (2) +tells the user that there is no warranty for the work (except to the +extent that warranties are provided), that licensees may convey the +work under this License, and how to view a copy of this License. If +the interface presents a list of user commands or options, such as a +menu, a prominent item in the list meets this criterion. + + 1. Source Code. + + The "source code" for a work means the preferred form of the work +for making modifications to it. "Object code" means any non-source +form of a work. + + A "Standard Interface" means an interface that either is an official +standard defined by a recognized standards body, or, in the case of +interfaces specified for a particular programming language, one that +is widely used among developers working in that language. + + The "System Libraries" of an executable work include anything, other +than the work as a whole, that (a) is included in the normal form of +packaging a Major Component, but which is not part of that Major +Component, and (b) serves only to enable use of the work with that +Major Component, or to implement a Standard Interface for which an +implementation is available to the public in source code form. A +"Major Component", in this context, means a major essential component +(kernel, window system, and so on) of the specific operating system +(if any) on which the executable work runs, or a compiler used to +produce the work, or an object code interpreter used to run it. + + The "Corresponding Source" for a work in object code form means all +the source code needed to generate, install, and (for an executable +work) run the object code and to modify the work, including scripts to +control those activities. However, it does not include the work's +System Libraries, or general-purpose tools or generally available free +programs which are used unmodified in performing those activities but +which are not part of the work. For example, Corresponding Source +includes interface definition files associated with source files for +the work, and the source code for shared libraries and dynamically +linked subprograms that the work is specifically designed to require, +such as by intimate data communication or control flow between those +subprograms and other parts of the work. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 2. Basic Permissions. + + All rights granted under this License are granted for the term of +copyright on the Program, and are irrevocable provided the stated +conditions are met. This License explicitly affirms your unlimited +permission to run the unmodified Program. The output from running a +covered work is covered by this License only if the output, given its +content, constitutes a covered work. This License acknowledges your +rights of fair use or other equivalent, as provided by copyright law. + + You may make, run and propagate covered works that you do not +convey, without conditions so long as your license otherwise remains +in force. You may convey covered works to others for the sole purpose +of having them make modifications exclusively for you, or provide you +with facilities for running those works, provided that you comply with +the terms of this License in conveying all material for which you do +not control copyright. Those thus making or running the covered works +for you must do so exclusively on your behalf, under your direction +and control, on terms that prohibit them from making any copies of +your copyrighted material outside their relationship with you. + + Conveying under any other circumstances is permitted solely under +the conditions stated below. Sublicensing is not allowed; section 10 +makes it unnecessary. + + 3. Protecting Users' Legal Rights From Anti-Circumvention Law. + + No covered work shall be deemed part of an effective technological +measure under any applicable law fulfilling obligations under article +11 of the WIPO copyright treaty adopted on 20 December 1996, or +similar laws prohibiting or restricting circumvention of such +measures. + + When you convey a covered work, you waive any legal power to forbid +circumvention of technological measures to the extent such circumvention +is effected by exercising rights under this License with respect to +the covered work, and you disclaim any intention to limit operation or +modification of the work as a means of enforcing, against the work's +users, your or third parties' legal rights to forbid circumvention of +technological measures. + + 4. Conveying Verbatim Copies. + + You may convey verbatim copies of the Program's source code as you +receive it, in any medium, provided that you conspicuously and +appropriately publish on each copy an appropriate copyright notice; +keep intact all notices stating that this License and any +non-permissive terms added in accord with section 7 apply to the code; +keep intact all notices of the absence of any warranty; and give all +recipients a copy of this License along with the Program. + + You may charge any price or no price for each copy that you convey, +and you may offer support or warranty protection for a fee. + + 5. Conveying Modified Source Versions. + + You may convey a work based on the Program, or the modifications to +produce it from the Program, in the form of source code under the +terms of section 4, provided that you also meet all of these conditions: + + a) The work must carry prominent notices stating that you modified + it, and giving a relevant date. + + b) The work must carry prominent notices stating that it is + released under this License and any conditions added under section + 7. This requirement modifies the requirement in section 4 to + "keep intact all notices". + + c) You must license the entire work, as a whole, under this + License to anyone who comes into possession of a copy. This + License will therefore apply, along with any applicable section 7 + additional terms, to the whole of the work, and all its parts, + regardless of how they are packaged. This License gives no + permission to license the work in any other way, but it does not + invalidate such permission if you have separately received it. + + d) If the work has interactive user interfaces, each must display + Appropriate Legal Notices; however, if the Program has interactive + interfaces that do not display Appropriate Legal Notices, your + work need not make them do so. + + A compilation of a covered work with other separate and independent +works, which are not by their nature extensions of the covered work, +and which are not combined with it such as to form a larger program, +in or on a volume of a storage or distribution medium, is called an +"aggregate" if the compilation and its resulting copyright are not +used to limit the access or legal rights of the compilation's users +beyond what the individual works permit. Inclusion of a covered work +in an aggregate does not cause this License to apply to the other +parts of the aggregate. + + 6. Conveying Non-Source Forms. + + You may convey a covered work in object code form under the terms +of sections 4 and 5, provided that you also convey the +machine-readable Corresponding Source under the terms of this License, +in one of these ways: + + a) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by the + Corresponding Source fixed on a durable physical medium + customarily used for software interchange. + + b) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by a + written offer, valid for at least three years and valid for as + long as you offer spare parts or customer support for that product + model, to give anyone who possesses the object code either (1) a + copy of the Corresponding Source for all the software in the + product that is covered by this License, on a durable physical + medium customarily used for software interchange, for a price no + more than your reasonable cost of physically performing this + conveying of source, or (2) access to copy the + Corresponding Source from a network server at no charge. + + c) Convey individual copies of the object code with a copy of the + written offer to provide the Corresponding Source. This + alternative is allowed only occasionally and noncommercially, and + only if you received the object code with such an offer, in accord + with subsection 6b. + + d) Convey the object code by offering access from a designated + place (gratis or for a charge), and offer equivalent access to the + Corresponding Source in the same way through the same place at no + further charge. You need not require recipients to copy the + Corresponding Source along with the object code. If the place to + copy the object code is a network server, the Corresponding Source + may be on a different server (operated by you or a third party) + that supports equivalent copying facilities, provided you maintain + clear directions next to the object code saying where to find the + Corresponding Source. Regardless of what server hosts the + Corresponding Source, you remain obligated to ensure that it is + available for as long as needed to satisfy these requirements. + + e) Convey the object code using peer-to-peer transmission, provided + you inform other peers where the object code and Corresponding + Source of the work are being offered to the general public at no + charge under subsection 6d. + + A separable portion of the object code, whose source code is excluded +from the Corresponding Source as a System Library, need not be +included in conveying the object code work. + + A "User Product" is either (1) a "consumer product", which means any +tangible personal property which is normally used for personal, family, +or household purposes, or (2) anything designed or sold for incorporation +into a dwelling. In determining whether a product is a consumer product, +doubtful cases shall be resolved in favor of coverage. For a particular +product received by a particular user, "normally used" refers to a +typical or common use of that class of product, regardless of the status +of the particular user or of the way in which the particular user +actually uses, or expects or is expected to use, the product. A product +is a consumer product regardless of whether the product has substantial +commercial, industrial or non-consumer uses, unless such uses represent +the only significant mode of use of the product. + + "Installation Information" for a User Product means any methods, +procedures, authorization keys, or other information required to install +and execute modified versions of a covered work in that User Product from +a modified version of its Corresponding Source. The information must +suffice to ensure that the continued functioning of the modified object +code is in no case prevented or interfered with solely because +modification has been made. + + If you convey an object code work under this section in, or with, or +specifically for use in, a User Product, and the conveying occurs as +part of a transaction in which the right of possession and use of the +User Product is transferred to the recipient in perpetuity or for a +fixed term (regardless of how the transaction is characterized), the +Corresponding Source conveyed under this section must be accompanied +by the Installation Information. But this requirement does not apply +if neither you nor any third party retains the ability to install +modified object code on the User Product (for example, the work has +been installed in ROM). + + The requirement to provide Installation Information does not include a +requirement to continue to provide support service, warranty, or updates +for a work that has been modified or installed by the recipient, or for +the User Product in which it has been modified or installed. Access to a +network may be denied when the modification itself materially and +adversely affects the operation of the network or violates the rules and +protocols for communication across the network. + + Corresponding Source conveyed, and Installation Information provided, +in accord with this section must be in a format that is publicly +documented (and with an implementation available to the public in +source code form), and must require no special password or key for +unpacking, reading or copying. + + 7. Additional Terms. + + "Additional permissions" are terms that supplement the terms of this +License by making exceptions from one or more of its conditions. +Additional permissions that are applicable to the entire Program shall +be treated as though they were included in this License, to the extent +that they are valid under applicable law. If additional permissions +apply only to part of the Program, that part may be used separately +under those permissions, but the entire Program remains governed by +this License without regard to the additional permissions. + + When you convey a copy of a covered work, you may at your option +remove any additional permissions from that copy, or from any part of +it. (Additional permissions may be written to require their own +removal in certain cases when you modify the work.) You may place +additional permissions on material, added by you to a covered work, +for which you have or can give appropriate copyright permission. + + Notwithstanding any other provision of this License, for material you +add to a covered work, you may (if authorized by the copyright holders of +that material) supplement the terms of this License with terms: + + a) Disclaiming warranty or limiting liability differently from the + terms of sections 15 and 16 of this License; or + + b) Requiring preservation of specified reasonable legal notices or + author attributions in that material or in the Appropriate Legal + Notices displayed by works containing it; or + + c) Prohibiting misrepresentation of the origin of that material, or + requiring that modified versions of such material be marked in + reasonable ways as different from the original version; or + + d) Limiting the use for publicity purposes of names of licensors or + authors of the material; or + + e) Declining to grant rights under trademark law for use of some + trade names, trademarks, or service marks; or + + f) Requiring indemnification of licensors and authors of that + material by anyone who conveys the material (or modified versions of + it) with contractual assumptions of liability to the recipient, for + any liability that these contractual assumptions directly impose on + those licensors and authors. + + All other non-permissive additional terms are considered "further +restrictions" within the meaning of section 10. If the Program as you +received it, or any part of it, contains a notice stating that it is +governed by this License along with a term that is a further +restriction, you may remove that term. If a license document contains +a further restriction but permits relicensing or conveying under this +License, you may add to a covered work material governed by the terms +of that license document, provided that the further restriction does +not survive such relicensing or conveying. + + If you add terms to a covered work in accord with this section, you +must place, in the relevant source files, a statement of the +additional terms that apply to those files, or a notice indicating +where to find the applicable terms. + + Additional terms, permissive or non-permissive, may be stated in the +form of a separately written license, or stated as exceptions; +the above requirements apply either way. + + 8. Termination. + + You may not propagate or modify a covered work except as expressly +provided under this License. Any attempt otherwise to propagate or +modify it is void, and will automatically terminate your rights under +this License (including any patent licenses granted under the third +paragraph of section 11). + + However, if you cease all violation of this License, then your +license from a particular copyright holder is reinstated (a) +provisionally, unless and until the copyright holder explicitly and +finally terminates your license, and (b) permanently, if the copyright +holder fails to notify you of the violation by some reasonable means +prior to 60 days after the cessation. + + Moreover, your license from a particular copyright holder is +reinstated permanently if the copyright holder notifies you of the +violation by some reasonable means, this is the first time you have +received notice of violation of this License (for any work) from that +copyright holder, and you cure the violation prior to 30 days after +your receipt of the notice. + + Termination of your rights under this section does not terminate the +licenses of parties who have received copies or rights from you under +this License. If your rights have been terminated and not permanently +reinstated, you do not qualify to receive new licenses for the same +material under section 10. + + 9. Acceptance Not Required for Having Copies. + + You are not required to accept this License in order to receive or +run a copy of the Program. Ancillary propagation of a covered work +occurring solely as a consequence of using peer-to-peer transmission +to receive a copy likewise does not require acceptance. However, +nothing other than this License grants you permission to propagate or +modify any covered work. These actions infringe copyright if you do +not accept this License. Therefore, by modifying or propagating a +covered work, you indicate your acceptance of this License to do so. + + 10. Automatic Licensing of Downstream Recipients. + + Each time you convey a covered work, the recipient automatically +receives a license from the original licensors, to run, modify and +propagate that work, subject to this License. You are not responsible +for enforcing compliance by third parties with this License. + + An "entity transaction" is a transaction transferring control of an +organization, or substantially all assets of one, or subdividing an +organization, or merging organizations. If propagation of a covered +work results from an entity transaction, each party to that +transaction who receives a copy of the work also receives whatever +licenses to the work the party's predecessor in interest had or could +give under the previous paragraph, plus a right to possession of the +Corresponding Source of the work from the predecessor in interest, if +the predecessor has it or can get it with reasonable efforts. + + You may not impose any further restrictions on the exercise of the +rights granted or affirmed under this License. For example, you may +not impose a license fee, royalty, or other charge for exercise of +rights granted under this License, and you may not initiate litigation +(including a cross-claim or counterclaim in a lawsuit) alleging that +any patent claim is infringed by making, using, selling, offering for +sale, or importing the Program or any portion of it. + + 11. Patents. + + A "contributor" is a copyright holder who authorizes use under this +License of the Program or a work on which the Program is based. The +work thus licensed is called the contributor's "contributor version". + + A contributor's "essential patent claims" are all patent claims +owned or controlled by the contributor, whether already acquired or +hereafter acquired, that would be infringed by some manner, permitted +by this License, of making, using, or selling its contributor version, +but do not include claims that would be infringed only as a +consequence of further modification of the contributor version. For +purposes of this definition, "control" includes the right to grant +patent sublicenses in a manner consistent with the requirements of +this License. + + Each contributor grants you a non-exclusive, worldwide, royalty-free +patent license under the contributor's essential patent claims, to +make, use, sell, offer for sale, import and otherwise run, modify and +propagate the contents of its contributor version. + + In the following three paragraphs, a "patent license" is any express +agreement or commitment, however denominated, not to enforce a patent +(such as an express permission to practice a patent or covenant not to +sue for patent infringement). To "grant" such a patent license to a +party means to make such an agreement or commitment not to enforce a +patent against the party. + + If you convey a covered work, knowingly relying on a patent license, +and the Corresponding Source of the work is not available for anyone +to copy, free of charge and under the terms of this License, through a +publicly available network server or other readily accessible means, +then you must either (1) cause the Corresponding Source to be so +available, or (2) arrange to deprive yourself of the benefit of the +patent license for this particular work, or (3) arrange, in a manner +consistent with the requirements of this License, to extend the patent +license to downstream recipients. "Knowingly relying" means you have +actual knowledge that, but for the patent license, your conveying the +covered work in a country, or your recipient's use of the covered work +in a country, would infringe one or more identifiable patents in that +country that you have reason to believe are valid. + + If, pursuant to or in connection with a single transaction or +arrangement, you convey, or propagate by procuring conveyance of, a +covered work, and grant a patent license to some of the parties +receiving the covered work authorizing them to use, propagate, modify +or convey a specific copy of the covered work, then the patent license +you grant is automatically extended to all recipients of the covered +work and works based on it. + + A patent license is "discriminatory" if it does not include within +the scope of its coverage, prohibits the exercise of, or is +conditioned on the non-exercise of one or more of the rights that are +specifically granted under this License. You may not convey a covered +work if you are a party to an arrangement with a third party that is +in the business of distributing software, under which you make payment +to the third party based on the extent of your activity of conveying +the work, and under which the third party grants, to any of the +parties who would receive the covered work from you, a discriminatory +patent license (a) in connection with copies of the covered work +conveyed by you (or copies made from those copies), or (b) primarily +for and in connection with specific products or compilations that +contain the covered work, unless you entered into that arrangement, +or that patent license was granted, prior to 28 March 2007. + + Nothing in this License shall be construed as excluding or limiting +any implied license or other defenses to infringement that may +otherwise be available to you under applicable patent law. + + 12. No Surrender of Others' Freedom. + + If conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot convey a +covered work so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you may +not convey it at all. For example, if you agree to terms that obligate you +to collect a royalty for further conveying from those to whom you convey +the Program, the only way you could satisfy both those terms and this +License would be to refrain entirely from conveying the Program. + + 13. Use with the GNU Affero General Public License. + + Notwithstanding any other provision of this License, you have +permission to link or combine any covered work with a work licensed +under version 3 of the GNU Affero General Public License into a single +combined work, and to convey the resulting work. The terms of this +License will continue to apply to the part which is the covered work, +but the special requirements of the GNU Affero General Public License, +section 13, concerning interaction through a network will apply to the +combination as such. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + + Each version is given a distinguishing version number. If the +Program specifies that a certain numbered version of the GNU General +Public License "or any later version" applies to it, you have the +option of following the terms and conditions either of that numbered +version or of any later version published by the Free Software +Foundation. If the Program does not specify a version number of the +GNU General Public License, you may choose any version ever published +by the Free Software Foundation. + + If the Program specifies that a proxy can decide which future +versions of the GNU General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . + +Also add information on how to contact you by electronic and paper mail. + + If the program does terminal interaction, make it output a short +notice like this when it starts in an interactive mode: + + Copyright (C) + This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, your program's commands +might be different; for a GUI interface, you would use an "about box". + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU GPL, see +. + + The GNU General Public License does not permit incorporating your program +into proprietary programs. If your program is a subroutine library, you +may consider it more useful to permit linking proprietary applications with +the library. If this is what you want to do, use the GNU Lesser General +Public License instead of this License. But first, please read +. diff --git a/README.md b/README.md new file mode 100644 index 0000000..c3cd839 --- /dev/null +++ b/README.md @@ -0,0 +1,2 @@ +# ChatWars Spy +Ищет пользователей ChatWars из информации о дуэлях, предложениях и сделках на бирже. \ No newline at end of file diff --git a/cmd/main/main.go b/cmd/main/main.go new file mode 100644 index 0000000..44385c3 --- /dev/null +++ b/cmd/main/main.go @@ -0,0 +1,170 @@ +package main + +import ( + "context" + "encoding/json" + "gitea.russia9.dev/Russia9/chatwars-spy/internal/utils" + "gitea.russia9.dev/Russia9/chatwars-spy/pkg/domain" + "gitea.russia9.dev/Russia9/chatwars-spy/pkg/message" + "gitea.russia9.dev/Russia9/chatwars-spy/pkg/storage/mongodb" + "github.com/confluentinc/confluent-kafka-go/kafka" + "github.com/google/uuid" + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" + "go.mongodb.org/mongo-driver/mongo" + "go.mongodb.org/mongo-driver/mongo/options" + "os" + "strconv" +) + +func main() { + // Log settings + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + pretty, err := strconv.ParseBool(os.Getenv("LOG_PRETTY")) + if err != nil { + pretty = false + } + if pretty { + log.Logger = log.Output(zerolog.ConsoleWriter{Out: os.Stderr}) + } + switch os.Getenv("LOG_LEVEL") { + case "DISABLED": + zerolog.SetGlobalLevel(zerolog.Disabled) + case "PANIC": + zerolog.SetGlobalLevel(zerolog.PanicLevel) + case "FATAL": + zerolog.SetGlobalLevel(zerolog.FatalLevel) + case "ERROR": + zerolog.SetGlobalLevel(zerolog.ErrorLevel) + case "WARN": + zerolog.SetGlobalLevel(zerolog.WarnLevel) + case "DEBUG": + zerolog.SetGlobalLevel(zerolog.DebugLevel) + case "TRACE": + zerolog.SetGlobalLevel(zerolog.TraceLevel) + default: + zerolog.SetGlobalLevel(zerolog.InfoLevel) + } + + version := utils.GetEnv("CHATWARS_VERSION", "cw3") + + // Kafka consumer init + consumer, err := kafka.NewConsumer(&kafka.ConfigMap{ + "bootstrap.servers": utils.GetEnv("KAFKA_ADDRESS", "digest-api.chtwrs.com:9092"), + "group.id": version + "-" + uuid.New().String(), + "auto.offset.reset": "latest", + }) + if err != nil { + log.Fatal().Err(err).Send() + } + + // DB init + client, err := mongo.Connect(context.Background(), options.Client().ApplyURI(os.Getenv("MONGO_URI"))) + if err != nil { + log.Fatal().Err(err).Send() + } + defer client.Disconnect(context.Background()) + err = client.Ping(context.Background(), nil) + if err != nil { + log.Fatal().Err(err).Send() + } + db := client.Database(utils.GetEnv("MONGO_DB", "chatwars")) + repo := mongodb.NewUserRepo(db) + + // Kafka subscribe + err = consumer.SubscribeTopics([]string{version + "-offers", version + "-deals", version + "-duels"}, nil) + if err != nil { + log.Fatal().Err(err).Send() + } + + log.Info().Msg("Started watching") + + // Get messages + for { + // Read message + raw, err := consumer.ReadMessage(-1) + if err != nil { + log.Error().Str("source", "consumer").Err(err).Send() + continue + } + + log.Trace().Bytes("message", raw.Value).Msg("received message") + + switch *raw.TopicPartition.Topic { + case version + "-offers": // Offers + // Decode message + var msg message.Offer + err = json.Unmarshal(raw.Value, &msg) + if err != nil { + log.Warn().Err(err).Str("topic", *raw.TopicPartition.Topic).Send() + continue + } + + // Store user + _ = repo.Store(context.Background(), + &domain.User{ + ID: msg.SellerID, + Castle: msg.SellerCastle, + Name: msg.SellerName, + Source: "offers", + }, + ) + case version + "-duels": // Duels + // Decode message + var msg message.Duel + err = json.Unmarshal(raw.Value, &msg) + if err != nil { + log.Warn().Err(err).Str("topic", *raw.TopicPartition.Topic).Send() + continue + } + + // Store users + _ = repo.Store(context.Background(), + &domain.User{ + ID: msg.Winner.ID, + Castle: msg.Winner.Castle, + Guild: msg.Winner.Tag, + Name: msg.Winner.Name, + Level: msg.Winner.Level, + Source: "duels", + }, + ) + _ = repo.Store(context.Background(), + &domain.User{ + ID: msg.Loser.ID, + Castle: msg.Loser.Castle, + Guild: msg.Loser.Tag, + Name: msg.Loser.Name, + Level: msg.Loser.Level, + Source: "duels", + }, + ) + case version + "-deals": // Deals + // Decode message + var msg message.Deal + err = json.Unmarshal(raw.Value, &msg) + if err != nil { + log.Warn().Err(err).Str("topic", *raw.TopicPartition.Topic).Send() + continue + } + + // Store users + _ = repo.Store(context.Background(), + &domain.User{ + ID: msg.SellerID, + Castle: msg.SellerCastle, + Name: msg.SellerName, + Source: "deals", + }, + ) + _ = repo.Store(context.Background(), + &domain.User{ + ID: msg.BuyerID, + Castle: msg.BuyerCastle, + Name: msg.BuyerName, + Source: "deals", + }, + ) + } + } +} diff --git a/go.mod b/go.mod new file mode 100644 index 0000000..4291930 --- /dev/null +++ b/go.mod @@ -0,0 +1,26 @@ +module gitea.russia9.dev/Russia9/chatwars-spy + +go 1.18 + +require ( + github.com/confluentinc/confluent-kafka-go v1.8.2 + github.com/google/uuid v1.3.0 + github.com/rs/zerolog v1.26.1 + go.mongodb.org/mongo-driver v1.9.0 +) + +require ( + github.com/go-stack/stack v1.8.1 // indirect + github.com/golang/snappy v0.0.4 // indirect + github.com/klauspost/compress v1.15.1 // indirect + github.com/pkg/errors v0.9.1 // indirect + github.com/stretchr/testify v1.7.0 // indirect + github.com/xdg-go/pbkdf2 v1.0.0 // indirect + github.com/xdg-go/scram v1.1.1 // indirect + github.com/xdg-go/stringprep v1.0.3 // indirect + github.com/youmark/pkcs8 v0.0.0-20201027041543-1326539a0a0a // indirect + golang.org/x/crypto v0.0.0-20220411220226-7b82a4e95df4 // indirect + golang.org/x/sync v0.0.0-20210220032951-036812b2e83c // indirect + golang.org/x/text v0.3.7 // indirect + gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b // indirect +) diff --git a/go.sum b/go.sum new file mode 100644 index 0000000..50465fc --- /dev/null +++ b/go.sum @@ -0,0 +1,96 @@ +github.com/confluentinc/confluent-kafka-go v1.8.2 h1:PBdbvYpyOdFLehj8j+9ba7FL4c4Moxn79gy9cYKxG5E= +github.com/confluentinc/confluent-kafka-go v1.8.2/go.mod h1:u2zNLny2xq+5rWeTQjFHbDzzNuba4P1vo31r9r4uAdg= +github.com/coreos/go-systemd/v22 v22.3.2/go.mod h1:Y58oyj3AT4RCenI/lSvhwexgC+NSVTIJ3seZv2GcEnc= +github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38= +github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c= +github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38= +github.com/go-stack/stack v1.8.0/go.mod h1:v0f6uXyyMGvRgIKkXu+yp6POWl0qKG85gN/melR3HDY= +github.com/go-stack/stack v1.8.1 h1:ntEHSVwIt7PNXNpgPmVfMrNhLtgjlmnZha2kOpuRiDw= +github.com/go-stack/stack v1.8.1/go.mod h1:dcoOX6HbPZSZptuspn9bctJ+N/CnF5gGygcUP3XYfe4= +github.com/godbus/dbus/v5 v5.0.4/go.mod h1:xhWf0FNVPg57R7Z0UbKHbJfkEywrmjJnf7w5xrFpKfA= +github.com/golang/snappy v0.0.1/go.mod h1:/XxbfmMg8lxefKM7IXC3fBNl/7bRcc72aCRzEWrmP2Q= +github.com/golang/snappy v0.0.4 h1:yAGX7huGHXlcLOEtBnF4w7FQwA26wojNCwOYAEhLjQM= +github.com/golang/snappy v0.0.4/go.mod h1:/XxbfmMg8lxefKM7IXC3fBNl/7bRcc72aCRzEWrmP2Q= +github.com/google/go-cmp v0.5.2 h1:X2ev0eStA3AbceY54o37/0PQ/UWqKEiiO2dKL5OPaFM= +github.com/google/go-cmp v0.5.2/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE= +github.com/google/uuid v1.3.0 h1:t6JiXgmwXMjEs8VusXIJk2BXHsn+wx8BZdTaoZ5fu7I= +github.com/google/uuid v1.3.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo= +github.com/klauspost/compress v1.13.6/go.mod h1:/3/Vjq9QcHkK5uEr5lBEmyoZ1iFhe47etQ6QUkpK6sk= +github.com/klauspost/compress v1.15.1 h1:y9FcTHGyrebwfP0ZZqFiaxTaiDnUrGkJkI+f583BL1A= +github.com/klauspost/compress v1.15.1/go.mod h1:/3/Vjq9QcHkK5uEr5lBEmyoZ1iFhe47etQ6QUkpK6sk= +github.com/kr/pretty v0.1.0/go.mod h1:dAy3ld7l9f0ibDNOQOHHMYYIIbhfbHSm3C4ZsoJORNo= +github.com/kr/pty v1.1.1/go.mod h1:pFQYn66WHrOpPYNljwOMqo10TkYh1fy3cYio2l3bCsQ= +github.com/kr/text v0.1.0/go.mod h1:4Jbv+DJW3UT/LiOwJeYQe1efqtUx/iVham/4vfdArNI= +github.com/montanaflynn/stats v0.0.0-20171201202039-1bf9dbcd8cbe/go.mod h1:wL8QJuTMNUDYhXwkmfOly8iTdp5TEcJFWZD2D7SIkUc= +github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4= +github.com/pkg/errors v0.9.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0= +github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM= +github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4= +github.com/rs/xid v1.3.0/go.mod h1:trrq9SKmegXys3aeAKXMUTdJsYXVwGY3RLcfgqegfbg= +github.com/rs/zerolog v1.26.1 h1:/ihwxqH+4z8UxyI70wM1z9yCvkWcfz/a3mj48k/Zngc= +github.com/rs/zerolog v1.26.1/go.mod h1:/wSSJWX7lVrsOwlbyTRSOJvqRlc+WjWlfes+CiJ+tmc= +github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME= +github.com/stretchr/testify v1.6.1/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg= +github.com/stretchr/testify v1.7.0 h1:nwc3DEeHmmLAfoZucVR881uASk0Mfjw8xYJ99tb5CcY= +github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg= +github.com/tidwall/pretty v1.0.0 h1:HsD+QiTn7sK6flMKIvNmpqz1qrpP3Ps6jOKIKMooyg4= +github.com/tidwall/pretty v1.0.0/go.mod h1:XNkn88O1ChpSDQmQeStsy+sBenx6DDtFZJxhVysOjyk= +github.com/xdg-go/pbkdf2 v1.0.0 h1:Su7DPu48wXMwC3bs7MCNG+z4FhcyEuz5dlvchbq0B0c= +github.com/xdg-go/pbkdf2 v1.0.0/go.mod h1:jrpuAogTd400dnrH08LKmI/xc1MbPOebTwRqcT5RDeI= +github.com/xdg-go/scram v1.0.2/go.mod h1:1WAq6h33pAW+iRreB34OORO2Nf7qel3VV3fjBj+hCSs= +github.com/xdg-go/scram v1.1.1 h1:VOMT+81stJgXW3CpHyqHN3AXDYIMsx56mEFrB37Mb/E= +github.com/xdg-go/scram v1.1.1/go.mod h1:RaEWvsqvNKKvBPvcKeFjrG2cJqOkHTiyTpzz23ni57g= +github.com/xdg-go/stringprep v1.0.2/go.mod h1:8F9zXuvzgwmyT5DUm4GUfZGDdT3W+LCvS6+da4O5kxM= +github.com/xdg-go/stringprep v1.0.3 h1:kdwGpVNwPFtjs98xCGkHjQtGKh86rDcRZN17QEMCOIs= +github.com/xdg-go/stringprep v1.0.3/go.mod h1:W3f5j4i+9rC0kuIEJL0ky1VpHXQU3ocBgklLGvcBnW8= +github.com/youmark/pkcs8 v0.0.0-20181117223130-1be2e3e5546d/go.mod h1:rHwXgn7JulP+udvsHwJoVG1YGAP6VLg4y9I5dyZdqmA= +github.com/youmark/pkcs8 v0.0.0-20201027041543-1326539a0a0a h1:fZHgsYlfvtyqToslyjUt3VOPF4J7aK/3MPcK7xp3PDk= +github.com/youmark/pkcs8 v0.0.0-20201027041543-1326539a0a0a/go.mod h1:ul22v+Nro/R083muKhosV54bj5niojjWZvU8xrevuH4= +github.com/yuin/goldmark v1.4.0/go.mod h1:mwnBkeHKe2W/ZEtQ+71ViKU8L12m81fl3OWwC1Zlc8k= +go.mongodb.org/mongo-driver v1.9.0 h1:f3aLGJvQmBl8d9S40IL+jEyBC6hfLPbJjv9t5hEM9ck= +go.mongodb.org/mongo-driver v1.9.0/go.mod h1:0sQWfOeY63QTntERDJJ/0SuKK0T1uVSgKCuAROlKEPY= +golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w= +golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI= +golang.org/x/crypto v0.0.0-20200302210943-78000ba7a073/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto= +golang.org/x/crypto v0.0.0-20201216223049-8b5274cf687f/go.mod h1:jdWPYTVW3xRLrWPugEBEK3UY2ZEsg3UU495nc5E+M+I= +golang.org/x/crypto v0.0.0-20211215165025-cf75a172585e/go.mod h1:P+XmwS30IXTQdn5tA2iutPOUgjI07+tq3H3K9MVA1s8= +golang.org/x/crypto v0.0.0-20220411220226-7b82a4e95df4 h1:kUhD7nTDoI3fVd9G4ORWrbV5NY0liEs/Jg2pv5f+bBA= +golang.org/x/crypto v0.0.0-20220411220226-7b82a4e95df4/go.mod h1:IxCIyHEi3zRg3s0A5j5BB6A9Jmi73HwBIUl50j+osU4= +golang.org/x/mod v0.4.2/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA= +golang.org/x/net v0.0.0-20190311183353-d8887717615a/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg= +golang.org/x/net v0.0.0-20190404232315-eb5bcb51f2a3/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg= +golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s= +golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg= +golang.org/x/net v0.0.0-20210805182204-aaa1db679c0d/go.mod h1:9nx3DQGgdP8bBQD5qxJ1jj9UTztislL4KSBs9R2vV5Y= +golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM= +golang.org/x/sync v0.0.0-20190911185100-cd5d95a43a6e/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM= +golang.org/x/sync v0.0.0-20210220032951-036812b2e83c h1:5KslGYwFpkhGh+Q16bwMP3cOontH8FOep7tGV86Y7SQ= +golang.org/x/sync v0.0.0-20210220032951-036812b2e83c/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM= +golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY= +golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= +golang.org/x/sys v0.0.0-20191026070338-33540a1f6037/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= +golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= +golang.org/x/sys v0.0.0-20210423082822-04245dca01da/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= +golang.org/x/sys v0.0.0-20210809222454-d867a43fc93e/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg= +golang.org/x/term v0.0.0-20201117132131-f5c789dd3221/go.mod h1:Nr5EML6q2oocZ2LXRh80K7BxOlk5/8JxuGnuhpl+muw= +golang.org/x/term v0.0.0-20201126162022-7de9c90e9dd1/go.mod h1:bj7SfCRtBDWHUb9snDiAeCFNEtKQo2Wmx5Cou7ajbmo= +golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ= +golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ= +golang.org/x/text v0.3.5/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ= +golang.org/x/text v0.3.6/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ= +golang.org/x/text v0.3.7 h1:olpwvP2KacW1ZWvsR7uQhoyTYvKAupfQrRGBFM352Gk= +golang.org/x/text v0.3.7/go.mod h1:u+2+/6zg+i71rQMx5EYifcz6MCKuco9NR6JIITiCfzQ= +golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ= +golang.org/x/tools v0.0.0-20190531172133-b3315ee88b7d/go.mod h1:/rFqwRUd4F7ZHNgwSSTFct+R/Kf4OFW1sUzUTQQTgfc= +golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo= +golang.org/x/tools v0.1.7/go.mod h1:LGqMHiF4EqQNHR1JncWGqT5BVaXmza+X+BDGol+dOxo= +golang.org/x/xerrors v0.0.0-20190717185122-a985d3407aa7/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0= +golang.org/x/xerrors v0.0.0-20191011141410-1b5146add898/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0= +golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0= +golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 h1:go1bK/D/BFZV2I8cIQd1NKEZ+0owSTG1fDTci4IqFcE= +golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0= +gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0= +gopkg.in/check.v1 v1.0.0-20180628173108-788fd7840127/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0= +gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM= +gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b h1:h8qDotaEPuJATrMmW04NCwg7v22aHH28wwpauUhK9Oo= +gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM= diff --git a/internal/utils/common.go b/internal/utils/common.go new file mode 100644 index 0000000..282a613 --- /dev/null +++ b/internal/utils/common.go @@ -0,0 +1,10 @@ +package utils + +import "os" + +func GetEnv(key, fallback string) string { + if value, ok := os.LookupEnv(key); ok { + return value + } + return fallback +} diff --git a/pkg/domain/user.go b/pkg/domain/user.go new file mode 100644 index 0000000..0231431 --- /dev/null +++ b/pkg/domain/user.go @@ -0,0 +1,22 @@ +package domain + +import ( + "context" + "time" +) + +type User struct { + ID string `bson:"_id"` + Castle string + Guild string + Name string + Level int + Source string + + FirstSeen time.Time + LastSeen time.Time +} + +type UserRepo interface { + Store(ctx context.Context, object *User) error +} diff --git a/pkg/message/deal.go b/pkg/message/deal.go new file mode 100644 index 0000000..44dfa82 --- /dev/null +++ b/pkg/message/deal.go @@ -0,0 +1,13 @@ +package message + +type Deal struct { + SellerID string `json:"sellerId"` + SellerCastle string `json:"sellerCastle"` + SellerName string `json:"sellerName"` + BuyerID string `json:"buyerId"` + BuyerCastle string `json:"buyerCastle"` + BuyerName string `json:"buyerName"` + Item string `json:"item"` + Quantity int `json:"qty"` + Price int `json:"price"` +} diff --git a/pkg/message/duel.go b/pkg/message/duel.go new file mode 100644 index 0000000..c365c97 --- /dev/null +++ b/pkg/message/duel.go @@ -0,0 +1,17 @@ +package message + +type Duel struct { + Winner DuelUser `json:"winner"` + Loser DuelUser `json:"loser"` + IsChallenge bool `json:"isChallenge"` + IsGuildDuel bool `json:"isGuildDuel"` +} + +type DuelUser struct { + ID string `json:"id"` + Name string `json:"name"` + Tag string `json:"tag,omitempty"` + Castle string `json:"castle"` + Level int `json:"level"` + Health int `json:"hp"` +} diff --git a/pkg/message/lot.go b/pkg/message/lot.go new file mode 100644 index 0000000..17e90cd --- /dev/null +++ b/pkg/message/lot.go @@ -0,0 +1,20 @@ +package message + +type Lot struct { + LotID string `json:"lotId"` + ItemName string `json:"itemName"` + SellerTag string `json:"sellerTag"` + SellerName string `json:"sellerName"` + Quality string `json:"quality"` + SellerCastle string `json:"sellerCastle"` + Condition string `json:"condition"` + EndAt string `json:"endAt"` + StartAt string `json:"startedAt"` + BuyerCastle string `json:"buyerCastle"` + Status string `json:"status"` + FinishedAt string `json:"finishedAt"` + BuyerTag string `json:"buyerTag"` + BuyerName string `json:"buyerName"` + Price int `json:"price"` + Stats map[string]int +} diff --git a/pkg/message/offer.go b/pkg/message/offer.go new file mode 100644 index 0000000..644baef --- /dev/null +++ b/pkg/message/offer.go @@ -0,0 +1,10 @@ +package message + +type Offer struct { + SellerID string `json:"sellerId"` + SellerName string `json:"sellerName"` + SellerCastle string `json:"sellerCastle"` + Item string `json:"item"` + Quantity int `json:"qty"` + Price int `json:"price"` +} diff --git a/pkg/message/store.go b/pkg/message/store.go new file mode 100644 index 0000000..dfce0b7 --- /dev/null +++ b/pkg/message/store.go @@ -0,0 +1,23 @@ +package message + +type Store struct { + Link string `json:"link"` + Name string `json:"name"` + OwnerName string `json:"ownerName"` + OwnerCastle string `json:"ownerCastle"` + Kind string `json:"kind"` + Mana int `json:"mana"` + Offers []StoreOffer `json:"offers"` + Specialization map[string]int `json:"specialization,omitempty"` + QualityCraftLevel int `json:"qualityCraftLevel"` + MaintenanceEnabled bool `json:"maintenanceEnabled"` + MaintenanceCost int `json:"maintenanceCost"` + GuildDiscount int `json:"guildDiscount"` + CastleDiscount int `json:"castleDiscount"` +} + +type StoreOffer struct { + Item string `json:"item"` + Price int `json:"price"` + Mana int `json:"mana"` +} diff --git a/pkg/storage/mongodb/user.go b/pkg/storage/mongodb/user.go new file mode 100644 index 0000000..ba85bd3 --- /dev/null +++ b/pkg/storage/mongodb/user.go @@ -0,0 +1,52 @@ +package mongodb + +import ( + "context" + "errors" + "gitea.russia9.dev/Russia9/chatwars-spy/pkg/domain" + "go.mongodb.org/mongo-driver/bson" + "go.mongodb.org/mongo-driver/mongo" + "time" +) + +func NewUserRepo(database *mongo.Database) domain.UserRepo { + return &userRepo{Collection: database.Collection("users")} +} + +type userRepo struct { + Collection *mongo.Collection +} + +func (r *userRepo) Store(ctx context.Context, object *domain.User) error { + cursor := r.Collection.FindOne(ctx, bson.M{"_id": object.ID}) + if errors.Is(cursor.Err(), mongo.ErrNoDocuments) { + object.FirstSeen = time.Now() + object.LastSeen = time.Now() + _, err := r.Collection.InsertOne(ctx, object) + return err + } else if cursor.Err() != nil { + return cursor.Err() + } + + // Get old user + var old domain.User + err := cursor.Decode(&old) + if err != nil { + return err + } + + // Get old guild if empty + if object.Guild == "" { + object.Guild = old.Guild + } + + // Set old source + object.Source = old.Source + + // Set LastSeen date + object.LastSeen = time.Now() + + // Update object in DB + _, err = r.Collection.ReplaceOne(ctx, bson.M{"_id": object.ID}, object) + return err +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/LICENSE b/vendor/github.com/confluentinc/confluent-kafka-go/LICENSE new file mode 100644 index 0000000..e06d208 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/LICENSE @@ -0,0 +1,202 @@ +Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "{}" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright {yyyy} {name of copyright owner} + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. + diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/.gitignore b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/.gitignore new file mode 100644 index 0000000..b1a2111 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/.gitignore @@ -0,0 +1,2 @@ +testconf.json +go_rdkafka_generr/go_rdkafka_generr diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/00version.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/00version.go new file mode 100644 index 0000000..8c2c7b4 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/00version.go @@ -0,0 +1,58 @@ +package kafka + +/** + * Copyright 2016-2019 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "fmt" +) + +/* +#include "select_rdkafka.h" + +//Minimum required librdkafka version. This is checked both during +//build-time and runtime. +//Make sure to keep the MIN_RD_KAFKA_VERSION, MIN_VER_ERRSTR and #error +//defines and strings in sync. +// + +#define MIN_RD_KAFKA_VERSION 0x01060000 + +#ifdef __APPLE__ +#define MIN_VER_ERRSTR "confluent-kafka-go requires librdkafka v1.6.0 or later. Install the latest version of librdkafka from Homebrew by running `brew install librdkafka` or `brew upgrade librdkafka`" +#else +#define MIN_VER_ERRSTR "confluent-kafka-go requires librdkafka v1.6.0 or later. Install the latest version of librdkafka from the Confluent repositories, see http://docs.confluent.io/current/installation.html" +#endif + +#if RD_KAFKA_VERSION < MIN_RD_KAFKA_VERSION +#ifdef __APPLE__ +#error "confluent-kafka-go requires librdkafka v1.6.0 or later. Install the latest version of librdkafka from Homebrew by running `brew install librdkafka` or `brew upgrade librdkafka`" +#else +#error "confluent-kafka-go requires librdkafka v1.6.0 or later. Install the latest version of librdkafka from the Confluent repositories, see http://docs.confluent.io/current/installation.html" +#endif +#endif +*/ +import "C" + +func versionCheck() error { + ver, verstr := LibraryVersion() + if ver < C.MIN_RD_KAFKA_VERSION { + return newErrorFromString(ErrNotImplemented, + fmt.Sprintf("%s: librdkafka version %s (0x%x) detected", + C.MIN_VER_ERRSTR, verstr, ver)) + } + return nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/README.md b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/README.md new file mode 100644 index 0000000..98152e9 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/README.md @@ -0,0 +1,145 @@ +# Information for confluent-kafka-go developers + +Whenever librdkafka error codes are updated make sure to run generate +before building: + +``` + $ make -f mk/Makefile generr + $ go build ./... +``` + + + + +## Testing + +Some of the tests included in this directory, the benchmark and integration tests in particular, +require an existing Kafka cluster and a testconf.json configuration file to +provide tests with bootstrap brokers, topic name, etc. + +The format of testconf.json is a JSON object: +``` +{ + "Brokers": "", + "Topic": "" +} +``` + +See testconf-example.json for an example and full set of available options. + + +To run unit-tests: +``` +$ go test +``` + +To run benchmark tests: +``` +$ go test -bench . +``` + +For the code coverage: +``` +$ go test -coverprofile=coverage.out -bench=. +$ go tool cover -func=coverage.out +``` + + +## Build tags + +Different build types are supported through Go build tags (`-tags ..`), +these tags should be specified on the **application** build/get/install command. + + * By default the bundled platform-specific static build of librdkafka will + be used. This works out of the box on Mac OSX and glibc-based Linux distros, + such as Ubuntu and CentOS. + * `-tags musl` - must be specified when building on/for musl-based Linux + distros, such as Alpine. Will use the bundled static musl build of + librdkafka. + * `-tags dynamic` - link librdkafka dynamically. A shared librdkafka library + must be installed manually through other means (apt-get, yum, build from + source, etc). + + + +## Generating HTML documentation + +To generate one-page HTML documentation run the mk/doc-gen.py script from the +top-level directory. This script requires the beautifulsoup4 Python package. + +``` +$ source .../your/virtualenv/bin/activate +$ pip install beautifulsoup4 +... +$ make -f mk/Makefile docs +``` + + +## Release process + +For each release candidate and final release, perform the following steps: + +### Update bundle to latest librdkafka + +See instructions in [kafka/librdkafka/README.md](kafka/librdkafka/README.md). + + +### Update librdkafka version requirement + +Update the minimum required librdkafka version in `kafka/00version.go` +and `README.md`. + + +### Update error codes + +Error codes can be automatically generated from the current librdkafka version. + + +Update generated error codes: + + $ make -f mk/Makefile generr + # Verify by building + + +### Rebuild everything + + $ go clean -i ./... + $ go build ./... + + +### Run full test suite + +Set up a test cluster using whatever mechanism you typically use +(docker, trivup, ccloud, ..). + +Make sure to update `kafka/testconf.json` as needed (broker list, $BROKERS) + +Run test suite: + + $ go test ./... + + +### Verify examples + +Manually verify that the examples/ applications work. + +Also make sure the examples in README.md work. + +Convert any examples using `github.com/confluentinc/confluent-kafka-go/kafka` to use +`gopkg.in/confluentinc/confluent-kafka-go.v1/kafka` import path. + + $ find examples/ -type f -name *\.go -exec sed -i -e 's|github\.com/confluentinc/confluent-kafka-go/kafka|gopkg\.in/confluentinc/confluent-kafka-go\.v1/kafka|g' {} + + +### Commit any changes + +Make sure to push to github before creating the tag to have CI tests pass. + + +### Create and push tag + + $ git tag v1.3.0 + $ git push --dry-run origin v1.3.0 + # Remove --dry-run and re-execute if it looks ok. + + +### Create release notes page on github diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/adminapi.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/adminapi.go new file mode 100644 index 0000000..e128129 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/adminapi.go @@ -0,0 +1,1025 @@ +/** + * Copyright 2018 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import ( + "context" + "fmt" + "strings" + "time" + "unsafe" +) + +/* +#include "select_rdkafka.h" +#include + +static const rd_kafka_topic_result_t * +topic_result_by_idx (const rd_kafka_topic_result_t **topics, size_t cnt, size_t idx) { + if (idx >= cnt) + return NULL; + return topics[idx]; +} + +static const rd_kafka_ConfigResource_t * +ConfigResource_by_idx (const rd_kafka_ConfigResource_t **res, size_t cnt, size_t idx) { + if (idx >= cnt) + return NULL; + return res[idx]; +} + +static const rd_kafka_ConfigEntry_t * +ConfigEntry_by_idx (const rd_kafka_ConfigEntry_t **entries, size_t cnt, size_t idx) { + if (idx >= cnt) + return NULL; + return entries[idx]; +} +*/ +import "C" + +// AdminClient is derived from an existing Producer or Consumer +type AdminClient struct { + handle *handle + isDerived bool // Derived from existing client handle +} + +func durationToMilliseconds(t time.Duration) int { + if t > 0 { + return (int)(t.Seconds() * 1000.0) + } + return (int)(t) +} + +// TopicResult provides per-topic operation result (error) information. +type TopicResult struct { + // Topic name + Topic string + // Error, if any, of result. Check with `Error.Code() != ErrNoError`. + Error Error +} + +// String returns a human-readable representation of a TopicResult. +func (t TopicResult) String() string { + if t.Error.code == 0 { + return t.Topic + } + return fmt.Sprintf("%s (%s)", t.Topic, t.Error.str) +} + +// TopicSpecification holds parameters for creating a new topic. +// TopicSpecification is analogous to NewTopic in the Java Topic Admin API. +type TopicSpecification struct { + // Topic name to create. + Topic string + // Number of partitions in topic. + NumPartitions int + // Default replication factor for the topic's partitions, or zero + // if an explicit ReplicaAssignment is set. + ReplicationFactor int + // (Optional) Explicit replica assignment. The outer array is + // indexed by the partition number, while the inner per-partition array + // contains the replica broker ids. The first broker in each + // broker id list will be the preferred replica. + ReplicaAssignment [][]int32 + // Topic configuration. + Config map[string]string +} + +// PartitionsSpecification holds parameters for creating additional partitions for a topic. +// PartitionsSpecification is analogous to NewPartitions in the Java Topic Admin API. +type PartitionsSpecification struct { + // Topic to create more partitions for. + Topic string + // New partition count for topic, must be higher than current partition count. + IncreaseTo int + // (Optional) Explicit replica assignment. The outer array is + // indexed by the new partition index (i.e., 0 for the first added + // partition), while the inner per-partition array + // contains the replica broker ids. The first broker in each + // broker id list will be the preferred replica. + ReplicaAssignment [][]int32 +} + +// ResourceType represents an Apache Kafka resource type +type ResourceType int + +const ( + // ResourceUnknown - Unknown + ResourceUnknown = ResourceType(C.RD_KAFKA_RESOURCE_UNKNOWN) + // ResourceAny - match any resource type (DescribeConfigs) + ResourceAny = ResourceType(C.RD_KAFKA_RESOURCE_ANY) + // ResourceTopic - Topic + ResourceTopic = ResourceType(C.RD_KAFKA_RESOURCE_TOPIC) + // ResourceGroup - Group + ResourceGroup = ResourceType(C.RD_KAFKA_RESOURCE_GROUP) + // ResourceBroker - Broker + ResourceBroker = ResourceType(C.RD_KAFKA_RESOURCE_BROKER) +) + +// String returns the human-readable representation of a ResourceType +func (t ResourceType) String() string { + return C.GoString(C.rd_kafka_ResourceType_name(C.rd_kafka_ResourceType_t(t))) +} + +// ResourceTypeFromString translates a resource type name/string to +// a ResourceType value. +func ResourceTypeFromString(typeString string) (ResourceType, error) { + switch strings.ToUpper(typeString) { + case "ANY": + return ResourceAny, nil + case "TOPIC": + return ResourceTopic, nil + case "GROUP": + return ResourceGroup, nil + case "BROKER": + return ResourceBroker, nil + default: + return ResourceUnknown, NewError(ErrInvalidArg, "Unknown resource type", false) + } +} + +// ConfigSource represents an Apache Kafka config source +type ConfigSource int + +const ( + // ConfigSourceUnknown is the default value + ConfigSourceUnknown = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_UNKNOWN_CONFIG) + // ConfigSourceDynamicTopic is dynamic topic config that is configured for a specific topic + ConfigSourceDynamicTopic = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DYNAMIC_TOPIC_CONFIG) + // ConfigSourceDynamicBroker is dynamic broker config that is configured for a specific broker + ConfigSourceDynamicBroker = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DYNAMIC_BROKER_CONFIG) + // ConfigSourceDynamicDefaultBroker is dynamic broker config that is configured as default for all brokers in the cluster + ConfigSourceDynamicDefaultBroker = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DYNAMIC_DEFAULT_BROKER_CONFIG) + // ConfigSourceStaticBroker is static broker config provided as broker properties at startup (e.g. from server.properties file) + ConfigSourceStaticBroker = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_STATIC_BROKER_CONFIG) + // ConfigSourceDefault is built-in default configuration for configs that have a default value + ConfigSourceDefault = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DEFAULT_CONFIG) +) + +// String returns the human-readable representation of a ConfigSource type +func (t ConfigSource) String() string { + return C.GoString(C.rd_kafka_ConfigSource_name(C.rd_kafka_ConfigSource_t(t))) +} + +// ConfigResource holds parameters for altering an Apache Kafka configuration resource +type ConfigResource struct { + // Type of resource to set. + Type ResourceType + // Name of resource to set. + Name string + // Config entries to set. + // Configuration updates are atomic, any configuration property not provided + // here will be reverted (by the broker) to its default value. + // Use DescribeConfigs to retrieve the list of current configuration entry values. + Config []ConfigEntry +} + +// String returns a human-readable representation of a ConfigResource +func (c ConfigResource) String() string { + return fmt.Sprintf("Resource(%s, %s)", c.Type, c.Name) +} + +// AlterOperation specifies the operation to perform on the ConfigEntry. +// Currently only AlterOperationSet. +type AlterOperation int + +const ( + // AlterOperationSet sets/overwrites the configuration setting. + AlterOperationSet = iota +) + +// String returns the human-readable representation of an AlterOperation +func (o AlterOperation) String() string { + switch o { + case AlterOperationSet: + return "Set" + default: + return fmt.Sprintf("Unknown%d?", int(o)) + } +} + +// ConfigEntry holds parameters for altering a resource's configuration. +type ConfigEntry struct { + // Name of configuration entry, e.g., topic configuration property name. + Name string + // Value of configuration entry. + Value string + // Operation to perform on the entry. + Operation AlterOperation +} + +// StringMapToConfigEntries creates a new map of ConfigEntry objects from the +// provided string map. The AlterOperation is set on each created entry. +func StringMapToConfigEntries(stringMap map[string]string, operation AlterOperation) []ConfigEntry { + var ceList []ConfigEntry + + for k, v := range stringMap { + ceList = append(ceList, ConfigEntry{Name: k, Value: v, Operation: operation}) + } + + return ceList +} + +// String returns a human-readable representation of a ConfigEntry. +func (c ConfigEntry) String() string { + return fmt.Sprintf("%v %s=\"%s\"", c.Operation, c.Name, c.Value) +} + +// ConfigEntryResult contains the result of a single configuration entry from a +// DescribeConfigs request. +type ConfigEntryResult struct { + // Name of configuration entry, e.g., topic configuration property name. + Name string + // Value of configuration entry. + Value string + // Source indicates the configuration source. + Source ConfigSource + // IsReadOnly indicates whether the configuration entry can be altered. + IsReadOnly bool + // IsSensitive indicates whether the configuration entry contains sensitive information, in which case the value will be unset. + IsSensitive bool + // IsSynonym indicates whether the configuration entry is a synonym for another configuration property. + IsSynonym bool + // Synonyms contains a map of configuration entries that are synonyms to this configuration entry. + Synonyms map[string]ConfigEntryResult +} + +// String returns a human-readable representation of a ConfigEntryResult. +func (c ConfigEntryResult) String() string { + return fmt.Sprintf("%s=\"%s\"", c.Name, c.Value) +} + +// setFromC sets up a ConfigEntryResult from a C ConfigEntry +func configEntryResultFromC(cEntry *C.rd_kafka_ConfigEntry_t) (entry ConfigEntryResult) { + entry.Name = C.GoString(C.rd_kafka_ConfigEntry_name(cEntry)) + cValue := C.rd_kafka_ConfigEntry_value(cEntry) + if cValue != nil { + entry.Value = C.GoString(cValue) + } + entry.Source = ConfigSource(C.rd_kafka_ConfigEntry_source(cEntry)) + entry.IsReadOnly = cint2bool(C.rd_kafka_ConfigEntry_is_read_only(cEntry)) + entry.IsSensitive = cint2bool(C.rd_kafka_ConfigEntry_is_sensitive(cEntry)) + entry.IsSynonym = cint2bool(C.rd_kafka_ConfigEntry_is_synonym(cEntry)) + + var cSynCnt C.size_t + cSyns := C.rd_kafka_ConfigEntry_synonyms(cEntry, &cSynCnt) + if cSynCnt > 0 { + entry.Synonyms = make(map[string]ConfigEntryResult) + } + + for si := 0; si < int(cSynCnt); si++ { + cSyn := C.ConfigEntry_by_idx(cSyns, cSynCnt, C.size_t(si)) + Syn := configEntryResultFromC(cSyn) + entry.Synonyms[Syn.Name] = Syn + } + + return entry +} + +// ConfigResourceResult provides the result for a resource from a AlterConfigs or +// DescribeConfigs request. +type ConfigResourceResult struct { + // Type of returned result resource. + Type ResourceType + // Name of returned result resource. + Name string + // Error, if any, of returned result resource. + Error Error + // Config entries, if any, of returned result resource. + Config map[string]ConfigEntryResult +} + +// String returns a human-readable representation of a ConfigResourceResult. +func (c ConfigResourceResult) String() string { + if c.Error.Code() != 0 { + return fmt.Sprintf("ResourceResult(%s, %s, \"%v\")", c.Type, c.Name, c.Error) + + } + return fmt.Sprintf("ResourceResult(%s, %s, %d config(s))", c.Type, c.Name, len(c.Config)) +} + +// waitResult waits for a result event on cQueue or the ctx to be cancelled, whichever happens +// first. +// The returned result event is checked for errors its error is returned if set. +func (a *AdminClient) waitResult(ctx context.Context, cQueue *C.rd_kafka_queue_t, cEventType C.rd_kafka_event_type_t) (rkev *C.rd_kafka_event_t, err error) { + + resultChan := make(chan *C.rd_kafka_event_t) + closeChan := make(chan bool) // never written to, just closed + + go func() { + for { + select { + case _, ok := <-closeChan: + if !ok { + // Context cancelled/timed out + close(resultChan) + return + } + + default: + // Wait for result event for at most 50ms + // to avoid blocking for too long if + // context is cancelled. + rkev := C.rd_kafka_queue_poll(cQueue, 50) + if rkev != nil { + resultChan <- rkev + close(resultChan) + return + } + } + } + }() + + select { + case rkev = <-resultChan: + // Result type check + if cEventType != C.rd_kafka_event_type(rkev) { + err = newErrorFromString(ErrInvalidType, + fmt.Sprintf("Expected %d result event, not %d", (int)(cEventType), (int)(C.rd_kafka_event_type(rkev)))) + C.rd_kafka_event_destroy(rkev) + return nil, err + } + + // Generic error handling + cErr := C.rd_kafka_event_error(rkev) + if cErr != 0 { + err = newErrorFromCString(cErr, C.rd_kafka_event_error_string(rkev)) + C.rd_kafka_event_destroy(rkev) + return nil, err + } + close(closeChan) + return rkev, nil + case <-ctx.Done(): + // signal close to go-routine + close(closeChan) + // wait for close from go-routine to make sure it is done + // using cQueue before we return. + rkev, ok := <-resultChan + if ok { + // throw away result since context was cancelled + C.rd_kafka_event_destroy(rkev) + } + return nil, ctx.Err() + } +} + +// cToTopicResults converts a C topic_result_t array to Go TopicResult list. +func (a *AdminClient) cToTopicResults(cTopicRes **C.rd_kafka_topic_result_t, cCnt C.size_t) (result []TopicResult, err error) { + + result = make([]TopicResult, int(cCnt)) + + for i := 0; i < int(cCnt); i++ { + cTopic := C.topic_result_by_idx(cTopicRes, cCnt, C.size_t(i)) + result[i].Topic = C.GoString(C.rd_kafka_topic_result_name(cTopic)) + result[i].Error = newErrorFromCString( + C.rd_kafka_topic_result_error(cTopic), + C.rd_kafka_topic_result_error_string(cTopic)) + } + + return result, nil +} + +// cConfigResourceToResult converts a C ConfigResource result array to Go ConfigResourceResult +func (a *AdminClient) cConfigResourceToResult(cRes **C.rd_kafka_ConfigResource_t, cCnt C.size_t) (result []ConfigResourceResult, err error) { + + result = make([]ConfigResourceResult, int(cCnt)) + + for i := 0; i < int(cCnt); i++ { + cRes := C.ConfigResource_by_idx(cRes, cCnt, C.size_t(i)) + result[i].Type = ResourceType(C.rd_kafka_ConfigResource_type(cRes)) + result[i].Name = C.GoString(C.rd_kafka_ConfigResource_name(cRes)) + result[i].Error = newErrorFromCString( + C.rd_kafka_ConfigResource_error(cRes), + C.rd_kafka_ConfigResource_error_string(cRes)) + var cConfigCnt C.size_t + cConfigs := C.rd_kafka_ConfigResource_configs(cRes, &cConfigCnt) + if cConfigCnt > 0 { + result[i].Config = make(map[string]ConfigEntryResult) + } + for ci := 0; ci < int(cConfigCnt); ci++ { + cEntry := C.ConfigEntry_by_idx(cConfigs, cConfigCnt, C.size_t(ci)) + entry := configEntryResultFromC(cEntry) + result[i].Config[entry.Name] = entry + } + } + + return result, nil +} + +// ClusterID returns the cluster ID as reported in broker metadata. +// +// Note on cancellation: Although the underlying C function respects the +// timeout, it currently cannot be manually cancelled. That means manually +// cancelling the context will block until the C function call returns. +// +// Requires broker version >= 0.10.0. +func (a *AdminClient) ClusterID(ctx context.Context) (clusterID string, err error) { + responseChan := make(chan *C.char, 1) + + go func() { + responseChan <- C.rd_kafka_clusterid(a.handle.rk, cTimeoutFromContext(ctx)) + }() + + select { + case <-ctx.Done(): + if cClusterID := <-responseChan; cClusterID != nil { + C.rd_kafka_mem_free(a.handle.rk, unsafe.Pointer(cClusterID)) + } + return "", ctx.Err() + + case cClusterID := <-responseChan: + if cClusterID == nil { // C timeout + <-ctx.Done() + return "", ctx.Err() + } + defer C.rd_kafka_mem_free(a.handle.rk, unsafe.Pointer(cClusterID)) + return C.GoString(cClusterID), nil + } +} + +// ControllerID returns the broker ID of the current controller as reported in +// broker metadata. +// +// Note on cancellation: Although the underlying C function respects the +// timeout, it currently cannot be manually cancelled. That means manually +// cancelling the context will block until the C function call returns. +// +// Requires broker version >= 0.10.0. +func (a *AdminClient) ControllerID(ctx context.Context) (controllerID int32, err error) { + responseChan := make(chan int32, 1) + + go func() { + responseChan <- int32(C.rd_kafka_controllerid(a.handle.rk, cTimeoutFromContext(ctx))) + }() + + select { + case <-ctx.Done(): + <-responseChan + return 0, ctx.Err() + + case controllerID := <-responseChan: + if controllerID < 0 { // C timeout + <-ctx.Done() + return 0, ctx.Err() + } + return controllerID, nil + } +} + +// CreateTopics creates topics in cluster. +// +// The list of TopicSpecification objects define the per-topic partition count, replicas, etc. +// +// Topic creation is non-atomic and may succeed for some topics but fail for others, +// make sure to check the result for topic-specific errors. +// +// Note: TopicSpecification is analogous to NewTopic in the Java Topic Admin API. +func (a *AdminClient) CreateTopics(ctx context.Context, topics []TopicSpecification, options ...CreateTopicsAdminOption) (result []TopicResult, err error) { + cTopics := make([]*C.rd_kafka_NewTopic_t, len(topics)) + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + // Convert Go TopicSpecifications to C TopicSpecifications + for i, topic := range topics { + + var cReplicationFactor C.int + if topic.ReplicationFactor == 0 { + cReplicationFactor = -1 + } else { + cReplicationFactor = C.int(topic.ReplicationFactor) + } + if topic.ReplicaAssignment != nil { + if cReplicationFactor != -1 { + return nil, newErrorFromString(ErrInvalidArg, + "TopicSpecification.ReplicationFactor and TopicSpecification.ReplicaAssignment are mutually exclusive") + } + + if len(topic.ReplicaAssignment) != topic.NumPartitions { + return nil, newErrorFromString(ErrInvalidArg, + "TopicSpecification.ReplicaAssignment must contain exactly TopicSpecification.NumPartitions partitions") + } + } + + cTopics[i] = C.rd_kafka_NewTopic_new( + C.CString(topic.Topic), + C.int(topic.NumPartitions), + cReplicationFactor, + cErrstr, cErrstrSize) + if cTopics[i] == nil { + return nil, newErrorFromString(ErrInvalidArg, + fmt.Sprintf("Topic %s: %s", topic.Topic, C.GoString(cErrstr))) + } + + defer C.rd_kafka_NewTopic_destroy(cTopics[i]) + + for p, replicas := range topic.ReplicaAssignment { + cReplicas := make([]C.int32_t, len(replicas)) + for ri, replica := range replicas { + cReplicas[ri] = C.int32_t(replica) + } + cErr := C.rd_kafka_NewTopic_set_replica_assignment( + cTopics[i], C.int32_t(p), + (*C.int32_t)(&cReplicas[0]), C.size_t(len(cReplicas)), + cErrstr, cErrstrSize) + if cErr != 0 { + return nil, newCErrorFromString(cErr, + fmt.Sprintf("Failed to set replica assignment for topic %s partition %d: %s", topic.Topic, p, C.GoString(cErrstr))) + } + } + + for key, value := range topic.Config { + cErr := C.rd_kafka_NewTopic_set_config( + cTopics[i], + C.CString(key), C.CString(value)) + if cErr != 0 { + return nil, newCErrorFromString(cErr, + fmt.Sprintf("Failed to set config %s=%s for topic %s", key, value, topic.Topic)) + } + } + } + + // Convert Go AdminOptions (if any) to C AdminOptions + genericOptions := make([]AdminOption, len(options)) + for i := range options { + genericOptions[i] = options[i] + } + cOptions, err := adminOptionsSetup(a.handle, C.RD_KAFKA_ADMIN_OP_CREATETOPICS, genericOptions) + if err != nil { + return nil, err + } + defer C.rd_kafka_AdminOptions_destroy(cOptions) + + // Create temporary queue for async operation + cQueue := C.rd_kafka_queue_new(a.handle.rk) + defer C.rd_kafka_queue_destroy(cQueue) + + // Asynchronous call + C.rd_kafka_CreateTopics( + a.handle.rk, + (**C.rd_kafka_NewTopic_t)(&cTopics[0]), + C.size_t(len(cTopics)), + cOptions, + cQueue) + + // Wait for result, error or context timeout + rkev, err := a.waitResult(ctx, cQueue, C.RD_KAFKA_EVENT_CREATETOPICS_RESULT) + if err != nil { + return nil, err + } + defer C.rd_kafka_event_destroy(rkev) + + cRes := C.rd_kafka_event_CreateTopics_result(rkev) + + // Convert result from C to Go + var cCnt C.size_t + cTopicRes := C.rd_kafka_CreateTopics_result_topics(cRes, &cCnt) + + return a.cToTopicResults(cTopicRes, cCnt) +} + +// DeleteTopics deletes a batch of topics. +// +// This operation is not transactional and may succeed for a subset of topics while +// failing others. +// It may take several seconds after the DeleteTopics result returns success for +// all the brokers to become aware that the topics are gone. During this time, +// topic metadata and configuration may continue to return information about deleted topics. +// +// Requires broker version >= 0.10.1.0 +func (a *AdminClient) DeleteTopics(ctx context.Context, topics []string, options ...DeleteTopicsAdminOption) (result []TopicResult, err error) { + cTopics := make([]*C.rd_kafka_DeleteTopic_t, len(topics)) + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + // Convert Go DeleteTopics to C DeleteTopics + for i, topic := range topics { + cTopics[i] = C.rd_kafka_DeleteTopic_new(C.CString(topic)) + if cTopics[i] == nil { + return nil, newErrorFromString(ErrInvalidArg, + fmt.Sprintf("Invalid arguments for topic %s", topic)) + } + + defer C.rd_kafka_DeleteTopic_destroy(cTopics[i]) + } + + // Convert Go AdminOptions (if any) to C AdminOptions + genericOptions := make([]AdminOption, len(options)) + for i := range options { + genericOptions[i] = options[i] + } + cOptions, err := adminOptionsSetup(a.handle, C.RD_KAFKA_ADMIN_OP_DELETETOPICS, genericOptions) + if err != nil { + return nil, err + } + defer C.rd_kafka_AdminOptions_destroy(cOptions) + + // Create temporary queue for async operation + cQueue := C.rd_kafka_queue_new(a.handle.rk) + defer C.rd_kafka_queue_destroy(cQueue) + + // Asynchronous call + C.rd_kafka_DeleteTopics( + a.handle.rk, + (**C.rd_kafka_DeleteTopic_t)(&cTopics[0]), + C.size_t(len(cTopics)), + cOptions, + cQueue) + + // Wait for result, error or context timeout + rkev, err := a.waitResult(ctx, cQueue, C.RD_KAFKA_EVENT_DELETETOPICS_RESULT) + if err != nil { + return nil, err + } + defer C.rd_kafka_event_destroy(rkev) + + cRes := C.rd_kafka_event_DeleteTopics_result(rkev) + + // Convert result from C to Go + var cCnt C.size_t + cTopicRes := C.rd_kafka_DeleteTopics_result_topics(cRes, &cCnt) + + return a.cToTopicResults(cTopicRes, cCnt) +} + +// CreatePartitions creates additional partitions for topics. +func (a *AdminClient) CreatePartitions(ctx context.Context, partitions []PartitionsSpecification, options ...CreatePartitionsAdminOption) (result []TopicResult, err error) { + cParts := make([]*C.rd_kafka_NewPartitions_t, len(partitions)) + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + // Convert Go PartitionsSpecification to C NewPartitions + for i, part := range partitions { + cParts[i] = C.rd_kafka_NewPartitions_new(C.CString(part.Topic), C.size_t(part.IncreaseTo), cErrstr, cErrstrSize) + if cParts[i] == nil { + return nil, newErrorFromString(ErrInvalidArg, + fmt.Sprintf("Topic %s: %s", part.Topic, C.GoString(cErrstr))) + } + + defer C.rd_kafka_NewPartitions_destroy(cParts[i]) + + for pidx, replicas := range part.ReplicaAssignment { + cReplicas := make([]C.int32_t, len(replicas)) + for ri, replica := range replicas { + cReplicas[ri] = C.int32_t(replica) + } + cErr := C.rd_kafka_NewPartitions_set_replica_assignment( + cParts[i], C.int32_t(pidx), + (*C.int32_t)(&cReplicas[0]), C.size_t(len(cReplicas)), + cErrstr, cErrstrSize) + if cErr != 0 { + return nil, newCErrorFromString(cErr, + fmt.Sprintf("Failed to set replica assignment for topic %s new partition index %d: %s", part.Topic, pidx, C.GoString(cErrstr))) + } + } + + } + + // Convert Go AdminOptions (if any) to C AdminOptions + genericOptions := make([]AdminOption, len(options)) + for i := range options { + genericOptions[i] = options[i] + } + cOptions, err := adminOptionsSetup(a.handle, C.RD_KAFKA_ADMIN_OP_CREATEPARTITIONS, genericOptions) + if err != nil { + return nil, err + } + defer C.rd_kafka_AdminOptions_destroy(cOptions) + + // Create temporary queue for async operation + cQueue := C.rd_kafka_queue_new(a.handle.rk) + defer C.rd_kafka_queue_destroy(cQueue) + + // Asynchronous call + C.rd_kafka_CreatePartitions( + a.handle.rk, + (**C.rd_kafka_NewPartitions_t)(&cParts[0]), + C.size_t(len(cParts)), + cOptions, + cQueue) + + // Wait for result, error or context timeout + rkev, err := a.waitResult(ctx, cQueue, C.RD_KAFKA_EVENT_CREATEPARTITIONS_RESULT) + if err != nil { + return nil, err + } + defer C.rd_kafka_event_destroy(rkev) + + cRes := C.rd_kafka_event_CreatePartitions_result(rkev) + + // Convert result from C to Go + var cCnt C.size_t + cTopicRes := C.rd_kafka_CreatePartitions_result_topics(cRes, &cCnt) + + return a.cToTopicResults(cTopicRes, cCnt) +} + +// AlterConfigs alters/updates cluster resource configuration. +// +// Updates are not transactional so they may succeed for a subset +// of the provided resources while others fail. +// The configuration for a particular resource is updated atomically, +// replacing values using the provided ConfigEntrys and reverting +// unspecified ConfigEntrys to their default values. +// +// Requires broker version >=0.11.0.0 +// +// AlterConfigs will replace all existing configuration for +// the provided resources with the new configuration given, +// reverting all other configuration to their default values. +// +// Multiple resources and resource types may be set, but at most one +// resource of type ResourceBroker is allowed per call since these +// resource requests must be sent to the broker specified in the resource. +func (a *AdminClient) AlterConfigs(ctx context.Context, resources []ConfigResource, options ...AlterConfigsAdminOption) (result []ConfigResourceResult, err error) { + cRes := make([]*C.rd_kafka_ConfigResource_t, len(resources)) + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + // Convert Go ConfigResources to C ConfigResources + for i, res := range resources { + cRes[i] = C.rd_kafka_ConfigResource_new( + C.rd_kafka_ResourceType_t(res.Type), C.CString(res.Name)) + if cRes[i] == nil { + return nil, newErrorFromString(ErrInvalidArg, + fmt.Sprintf("Invalid arguments for resource %v", res)) + } + + defer C.rd_kafka_ConfigResource_destroy(cRes[i]) + + for _, entry := range res.Config { + var cErr C.rd_kafka_resp_err_t + switch entry.Operation { + case AlterOperationSet: + cErr = C.rd_kafka_ConfigResource_set_config( + cRes[i], C.CString(entry.Name), C.CString(entry.Value)) + default: + panic(fmt.Sprintf("Invalid ConfigEntry.Operation: %v", entry.Operation)) + } + + if cErr != 0 { + return nil, + newCErrorFromString(cErr, + fmt.Sprintf("Failed to add configuration %s: %s", + entry, C.GoString(C.rd_kafka_err2str(cErr)))) + } + } + } + + // Convert Go AdminOptions (if any) to C AdminOptions + genericOptions := make([]AdminOption, len(options)) + for i := range options { + genericOptions[i] = options[i] + } + cOptions, err := adminOptionsSetup(a.handle, C.RD_KAFKA_ADMIN_OP_ALTERCONFIGS, genericOptions) + if err != nil { + return nil, err + } + defer C.rd_kafka_AdminOptions_destroy(cOptions) + + // Create temporary queue for async operation + cQueue := C.rd_kafka_queue_new(a.handle.rk) + defer C.rd_kafka_queue_destroy(cQueue) + + // Asynchronous call + C.rd_kafka_AlterConfigs( + a.handle.rk, + (**C.rd_kafka_ConfigResource_t)(&cRes[0]), + C.size_t(len(cRes)), + cOptions, + cQueue) + + // Wait for result, error or context timeout + rkev, err := a.waitResult(ctx, cQueue, C.RD_KAFKA_EVENT_ALTERCONFIGS_RESULT) + if err != nil { + return nil, err + } + defer C.rd_kafka_event_destroy(rkev) + + cResult := C.rd_kafka_event_AlterConfigs_result(rkev) + + // Convert results from C to Go + var cCnt C.size_t + cResults := C.rd_kafka_AlterConfigs_result_resources(cResult, &cCnt) + + return a.cConfigResourceToResult(cResults, cCnt) +} + +// DescribeConfigs retrieves configuration for cluster resources. +// +// The returned configuration includes default values, use +// ConfigEntryResult.IsDefault or ConfigEntryResult.Source to distinguish +// default values from manually configured settings. +// +// The value of config entries where .IsSensitive is true +// will always be nil to avoid disclosing sensitive +// information, such as security settings. +// +// Configuration entries where .IsReadOnly is true can't be modified +// (with AlterConfigs). +// +// Synonym configuration entries are returned if the broker supports +// it (broker version >= 1.1.0). See .Synonyms. +// +// Requires broker version >=0.11.0.0 +// +// Multiple resources and resource types may be requested, but at most +// one resource of type ResourceBroker is allowed per call +// since these resource requests must be sent to the broker specified +// in the resource. +func (a *AdminClient) DescribeConfigs(ctx context.Context, resources []ConfigResource, options ...DescribeConfigsAdminOption) (result []ConfigResourceResult, err error) { + cRes := make([]*C.rd_kafka_ConfigResource_t, len(resources)) + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + // Convert Go ConfigResources to C ConfigResources + for i, res := range resources { + cRes[i] = C.rd_kafka_ConfigResource_new( + C.rd_kafka_ResourceType_t(res.Type), C.CString(res.Name)) + if cRes[i] == nil { + return nil, newErrorFromString(ErrInvalidArg, + fmt.Sprintf("Invalid arguments for resource %v", res)) + } + + defer C.rd_kafka_ConfigResource_destroy(cRes[i]) + } + + // Convert Go AdminOptions (if any) to C AdminOptions + genericOptions := make([]AdminOption, len(options)) + for i := range options { + genericOptions[i] = options[i] + } + cOptions, err := adminOptionsSetup(a.handle, C.RD_KAFKA_ADMIN_OP_DESCRIBECONFIGS, genericOptions) + if err != nil { + return nil, err + } + defer C.rd_kafka_AdminOptions_destroy(cOptions) + + // Create temporary queue for async operation + cQueue := C.rd_kafka_queue_new(a.handle.rk) + defer C.rd_kafka_queue_destroy(cQueue) + + // Asynchronous call + C.rd_kafka_DescribeConfigs( + a.handle.rk, + (**C.rd_kafka_ConfigResource_t)(&cRes[0]), + C.size_t(len(cRes)), + cOptions, + cQueue) + + // Wait for result, error or context timeout + rkev, err := a.waitResult(ctx, cQueue, C.RD_KAFKA_EVENT_DESCRIBECONFIGS_RESULT) + if err != nil { + return nil, err + } + defer C.rd_kafka_event_destroy(rkev) + + cResult := C.rd_kafka_event_DescribeConfigs_result(rkev) + + // Convert results from C to Go + var cCnt C.size_t + cResults := C.rd_kafka_DescribeConfigs_result_resources(cResult, &cCnt) + + return a.cConfigResourceToResult(cResults, cCnt) +} + +// GetMetadata queries broker for cluster and topic metadata. +// If topic is non-nil only information about that topic is returned, else if +// allTopics is false only information about locally used topics is returned, +// else information about all topics is returned. +// GetMetadata is equivalent to listTopics, describeTopics and describeCluster in the Java API. +func (a *AdminClient) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error) { + return getMetadata(a, topic, allTopics, timeoutMs) +} + +// String returns a human readable name for an AdminClient instance +func (a *AdminClient) String() string { + return fmt.Sprintf("admin-%s", a.handle.String()) +} + +// get_handle implements the Handle interface +func (a *AdminClient) gethandle() *handle { + return a.handle +} + +// SetOAuthBearerToken sets the the data to be transmitted +// to a broker during SASL/OAUTHBEARER authentication. It will return nil +// on success, otherwise an error if: +// 1) the token data is invalid (meaning an expiration time in the past +// or either a token value or an extension key or value that does not meet +// the regular expression requirements as per +// https://tools.ietf.org/html/rfc7628#section-3.1); +// 2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +// 3) SASL/OAUTHBEARER is supported but is not configured as the client's +// authentication mechanism. +func (a *AdminClient) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error { + return a.handle.setOAuthBearerToken(oauthBearerToken) +} + +// SetOAuthBearerTokenFailure sets the error message describing why token +// retrieval/setting failed; it also schedules a new token refresh event for 10 +// seconds later so the attempt may be retried. It will return nil on +// success, otherwise an error if: +// 1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +// 2) SASL/OAUTHBEARER is supported but is not configured as the client's +// authentication mechanism. +func (a *AdminClient) SetOAuthBearerTokenFailure(errstr string) error { + return a.handle.setOAuthBearerTokenFailure(errstr) +} + +// Close an AdminClient instance. +func (a *AdminClient) Close() { + if a.isDerived { + // Derived AdminClient needs no cleanup. + a.handle = &handle{} + return + } + + a.handle.cleanup() + + C.rd_kafka_destroy(a.handle.rk) +} + +// NewAdminClient creats a new AdminClient instance with a new underlying client instance +func NewAdminClient(conf *ConfigMap) (*AdminClient, error) { + + err := versionCheck() + if err != nil { + return nil, err + } + + a := &AdminClient{} + a.handle = &handle{} + + // Convert ConfigMap to librdkafka conf_t + cConf, err := conf.convert() + if err != nil { + return nil, err + } + + cErrstr := (*C.char)(C.malloc(C.size_t(256))) + defer C.free(unsafe.Pointer(cErrstr)) + + C.rd_kafka_conf_set_events(cConf, C.RD_KAFKA_EVENT_STATS|C.RD_KAFKA_EVENT_ERROR|C.RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH) + + // Create librdkafka producer instance. The Producer is somewhat cheaper than + // the consumer, but any instance type can be used for Admin APIs. + a.handle.rk = C.rd_kafka_new(C.RD_KAFKA_PRODUCER, cConf, cErrstr, 256) + if a.handle.rk == nil { + return nil, newErrorFromCString(C.RD_KAFKA_RESP_ERR__INVALID_ARG, cErrstr) + } + + a.isDerived = false + a.handle.setup() + + return a, nil +} + +// NewAdminClientFromProducer derives a new AdminClient from an existing Producer instance. +// The AdminClient will use the same configuration and connections as the parent instance. +func NewAdminClientFromProducer(p *Producer) (a *AdminClient, err error) { + if p.handle.rk == nil { + return nil, newErrorFromString(ErrInvalidArg, "Can't derive AdminClient from closed producer") + } + + a = &AdminClient{} + a.handle = &p.handle + a.isDerived = true + return a, nil +} + +// NewAdminClientFromConsumer derives a new AdminClient from an existing Consumer instance. +// The AdminClient will use the same configuration and connections as the parent instance. +func NewAdminClientFromConsumer(c *Consumer) (a *AdminClient, err error) { + if c.handle.rk == nil { + return nil, newErrorFromString(ErrInvalidArg, "Can't derive AdminClient from closed consumer") + } + + a = &AdminClient{} + a.handle = &c.handle + a.isDerived = true + return a, nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/adminoptions.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/adminoptions.go new file mode 100644 index 0000000..842631b --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/adminoptions.go @@ -0,0 +1,264 @@ +/** + * Copyright 2018 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import ( + "fmt" + "time" + "unsafe" +) + +/* +#include "select_rdkafka.h" +#include +*/ +import "C" + +// AdminOptionOperationTimeout sets the broker's operation timeout, such as the +// timeout for CreateTopics to complete the creation of topics on the controller +// before returning a result to the application. +// +// CreateTopics, DeleteTopics, CreatePartitions: +// a value 0 will return immediately after triggering topic +// creation, while > 0 will wait this long for topic creation to propagate +// in cluster. +// +// Default: 0 (return immediately). +// +// Valid for CreateTopics, DeleteTopics, CreatePartitions. +type AdminOptionOperationTimeout struct { + isSet bool + val time.Duration +} + +func (ao AdminOptionOperationTimeout) supportsCreateTopics() { +} +func (ao AdminOptionOperationTimeout) supportsDeleteTopics() { +} +func (ao AdminOptionOperationTimeout) supportsCreatePartitions() { +} + +func (ao AdminOptionOperationTimeout) apply(cOptions *C.rd_kafka_AdminOptions_t) error { + if !ao.isSet { + return nil + } + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + cErr := C.rd_kafka_AdminOptions_set_operation_timeout( + cOptions, C.int(durationToMilliseconds(ao.val)), + cErrstr, cErrstrSize) + if cErr != 0 { + C.rd_kafka_AdminOptions_destroy(cOptions) + return newCErrorFromString(cErr, + fmt.Sprintf("Failed to set operation timeout: %s", C.GoString(cErrstr))) + + } + + return nil +} + +// SetAdminOperationTimeout sets the broker's operation timeout, such as the +// timeout for CreateTopics to complete the creation of topics on the controller +// before returning a result to the application. +// +// CreateTopics, DeleteTopics, CreatePartitions: +// a value 0 will return immediately after triggering topic +// creation, while > 0 will wait this long for topic creation to propagate +// in cluster. +// +// Default: 0 (return immediately). +// +// Valid for CreateTopics, DeleteTopics, CreatePartitions. +func SetAdminOperationTimeout(t time.Duration) (ao AdminOptionOperationTimeout) { + ao.isSet = true + ao.val = t + return ao +} + +// AdminOptionRequestTimeout sets the overall request timeout, including broker +// lookup, request transmission, operation time on broker, and response. +// +// Default: `socket.timeout.ms`. +// +// Valid for all Admin API methods. +type AdminOptionRequestTimeout struct { + isSet bool + val time.Duration +} + +func (ao AdminOptionRequestTimeout) supportsCreateTopics() { +} +func (ao AdminOptionRequestTimeout) supportsDeleteTopics() { +} +func (ao AdminOptionRequestTimeout) supportsCreatePartitions() { +} +func (ao AdminOptionRequestTimeout) supportsAlterConfigs() { +} +func (ao AdminOptionRequestTimeout) supportsDescribeConfigs() { +} + +func (ao AdminOptionRequestTimeout) apply(cOptions *C.rd_kafka_AdminOptions_t) error { + if !ao.isSet { + return nil + } + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + cErr := C.rd_kafka_AdminOptions_set_request_timeout( + cOptions, C.int(durationToMilliseconds(ao.val)), + cErrstr, cErrstrSize) + if cErr != 0 { + C.rd_kafka_AdminOptions_destroy(cOptions) + return newCErrorFromString(cErr, + fmt.Sprintf("%s", C.GoString(cErrstr))) + + } + + return nil +} + +// SetAdminRequestTimeout sets the overall request timeout, including broker +// lookup, request transmission, operation time on broker, and response. +// +// Default: `socket.timeout.ms`. +// +// Valid for all Admin API methods. +func SetAdminRequestTimeout(t time.Duration) (ao AdminOptionRequestTimeout) { + ao.isSet = true + ao.val = t + return ao +} + +// AdminOptionValidateOnly tells the broker to only validate the request, +// without performing the requested operation (create topics, etc). +// +// Default: false. +// +// Valid for CreateTopics, CreatePartitions, AlterConfigs +type AdminOptionValidateOnly struct { + isSet bool + val bool +} + +func (ao AdminOptionValidateOnly) supportsCreateTopics() { +} +func (ao AdminOptionValidateOnly) supportsCreatePartitions() { +} +func (ao AdminOptionValidateOnly) supportsAlterConfigs() { +} + +func (ao AdminOptionValidateOnly) apply(cOptions *C.rd_kafka_AdminOptions_t) error { + if !ao.isSet { + return nil + } + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + cErr := C.rd_kafka_AdminOptions_set_validate_only( + cOptions, bool2cint(ao.val), + cErrstr, cErrstrSize) + if cErr != 0 { + C.rd_kafka_AdminOptions_destroy(cOptions) + return newCErrorFromString(cErr, + fmt.Sprintf("%s", C.GoString(cErrstr))) + + } + + return nil +} + +// SetAdminValidateOnly tells the broker to only validate the request, +// without performing the requested operation (create topics, etc). +// +// Default: false. +// +// Valid for CreateTopics, DeleteTopics, CreatePartitions, AlterConfigs +func SetAdminValidateOnly(validateOnly bool) (ao AdminOptionValidateOnly) { + ao.isSet = true + ao.val = validateOnly + return ao +} + +// CreateTopicsAdminOption - see setters. +// +// See SetAdminRequestTimeout, SetAdminOperationTimeout, SetAdminValidateOnly. +type CreateTopicsAdminOption interface { + supportsCreateTopics() + apply(cOptions *C.rd_kafka_AdminOptions_t) error +} + +// DeleteTopicsAdminOption - see setters. +// +// See SetAdminRequestTimeout, SetAdminOperationTimeout. +type DeleteTopicsAdminOption interface { + supportsDeleteTopics() + apply(cOptions *C.rd_kafka_AdminOptions_t) error +} + +// CreatePartitionsAdminOption - see setters. +// +// See SetAdminRequestTimeout, SetAdminOperationTimeout, SetAdminValidateOnly. +type CreatePartitionsAdminOption interface { + supportsCreatePartitions() + apply(cOptions *C.rd_kafka_AdminOptions_t) error +} + +// AlterConfigsAdminOption - see setters. +// +// See SetAdminRequestTimeout, SetAdminValidateOnly, SetAdminIncremental. +type AlterConfigsAdminOption interface { + supportsAlterConfigs() + apply(cOptions *C.rd_kafka_AdminOptions_t) error +} + +// DescribeConfigsAdminOption - see setters. +// +// See SetAdminRequestTimeout. +type DescribeConfigsAdminOption interface { + supportsDescribeConfigs() + apply(cOptions *C.rd_kafka_AdminOptions_t) error +} + +// AdminOption is a generic type not to be used directly. +// +// See CreateTopicsAdminOption et.al. +type AdminOption interface { + apply(cOptions *C.rd_kafka_AdminOptions_t) error +} + +func adminOptionsSetup(h *handle, opType C.rd_kafka_admin_op_t, options []AdminOption) (*C.rd_kafka_AdminOptions_t, error) { + + cOptions := C.rd_kafka_AdminOptions_new(h.rk, opType) + for _, opt := range options { + if opt == nil { + continue + } + err := opt.apply(cOptions) + if err != nil { + return nil, err + } + } + + return cOptions, nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/api.html b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/api.html new file mode 100644 index 0000000..8c495b3 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/api.html @@ -0,0 +1,4737 @@ + + + + + + + + kafka - Go Documentation Server + + + + + + + + +
+ ... +
+ +
+
+

+ Package kafka + + +

+ + + + +
+
+
+ + import "github.com/confluentinc/confluent-kafka-go/kafka" + +
+
+
+
+ + Overview + +
+
+ + Index + +
+
+
+
+
+ +
+ +
+

+ Overview ▾ +

+

+ Package kafka provides high-level Apache Kafka producer and consumers +using bindings on-top of the librdkafka C library. +

+

+ High-level Consumer +

+

+ * Decide if you want to read messages and events by calling `.Poll()` or +the deprecated option of using the `.Events()` channel. (If you want to use +`.Events()` channel then set `"go.events.channel.enable": true`). +

+

+ * Create a Consumer with `kafka.NewConsumer()` providing at +least the `bootstrap.servers` and `group.id` configuration properties. +

+

+ * Call `.Subscribe()` or (`.SubscribeTopics()` to subscribe to multiple topics) +to join the group with the specified subscription set. +Subscriptions are atomic, calling `.Subscribe*()` again will leave +the group and rejoin with the new set of topics. +

+

+ * Start reading events and messages from either the `.Events` channel +or by calling `.Poll()`. +

+

+ * When the group has rebalanced each client member is assigned a +(sub-)set of topic+partitions. +By default the consumer will start fetching messages for its assigned +partitions at this point, but your application may enable rebalance +events to get an insight into what the assigned partitions where +as well as set the initial offsets. To do this you need to pass +`"go.application.rebalance.enable": true` to the `NewConsumer()` call +mentioned above. You will (eventually) see a `kafka.AssignedPartitions` event +with the assigned partition set. You can optionally modify the initial +offsets (they'll default to stored offsets and if there are no previously stored +offsets it will fall back to `"auto.offset.reset"` +which defaults to the `latest` message) and then call `.Assign(partitions)` +to start consuming. If you don't need to modify the initial offsets you will +not need to call `.Assign()`, the client will do so automatically for you if +you dont, unless you are using the channel-based consumer in which case +you MUST call `.Assign()` when receiving the `AssignedPartitions` and +`RevokedPartitions` events. +

+

+ * As messages are fetched they will be made available on either the +`.Events` channel or by calling `.Poll()`, look for event type `*kafka.Message`. +

+

+ * Handle messages, events and errors to your liking. +

+

+ * When you are done consuming call `.Close()` to commit final offsets +and leave the consumer group. +

+

+ Producer +

+

+ * Create a Producer with `kafka.NewProducer()` providing at least +the `bootstrap.servers` configuration properties. +

+

+ * Messages may now be produced either by sending a `*kafka.Message` +on the `.ProduceChannel` or by calling `.Produce()`. +

+

+ * Producing is an asynchronous operation so the client notifies the application +of per-message produce success or failure through something called delivery reports. +Delivery reports are by default emitted on the `.Events()` channel as `*kafka.Message` +and you should check `msg.TopicPartition.Error` for `nil` to find out if the message +was succesfully delivered or not. +It is also possible to direct delivery reports to alternate channels +by providing a non-nil `chan Event` channel to `.Produce()`. +If no delivery reports are wanted they can be completely disabled by +setting configuration property `"go.delivery.reports": false`. +

+

+ * When you are done producing messages you will need to make sure all messages +are indeed delivered to the broker (or failed), remember that this is +an asynchronous client so some of your messages may be lingering in internal +channels or tranmission queues. +To do this you can either keep track of the messages you've produced +and wait for their corresponding delivery reports, or call the convenience +function `.Flush()` that will block until all message deliveries are done +or the provided timeout elapses. +

+

+ * Finally call `.Close()` to decommission the producer. +

+

+ Transactional producer API +

+

+ The transactional producer operates on top of the idempotent producer, +and provides full exactly-once semantics (EOS) for Apache Kafka when used +with the transaction aware consumer (`isolation.level=read_committed`). +

+

+ A producer instance is configured for transactions by setting the +`transactional.id` to an identifier unique for the application. This +id will be used to fence stale transactions from previous instances of +the application, typically following an outage or crash. +

+

+ After creating the transactional producer instance using `NewProducer()` +the transactional state must be initialized by calling +`InitTransactions()`. This is a blocking call that will +acquire a runtime producer id from the transaction coordinator broker +as well as abort any stale transactions and fence any still running producer +instances with the same `transactional.id`. +

+

+ Once transactions are initialized the application may begin a new +transaction by calling `BeginTransaction()`. +A producer instance may only have one single on-going transaction. +

+

+ Any messages produced after the transaction has been started will +belong to the ongoing transaction and will be committed or aborted +atomically. +It is not permitted to produce messages outside a transaction +boundary, e.g., before `BeginTransaction()` or after `CommitTransaction()`, +`AbortTransaction()` or if the current transaction has failed. +

+

+ If consumed messages are used as input to the transaction, the consumer +instance must be configured with `enable.auto.commit` set to `false`. +To commit the consumed offsets along with the transaction pass the +list of consumed partitions and the last offset processed + 1 to +`SendOffsetsToTransaction()` prior to committing the transaction. +This allows an aborted transaction to be restarted using the previously +committed offsets. +

+

+ To commit the produced messages, and any consumed offsets, to the +current transaction, call `CommitTransaction()`. +This call will block until the transaction has been fully committed or +failed (typically due to fencing by a newer producer instance). +

+

+ Alternatively, if processing fails, or an abortable transaction error is +raised, the transaction needs to be aborted by calling +`AbortTransaction()` which marks any produced messages and +offset commits as aborted. +

+

+ After the current transaction has been committed or aborted a new +transaction may be started by calling `BeginTransaction()` again. +

+

+ Retriable errors: +Some error cases allow the attempted operation to be retried, this is +indicated by the error object having the retriable flag set which can +be detected by calling `err.(kafka.Error).IsRetriable()`. +When this flag is set the application may retry the operation immediately +or preferably after a shorter grace period (to avoid busy-looping). +Retriable errors include timeouts, broker transport failures, etc. +

+

+ Abortable errors: +An ongoing transaction may fail permanently due to various errors, +such as transaction coordinator becoming unavailable, write failures to the +Apache Kafka log, under-replicated partitions, etc. +At this point the producer application must abort the current transaction +using `AbortTransaction()` and optionally start a new transaction +by calling `BeginTransaction()`. +Whether an error is abortable or not is detected by calling +`err.(kafka.Error).TxnRequiresAbort()` on the returned error object. +

+

+ Fatal errors: +While the underlying idempotent producer will typically only raise +fatal errors for unrecoverable cluster errors where the idempotency +guarantees can't be maintained, most of these are treated as abortable by +the transactional producer since transactions may be aborted and retried +in their entirety; +The transactional producer on the other hand introduces a set of additional +fatal errors which the application needs to handle by shutting down the +producer and terminate. There is no way for a producer instance to recover +from fatal errors. +Whether an error is fatal or not is detected by calling +`err.(kafka.Error).IsFatal()` on the returned error object or by checking +the global `GetFatalError()`. +

+

+ Handling of other errors: +For errors that have neither retriable, abortable or the fatal flag set +it is not always obvious how to handle them. While some of these errors +may be indicative of bugs in the application code, such as when +an invalid parameter is passed to a method, other errors might originate +from the broker and be passed thru as-is to the application. +The general recommendation is to treat these errors, that have +neither the retriable or abortable flags set, as fatal. +

+

+ Error handling example: +

+ 
retry:
+
+   err := producer.CommitTransaction(...)
+   if err == nil {
+       return nil
+   } else if err.(kafka.Error).TxnRequiresAbort() {
+       do_abort_transaction_and_reset_inputs()
+   } else if err.(kafka.Error).IsRetriable() {
+       goto retry
+   } else { // treat all other errors as fatal errors
+       panic(err)
+   }
+
+

+ Events +

+

+ Apart from emitting messages and delivery reports the client also communicates +with the application through a number of different event types. +An application may choose to handle or ignore these events. +

+

+ Consumer events +

+

+ * `*kafka.Message` - a fetched message. +

+

+ * `AssignedPartitions` - The assigned partition set for this client following a rebalance. +Requires `go.application.rebalance.enable` +

+

+ * `RevokedPartitions` - The counter part to `AssignedPartitions` following a rebalance. +`AssignedPartitions` and `RevokedPartitions` are symmetrical. +Requires `go.application.rebalance.enable` +

+

+ * `PartitionEOF` - Consumer has reached the end of a partition. +NOTE: The consumer will keep trying to fetch new messages for the partition. +

+

+ * `OffsetsCommitted` - Offset commit results (when `enable.auto.commit` is enabled). +

+

+ Producer events +

+

+ * `*kafka.Message` - delivery report for produced message. +Check `.TopicPartition.Error` for delivery result. +

+

+ Generic events for both Consumer and Producer +

+

+ * `KafkaError` - client (error codes are prefixed with _) or broker error. +These errors are normally just informational since the +client will try its best to automatically recover (eventually). +

+

+ * `OAuthBearerTokenRefresh` - retrieval of a new SASL/OAUTHBEARER token is required. +This event only occurs with sasl.mechanism=OAUTHBEARER. +Be sure to invoke SetOAuthBearerToken() on the Producer/Consumer/AdminClient +instance when a successful token retrieval is completed, otherwise be sure to +invoke SetOAuthBearerTokenFailure() to indicate that retrieval failed (or +if setting the token failed, which could happen if an extension doesn't meet +the required regular expression); invoking SetOAuthBearerTokenFailure() will +schedule a new event for 10 seconds later so another retrieval can be attempted. +

+

+ Hint: If your application registers a signal notification +(signal.Notify) makes sure the signals channel is buffered to avoid +possible complications with blocking Poll() calls. +

+

+ Note: The Confluent Kafka Go client is safe for concurrent use. +

+
+
+
+ +
+

+ Index ▾ +

+ +
+
+
+ + Constants + +
+
+ + func LibraryVersion() (int, string) + +
+
+ + func WriteErrorCodes(f *os.File) + +
+
+ + type AdminClient + +
+
+ + func NewAdminClient(conf *ConfigMap) (*AdminClient, error) + +
+
+ + func NewAdminClientFromConsumer(c *Consumer) (a *AdminClient, err error) + +
+
+ + func NewAdminClientFromProducer(p *Producer) (a *AdminClient, err error) + +
+
+ + func (a *AdminClient) AlterConfigs(ctx context.Context, resources []ConfigResource, options ...AlterConfigsAdminOption) (result []ConfigResourceResult, err error) + +
+
+ + func (a *AdminClient) Close() + +
+
+ + func (a *AdminClient) ClusterID(ctx context.Context) (clusterID string, err error) + +
+
+ + func (a *AdminClient) ControllerID(ctx context.Context) (controllerID int32, err error) + +
+
+ + func (a *AdminClient) CreatePartitions(ctx context.Context, partitions []PartitionsSpecification, options ...CreatePartitionsAdminOption) (result []TopicResult, err error) + +
+
+ + func (a *AdminClient) CreateTopics(ctx context.Context, topics []TopicSpecification, options ...CreateTopicsAdminOption) (result []TopicResult, err error) + +
+
+ + func (a *AdminClient) DeleteTopics(ctx context.Context, topics []string, options ...DeleteTopicsAdminOption) (result []TopicResult, err error) + +
+
+ + func (a *AdminClient) DescribeConfigs(ctx context.Context, resources []ConfigResource, options ...DescribeConfigsAdminOption) (result []ConfigResourceResult, err error) + +
+
+ + func (a *AdminClient) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error) + +
+
+ + func (a *AdminClient) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error + +
+
+ + func (a *AdminClient) SetOAuthBearerTokenFailure(errstr string) error + +
+
+ + func (a *AdminClient) String() string + +
+
+ + type AdminOption + +
+
+ + type AdminOptionOperationTimeout + +
+
+ + func SetAdminOperationTimeout(t time.Duration) (ao AdminOptionOperationTimeout) + +
+
+ + type AdminOptionRequestTimeout + +
+
+ + func SetAdminRequestTimeout(t time.Duration) (ao AdminOptionRequestTimeout) + +
+
+ + type AdminOptionValidateOnly + +
+
+ + func SetAdminValidateOnly(validateOnly bool) (ao AdminOptionValidateOnly) + +
+
+ + type AlterConfigsAdminOption + +
+
+ + type AlterOperation + +
+
+ + func (o AlterOperation) String() string + +
+
+ + type AssignedPartitions + +
+
+ + func (e AssignedPartitions) String() string + +
+
+ + type BrokerMetadata + +
+
+ + type ConfigEntry + +
+
+ + func StringMapToConfigEntries(stringMap map[string]string, operation AlterOperation) []ConfigEntry + +
+
+ + func (c ConfigEntry) String() string + +
+
+ + type ConfigEntryResult + +
+
+ + func (c ConfigEntryResult) String() string + +
+
+ + type ConfigMap + +
+
+ + func (m ConfigMap) Get(key string, defval ConfigValue) (ConfigValue, error) + +
+
+ + func (m ConfigMap) Set(kv string) error + +
+
+ + func (m ConfigMap) SetKey(key string, value ConfigValue) error + +
+
+ + type ConfigResource + +
+
+ + func (c ConfigResource) String() string + +
+
+ + type ConfigResourceResult + +
+
+ + func (c ConfigResourceResult) String() string + +
+
+ + type ConfigSource + +
+
+ + func (t ConfigSource) String() string + +
+
+ + type ConfigValue + +
+
+ + type Consumer + +
+
+ + func NewConsumer(conf *ConfigMap) (*Consumer, error) + +
+
+ + func (c *Consumer) Assign(partitions []TopicPartition) (err error) + +
+
+ + func (c *Consumer) Assignment() (partitions []TopicPartition, err error) + +
+
+ + func (c *Consumer) AssignmentLost() bool + +
+
+ + func (c *Consumer) Close() (err error) + +
+
+ + func (c *Consumer) Commit() ([]TopicPartition, error) + +
+
+ + func (c *Consumer) CommitMessage(m *Message) ([]TopicPartition, error) + +
+
+ + func (c *Consumer) CommitOffsets(offsets []TopicPartition) ([]TopicPartition, error) + +
+
+ + func (c *Consumer) Committed(partitions []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) + +
+
+ + func (c *Consumer) Events() chan Event + +
+
+ + func (c *Consumer) GetConsumerGroupMetadata() (*ConsumerGroupMetadata, error) + +
+
+ + func (c *Consumer) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error) + +
+
+ + func (c *Consumer) GetRebalanceProtocol() string + +
+
+ + func (c *Consumer) GetWatermarkOffsets(topic string, partition int32) (low, high int64, err error) + +
+
+ + func (c *Consumer) IncrementalAssign(partitions []TopicPartition) (err error) + +
+
+ + func (c *Consumer) IncrementalUnassign(partitions []TopicPartition) (err error) + +
+
+ + func (c *Consumer) Logs() chan LogEvent + +
+
+ + func (c *Consumer) OffsetsForTimes(times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) + +
+
+ + func (c *Consumer) Pause(partitions []TopicPartition) (err error) + +
+
+ + func (c *Consumer) Poll(timeoutMs int) (event Event) + +
+
+ + func (c *Consumer) Position(partitions []TopicPartition) (offsets []TopicPartition, err error) + +
+
+ + func (c *Consumer) QueryWatermarkOffsets(topic string, partition int32, timeoutMs int) (low, high int64, err error) + +
+
+ + func (c *Consumer) ReadMessage(timeout time.Duration) (*Message, error) + +
+
+ + func (c *Consumer) Resume(partitions []TopicPartition) (err error) + +
+
+ + func (c *Consumer) Seek(partition TopicPartition, timeoutMs int) error + +
+
+ + func (c *Consumer) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error + +
+
+ + func (c *Consumer) SetOAuthBearerTokenFailure(errstr string) error + +
+
+ + func (c *Consumer) StoreOffsets(offsets []TopicPartition) (storedOffsets []TopicPartition, err error) + +
+
+ + func (c *Consumer) String() string + +
+
+ + func (c *Consumer) Subscribe(topic string, rebalanceCb RebalanceCb) error + +
+
+ + func (c *Consumer) SubscribeTopics(topics []string, rebalanceCb RebalanceCb) (err error) + +
+
+ + func (c *Consumer) Subscription() (topics []string, err error) + +
+
+ + func (c *Consumer) Unassign() (err error) + +
+
+ + func (c *Consumer) Unsubscribe() (err error) + +
+
+ + type ConsumerGroupMetadata + +
+
+ + func NewTestConsumerGroupMetadata(groupID string) (*ConsumerGroupMetadata, error) + +
+
+ + type CreatePartitionsAdminOption + +
+
+ + type CreateTopicsAdminOption + +
+
+ + type DeleteTopicsAdminOption + +
+
+ + type DescribeConfigsAdminOption + +
+
+ + type Error + +
+
+ + func NewError(code ErrorCode, str string, fatal bool) (err Error) + +
+
+ + func (e Error) Code() ErrorCode + +
+
+ + func (e Error) Error() string + +
+
+ + func (e Error) IsFatal() bool + +
+
+ + func (e Error) IsRetriable() bool + +
+
+ + func (e Error) String() string + +
+
+ + func (e Error) TxnRequiresAbort() bool + +
+
+ + type ErrorCode + +
+
+ + func (c ErrorCode) String() string + +
+
+ + type Event + +
+
+ + type Handle + +
+
+ + type Header + +
+
+ + func (h Header) String() string + +
+
+ + type LogEvent + +
+
+ + func (logEvent LogEvent) String() string + +
+
+ + type Message + +
+
+ + func (m *Message) String() string + +
+
+ + type Metadata + +
+
+ + type OAuthBearerToken + +
+
+ + type OAuthBearerTokenRefresh + +
+
+ + func (o OAuthBearerTokenRefresh) String() string + +
+
+ + type Offset + +
+
+ + func NewOffset(offset interface{}) (Offset, error) + +
+
+ + func OffsetTail(relativeOffset Offset) Offset + +
+
+ + func (o *Offset) Set(offset interface{}) error + +
+
+ + func (o Offset) String() string + +
+
+ + type OffsetsCommitted + +
+
+ + func (o OffsetsCommitted) String() string + +
+
+ + type PartitionEOF + +
+
+ + func (p PartitionEOF) String() string + +
+
+ + type PartitionMetadata + +
+
+ + type PartitionsSpecification + +
+
+ + type Producer + +
+
+ + func NewProducer(conf *ConfigMap) (*Producer, error) + +
+
+ + func (p *Producer) AbortTransaction(ctx context.Context) error + +
+
+ + func (p *Producer) BeginTransaction() error + +
+
+ + func (p *Producer) Close() + +
+
+ + func (p *Producer) CommitTransaction(ctx context.Context) error + +
+
+ + func (p *Producer) Events() chan Event + +
+
+ + func (p *Producer) Flush(timeoutMs int) int + +
+
+ + func (p *Producer) GetFatalError() error + +
+
+ + func (p *Producer) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error) + +
+
+ + func (p *Producer) InitTransactions(ctx context.Context) error + +
+
+ + func (p *Producer) Len() int + +
+
+ + func (p *Producer) Logs() chan LogEvent + +
+
+ + func (p *Producer) OffsetsForTimes(times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) + +
+
+ + func (p *Producer) Produce(msg *Message, deliveryChan chan Event) error + +
+
+ + func (p *Producer) ProduceChannel() chan *Message + +
+
+ + func (p *Producer) Purge(flags int) error + +
+
+ + func (p *Producer) QueryWatermarkOffsets(topic string, partition int32, timeoutMs int) (low, high int64, err error) + +
+
+ + func (p *Producer) SendOffsetsToTransaction(ctx context.Context, offsets []TopicPartition, consumerMetadata *ConsumerGroupMetadata) error + +
+
+ + func (p *Producer) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error + +
+
+ + func (p *Producer) SetOAuthBearerTokenFailure(errstr string) error + +
+
+ + func (p *Producer) String() string + +
+
+ + func (p *Producer) TestFatalError(code ErrorCode, str string) ErrorCode + +
+
+ + type RebalanceCb + +
+
+ + type ResourceType + +
+
+ + func ResourceTypeFromString(typeString string) (ResourceType, error) + +
+
+ + func (t ResourceType) String() string + +
+
+ + type RevokedPartitions + +
+
+ + func (e RevokedPartitions) String() string + +
+
+ + type Stats + +
+
+ + func (e Stats) String() string + +
+
+ + type TimestampType + +
+
+ + func (t TimestampType) String() string + +
+
+ + type TopicMetadata + +
+
+ + type TopicPartition + +
+
+ + func (p TopicPartition) String() string + +
+
+ + type TopicPartitions + +
+
+ + func (tps TopicPartitions) Len() int + +
+
+ + func (tps TopicPartitions) Less(i, j int) bool + +
+
+ + func (tps TopicPartitions) Swap(i, j int) + +
+
+ + type TopicResult + +
+
+ + func (t TopicResult) String() string + +
+
+ + type TopicSpecification + +
+
+
+ +

+ Package files +

+

+ + + 00version.go + + + adminapi.go + + + adminoptions.go + + + build_glibc_linux.go + + + config.go + + + consumer.go + + + context.go + + + error.go + + + error_gen.go + + + event.go + + + generated_errors.go + + + handle.go + + + header.go + + + kafka.go + + + log.go + + + message.go + + + metadata.go + + + misc.go + + + offset.go + + + producer.go + + + testhelpers.go + + + time.go + + +

+
+ +
+ +

+ Constants +

+ 
const (
+    // ResourceUnknown - Unknown
+    ResourceUnknown = ResourceType(C.RD_KAFKA_RESOURCE_UNKNOWN)
+    // ResourceAny - match any resource type (DescribeConfigs)
+    ResourceAny = ResourceType(C.RD_KAFKA_RESOURCE_ANY)
+    // ResourceTopic - Topic
+    ResourceTopic = ResourceType(C.RD_KAFKA_RESOURCE_TOPIC)
+    // ResourceGroup - Group
+    ResourceGroup = ResourceType(C.RD_KAFKA_RESOURCE_GROUP)
+    // ResourceBroker - Broker
+    ResourceBroker = ResourceType(C.RD_KAFKA_RESOURCE_BROKER)
+)
+ 
const (
+    // ConfigSourceUnknown is the default value
+    ConfigSourceUnknown = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_UNKNOWN_CONFIG)
+    // ConfigSourceDynamicTopic is dynamic topic config that is configured for a specific topic
+    ConfigSourceDynamicTopic = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DYNAMIC_TOPIC_CONFIG)
+    // ConfigSourceDynamicBroker is dynamic broker config that is configured for a specific broker
+    ConfigSourceDynamicBroker = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DYNAMIC_BROKER_CONFIG)
+    // ConfigSourceDynamicDefaultBroker is dynamic broker config that is configured as default for all brokers in the cluster
+    ConfigSourceDynamicDefaultBroker = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DYNAMIC_DEFAULT_BROKER_CONFIG)
+    // ConfigSourceStaticBroker is static broker config provided as broker properties at startup (e.g. from server.properties file)
+    ConfigSourceStaticBroker = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_STATIC_BROKER_CONFIG)
+    // ConfigSourceDefault is built-in default configuration for configs that have a default value
+    ConfigSourceDefault = ConfigSource(C.RD_KAFKA_CONFIG_SOURCE_DEFAULT_CONFIG)
+)
+ 
const (
+    // TimestampNotAvailable indicates no timestamp was set, or not available due to lacking broker support
+    TimestampNotAvailable = TimestampType(C.RD_KAFKA_TIMESTAMP_NOT_AVAILABLE)
+    // TimestampCreateTime indicates timestamp set by producer (source time)
+    TimestampCreateTime = TimestampType(C.RD_KAFKA_TIMESTAMP_CREATE_TIME)
+    // TimestampLogAppendTime indicates timestamp set set by broker (store time)
+    TimestampLogAppendTime = TimestampType(C.RD_KAFKA_TIMESTAMP_LOG_APPEND_TIME)
+)
+ 
const (
+    // PurgeInFlight purges messages in-flight to or from the broker.
+    // Purging these messages will void any future acknowledgements from the
+    // broker, making it impossible for the application to know if these
+    // messages were successfully delivered or not.
+    // Retrying these messages may lead to duplicates.
+    PurgeInFlight = int(C.RD_KAFKA_PURGE_F_INFLIGHT)
+
+    // PurgeQueue Purge messages in internal queues.
+    PurgeQueue = int(C.RD_KAFKA_PURGE_F_QUEUE)
+
+    // PurgeNonBlocking Don't wait for background thread queue purging to finish.
+    PurgeNonBlocking = int(C.RD_KAFKA_PURGE_F_NON_BLOCKING)
+)
+ 
const (
+    // AlterOperationSet sets/overwrites the configuration setting.
+    AlterOperationSet = iota
+)
+

+ LibrdkafkaLinkInfo explains how librdkafka was linked to the Go client +

+ 
const LibrdkafkaLinkInfo = "static glibc_linux from librdkafka-static-bundle-v1.8.2.tgz"
+

+ OffsetBeginning represents the earliest offset (logical) +

+ 
const OffsetBeginning = Offset(C.RD_KAFKA_OFFSET_BEGINNING)
+

+ OffsetEnd represents the latest offset (logical) +

+ 
const OffsetEnd = Offset(C.RD_KAFKA_OFFSET_END)
+

+ OffsetInvalid represents an invalid/unspecified offset +

+ 
const OffsetInvalid = Offset(C.RD_KAFKA_OFFSET_INVALID)
+

+ OffsetStored represents a stored offset +

+ 
const OffsetStored = Offset(C.RD_KAFKA_OFFSET_STORED)
+

+ PartitionAny represents any partition (for partitioning), +or unspecified value (for all other cases) +

+ 
const PartitionAny = int32(C.RD_KAFKA_PARTITION_UA)
+

+ func + + LibraryVersion + + +

+ 
func LibraryVersion() (int, string)
+

+ LibraryVersion returns the underlying librdkafka library version as a +(version_int, version_str) tuple. +

+

+ func + + WriteErrorCodes + + +

+ 
func WriteErrorCodes(f *os.File)
+

+ WriteErrorCodes writes Go error code constants to file from the +librdkafka error codes. +This function is not intended for public use. +

+

+ type + + AdminClient + + +

+

+ AdminClient is derived from an existing Producer or Consumer +

+ 
type AdminClient struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + NewAdminClient + + +

+ 
func NewAdminClient(conf *ConfigMap) (*AdminClient, error)
+

+ NewAdminClient creats a new AdminClient instance with a new underlying client instance +

+

+ func + + NewAdminClientFromConsumer + + +

+ 
func NewAdminClientFromConsumer(c *Consumer) (a *AdminClient, err error)
+

+ NewAdminClientFromConsumer derives a new AdminClient from an existing Consumer instance. +The AdminClient will use the same configuration and connections as the parent instance. +

+

+ func + + NewAdminClientFromProducer + + +

+ 
func NewAdminClientFromProducer(p *Producer) (a *AdminClient, err error)
+

+ NewAdminClientFromProducer derives a new AdminClient from an existing Producer instance. +The AdminClient will use the same configuration and connections as the parent instance. +

+

+ func (*AdminClient) + + AlterConfigs + + +

+ 
func (a *AdminClient) AlterConfigs(ctx context.Context, resources []ConfigResource, options ...AlterConfigsAdminOption) (result []ConfigResourceResult, err error)
+

+ AlterConfigs alters/updates cluster resource configuration. +

+

+ Updates are not transactional so they may succeed for a subset +of the provided resources while others fail. +The configuration for a particular resource is updated atomically, +replacing values using the provided ConfigEntrys and reverting +unspecified ConfigEntrys to their default values. +

+

+ Requires broker version >=0.11.0.0 +

+

+ AlterConfigs will replace all existing configuration for +the provided resources with the new configuration given, +reverting all other configuration to their default values. +

+

+ Multiple resources and resource types may be set, but at most one +resource of type ResourceBroker is allowed per call since these +resource requests must be sent to the broker specified in the resource. +

+

+ func (*AdminClient) + + Close + + +

+ 
func (a *AdminClient) Close()
+

+ Close an AdminClient instance. +

+

+ func (*AdminClient) + + ClusterID + + +

+ 
func (a *AdminClient) ClusterID(ctx context.Context) (clusterID string, err error)
+

+ ClusterID returns the cluster ID as reported in broker metadata. +

+

+ Note on cancellation: Although the underlying C function respects the +timeout, it currently cannot be manually cancelled. That means manually +cancelling the context will block until the C function call returns. +

+

+ Requires broker version >= 0.10.0. +

+

+ func (*AdminClient) + + ControllerID + + +

+ 
func (a *AdminClient) ControllerID(ctx context.Context) (controllerID int32, err error)
+

+ ControllerID returns the broker ID of the current controller as reported in +broker metadata. +

+

+ Note on cancellation: Although the underlying C function respects the +timeout, it currently cannot be manually cancelled. That means manually +cancelling the context will block until the C function call returns. +

+

+ Requires broker version >= 0.10.0. +

+

+ func (*AdminClient) + + CreatePartitions + + +

+ 
func (a *AdminClient) CreatePartitions(ctx context.Context, partitions []PartitionsSpecification, options ...CreatePartitionsAdminOption) (result []TopicResult, err error)
+

+ CreatePartitions creates additional partitions for topics. +

+

+ func (*AdminClient) + + CreateTopics + + +

+ 
func (a *AdminClient) CreateTopics(ctx context.Context, topics []TopicSpecification, options ...CreateTopicsAdminOption) (result []TopicResult, err error)
+

+ CreateTopics creates topics in cluster. +

+

+ The list of TopicSpecification objects define the per-topic partition count, replicas, etc. +

+

+ Topic creation is non-atomic and may succeed for some topics but fail for others, +make sure to check the result for topic-specific errors. +

+

+ Note: TopicSpecification is analogous to NewTopic in the Java Topic Admin API. +

+

+ func (*AdminClient) + + DeleteTopics + + +

+ 
func (a *AdminClient) DeleteTopics(ctx context.Context, topics []string, options ...DeleteTopicsAdminOption) (result []TopicResult, err error)
+

+ DeleteTopics deletes a batch of topics. +

+

+ This operation is not transactional and may succeed for a subset of topics while +failing others. +It may take several seconds after the DeleteTopics result returns success for +all the brokers to become aware that the topics are gone. During this time, +topic metadata and configuration may continue to return information about deleted topics. +

+

+ Requires broker version >= 0.10.1.0 +

+

+ func (*AdminClient) + + DescribeConfigs + + +

+ 
func (a *AdminClient) DescribeConfigs(ctx context.Context, resources []ConfigResource, options ...DescribeConfigsAdminOption) (result []ConfigResourceResult, err error)
+

+ DescribeConfigs retrieves configuration for cluster resources. +

+

+ The returned configuration includes default values, use +ConfigEntryResult.IsDefault or ConfigEntryResult.Source to distinguish +default values from manually configured settings. +

+

+ The value of config entries where .IsSensitive is true +will always be nil to avoid disclosing sensitive +information, such as security settings. +

+

+ Configuration entries where .IsReadOnly is true can't be modified +(with AlterConfigs). +

+

+ Synonym configuration entries are returned if the broker supports +it (broker version >= 1.1.0). See .Synonyms. +

+

+ Requires broker version >=0.11.0.0 +

+

+ Multiple resources and resource types may be requested, but at most +one resource of type ResourceBroker is allowed per call +since these resource requests must be sent to the broker specified +in the resource. +

+

+ func (*AdminClient) + + GetMetadata + + +

+ 
func (a *AdminClient) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error)
+

+ GetMetadata queries broker for cluster and topic metadata. +If topic is non-nil only information about that topic is returned, else if +allTopics is false only information about locally used topics is returned, +else information about all topics is returned. +GetMetadata is equivalent to listTopics, describeTopics and describeCluster in the Java API. +

+

+ func (*AdminClient) + + SetOAuthBearerToken + + +

+ 
func (a *AdminClient) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error
+

+ SetOAuthBearerToken sets the the data to be transmitted +to a broker during SASL/OAUTHBEARER authentication. It will return nil +on success, otherwise an error if: +1) the token data is invalid (meaning an expiration time in the past +or either a token value or an extension key or value that does not meet +the regular expression requirements as per + + https://tools.ietf.org/html/rfc7628#section-3.1 + + ); +2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +3) SASL/OAUTHBEARER is supported but is not configured as the client's +authentication mechanism. +

+

+ func (*AdminClient) + + SetOAuthBearerTokenFailure + + +

+ 
func (a *AdminClient) SetOAuthBearerTokenFailure(errstr string) error
+

+ SetOAuthBearerTokenFailure sets the error message describing why token +retrieval/setting failed; it also schedules a new token refresh event for 10 +seconds later so the attempt may be retried. It will return nil on +success, otherwise an error if: +1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +2) SASL/OAUTHBEARER is supported but is not configured as the client's +authentication mechanism. +

+

+ func (*AdminClient) + + String + + +

+ 
func (a *AdminClient) String() string
+

+ String returns a human readable name for an AdminClient instance +

+

+ type + + AdminOption + + +

+

+ AdminOption is a generic type not to be used directly. +

+

+ See CreateTopicsAdminOption et.al. +

+ 
type AdminOption interface {
+    // contains filtered or unexported methods
+}
+

+ type + + AdminOptionOperationTimeout + + +

+

+ AdminOptionOperationTimeout sets the broker's operation timeout, such as the +timeout for CreateTopics to complete the creation of topics on the controller +before returning a result to the application. +

+

+ CreateTopics, DeleteTopics, CreatePartitions: +a value 0 will return immediately after triggering topic +creation, while > 0 will wait this long for topic creation to propagate +in cluster. +

+

+ Default: 0 (return immediately). +

+

+ Valid for CreateTopics, DeleteTopics, CreatePartitions. +

+ 
type AdminOptionOperationTimeout struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + SetAdminOperationTimeout + + +

+ 
func SetAdminOperationTimeout(t time.Duration) (ao AdminOptionOperationTimeout)
+

+ SetAdminOperationTimeout sets the broker's operation timeout, such as the +timeout for CreateTopics to complete the creation of topics on the controller +before returning a result to the application. +

+

+ CreateTopics, DeleteTopics, CreatePartitions: +a value 0 will return immediately after triggering topic +creation, while > 0 will wait this long for topic creation to propagate +in cluster. +

+

+ Default: 0 (return immediately). +

+

+ Valid for CreateTopics, DeleteTopics, CreatePartitions. +

+

+ type + + AdminOptionRequestTimeout + + +

+

+ AdminOptionRequestTimeout sets the overall request timeout, including broker +lookup, request transmission, operation time on broker, and response. +

+

+ Default: `socket.timeout.ms`. +

+

+ Valid for all Admin API methods. +

+ 
type AdminOptionRequestTimeout struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + SetAdminRequestTimeout + + +

+ 
func SetAdminRequestTimeout(t time.Duration) (ao AdminOptionRequestTimeout)
+

+ SetAdminRequestTimeout sets the overall request timeout, including broker +lookup, request transmission, operation time on broker, and response. +

+

+ Default: `socket.timeout.ms`. +

+

+ Valid for all Admin API methods. +

+

+ type + + AdminOptionValidateOnly + + +

+

+ AdminOptionValidateOnly tells the broker to only validate the request, +without performing the requested operation (create topics, etc). +

+

+ Default: false. +

+

+ Valid for CreateTopics, CreatePartitions, AlterConfigs +

+ 
type AdminOptionValidateOnly struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + SetAdminValidateOnly + + +

+ 
func SetAdminValidateOnly(validateOnly bool) (ao AdminOptionValidateOnly)
+

+ SetAdminValidateOnly tells the broker to only validate the request, +without performing the requested operation (create topics, etc). +

+

+ Default: false. +

+

+ Valid for CreateTopics, DeleteTopics, CreatePartitions, AlterConfigs +

+

+ type + + AlterConfigsAdminOption + + +

+

+ AlterConfigsAdminOption - see setters. +

+

+ See SetAdminRequestTimeout, SetAdminValidateOnly, SetAdminIncremental. +

+ 
type AlterConfigsAdminOption interface {
+    // contains filtered or unexported methods
+}
+

+ type + + AlterOperation + + +

+

+ AlterOperation specifies the operation to perform on the ConfigEntry. +Currently only AlterOperationSet. +

+ 
type AlterOperation int
+

+ func (AlterOperation) + + String + + +

+ 
func (o AlterOperation) String() string
+

+ String returns the human-readable representation of an AlterOperation +

+

+ type + + AssignedPartitions + + +

+

+ AssignedPartitions consumer group rebalance event: assigned partition set +

+ 
type AssignedPartitions struct {
+    Partitions []TopicPartition
+}
+
+

+ func (AssignedPartitions) + + String + + +

+ 
func (e AssignedPartitions) String() string
+

+ type + + BrokerMetadata + + +

+

+ BrokerMetadata contains per-broker metadata +

+ 
type BrokerMetadata struct {
+    ID   int32
+    Host string
+    Port int
+}
+
+

+ type + + ConfigEntry + + +

+

+ ConfigEntry holds parameters for altering a resource's configuration. +

+ 
type ConfigEntry struct {
+    // Name of configuration entry, e.g., topic configuration property name.
+    Name string
+    // Value of configuration entry.
+    Value string
+    // Operation to perform on the entry.
+    Operation AlterOperation
+}
+
+

+ func + + StringMapToConfigEntries + + +

+ 
func StringMapToConfigEntries(stringMap map[string]string, operation AlterOperation) []ConfigEntry
+

+ StringMapToConfigEntries creates a new map of ConfigEntry objects from the +provided string map. The AlterOperation is set on each created entry. +

+

+ func (ConfigEntry) + + String + + +

+ 
func (c ConfigEntry) String() string
+

+ String returns a human-readable representation of a ConfigEntry. +

+

+ type + + ConfigEntryResult + + +

+

+ ConfigEntryResult contains the result of a single configuration entry from a +DescribeConfigs request. +

+ 
type ConfigEntryResult struct {
+    // Name of configuration entry, e.g., topic configuration property name.
+    Name string
+    // Value of configuration entry.
+    Value string
+    // Source indicates the configuration source.
+    Source ConfigSource
+    // IsReadOnly indicates whether the configuration entry can be altered.
+    IsReadOnly bool
+    // IsSensitive indicates whether the configuration entry contains sensitive information, in which case the value will be unset.
+    IsSensitive bool
+    // IsSynonym indicates whether the configuration entry is a synonym for another configuration property.
+    IsSynonym bool
+    // Synonyms contains a map of configuration entries that are synonyms to this configuration entry.
+    Synonyms map[string]ConfigEntryResult
+}
+
+

+ func (ConfigEntryResult) + + String + + +

+ 
func (c ConfigEntryResult) String() string
+

+ String returns a human-readable representation of a ConfigEntryResult. +

+

+ type + + ConfigMap + + +

+

+ ConfigMap is a map containing standard librdkafka configuration properties as documented in: + + https://github.com/edenhill/librdkafka/tree/master/CONFIGURATION.md + +

+

+ The special property "default.topic.config" (optional) is a ConfigMap +containing default topic configuration properties. +

+

+ The use of "default.topic.config" is deprecated, +topic configuration properties shall be specified in the standard ConfigMap. +For backwards compatibility, "default.topic.config" (if supplied) +takes precedence. +

+ 
type ConfigMap map[string]ConfigValue
+

+ func (ConfigMap) + + Get + + +

+ 
func (m ConfigMap) Get(key string, defval ConfigValue) (ConfigValue, error)
+

+ Get finds the given key in the ConfigMap and returns its value. +If the key is not found `defval` is returned. +If the key is found but the type does not match that of `defval` (unless nil) +an ErrInvalidArg error is returned. +

+

+ func (ConfigMap) + + Set + + +

+ 
func (m ConfigMap) Set(kv string) error
+

+ Set implements flag.Set (command line argument parser) as a convenience +for `-X key=value` config. +

+

+ func (ConfigMap) + + SetKey + + +

+ 
func (m ConfigMap) SetKey(key string, value ConfigValue) error
+

+ SetKey sets configuration property key to value. +

+

+ For user convenience a key prefixed with {topic}. will be +set on the "default.topic.config" sub-map, this use is deprecated. +

+

+ type + + ConfigResource + + +

+

+ ConfigResource holds parameters for altering an Apache Kafka configuration resource +

+ 
type ConfigResource struct {
+    // Type of resource to set.
+    Type ResourceType
+    // Name of resource to set.
+    Name string
+    // Config entries to set.
+    // Configuration updates are atomic, any configuration property not provided
+    // here will be reverted (by the broker) to its default value.
+    // Use DescribeConfigs to retrieve the list of current configuration entry values.
+    Config []ConfigEntry
+}
+
+

+ func (ConfigResource) + + String + + +

+ 
func (c ConfigResource) String() string
+

+ String returns a human-readable representation of a ConfigResource +

+

+ type + + ConfigResourceResult + + +

+

+ ConfigResourceResult provides the result for a resource from a AlterConfigs or +DescribeConfigs request. +

+ 
type ConfigResourceResult struct {
+    // Type of returned result resource.
+    Type ResourceType
+    // Name of returned result resource.
+    Name string
+    // Error, if any, of returned result resource.
+    Error Error
+    // Config entries, if any, of returned result resource.
+    Config map[string]ConfigEntryResult
+}
+
+

+ func (ConfigResourceResult) + + String + + +

+ 
func (c ConfigResourceResult) String() string
+

+ String returns a human-readable representation of a ConfigResourceResult. +

+

+ type + + ConfigSource + + +

+

+ ConfigSource represents an Apache Kafka config source +

+ 
type ConfigSource int
+

+ func (ConfigSource) + + String + + +

+ 
func (t ConfigSource) String() string
+

+ String returns the human-readable representation of a ConfigSource type +

+

+ type + + ConfigValue + + +

+

+ ConfigValue supports the following types: +

+ 
bool, int, string, any type with the standard String() interface
+
+ 
type ConfigValue interface{}
+

+ type + + Consumer + + +

+

+ Consumer implements a High-level Apache Kafka Consumer instance +

+ 
type Consumer struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + NewConsumer + + +

+ 
func NewConsumer(conf *ConfigMap) (*Consumer, error)
+

+ NewConsumer creates a new high-level Consumer instance. +

+

+ conf is a *ConfigMap with standard librdkafka configuration properties. +

+

+ Supported special configuration properties: +

+ 
go.application.rebalance.enable (bool, false) - Forward rebalancing responsibility to application via the Events() channel.
+                                     If set to true the app must handle the AssignedPartitions and
+                                     RevokedPartitions events and call Assign() and Unassign()
+                                     respectively.
+go.events.channel.enable (bool, false) - [deprecated] Enable the Events() channel. Messages and events will be pushed on the Events() channel and the Poll() interface will be disabled.
+go.events.channel.size (int, 1000) - Events() channel size
+go.logs.channel.enable (bool, false) - Forward log to Logs() channel.
+go.logs.channel (chan kafka.LogEvent, nil) - Forward logs to application-provided channel instead of Logs(). Requires go.logs.channel.enable=true.
+
+

+ WARNING: Due to the buffering nature of channels (and queues in general) the +use of the events channel risks receiving outdated events and +messages. Minimizing go.events.channel.size reduces the risk +and number of outdated events and messages but does not eliminate +the factor completely. With a channel size of 1 at most one +event or message may be outdated. +

+

+ func (*Consumer) + + Assign + + +

+ 
func (c *Consumer) Assign(partitions []TopicPartition) (err error)
+

+ Assign an atomic set of partitions to consume. +

+

+ The .Offset field of each TopicPartition must either be set to an absolute +starting offset (>= 0), or one of the logical offsets (`kafka.OffsetEnd` etc), +but should typically be set to `kafka.OffsetStored` to have the consumer +use the committed offset as a start position, with a fallback to +`auto.offset.reset` if there is no committed offset. +

+

+ This replaces the current assignment. +

+

+ func (*Consumer) + + Assignment + + +

+ 
func (c *Consumer) Assignment() (partitions []TopicPartition, err error)
+

+ Assignment returns the current partition assignments +

+

+ func (*Consumer) + + AssignmentLost + + +

+ 
func (c *Consumer) AssignmentLost() bool
+

+ AssignmentLost returns true if current partition assignment has been lost. +This method is only applicable for use with a subscribing consumer when +handling a rebalance event or callback. +Partitions that have been lost may already be owned by other members in the +group and therefore commiting offsets, for example, may fail. +

+

+ func (*Consumer) + + Close + + +

+ 
func (c *Consumer) Close() (err error)
+

+ Close Consumer instance. +The object is no longer usable after this call. +

+

+ func (*Consumer) + + Commit + + +

+ 
func (c *Consumer) Commit() ([]TopicPartition, error)
+

+ Commit offsets for currently assigned partitions +This is a blocking call. +Returns the committed offsets on success. +

+

+ func (*Consumer) + + CommitMessage + + +

+ 
func (c *Consumer) CommitMessage(m *Message) ([]TopicPartition, error)
+

+ CommitMessage commits offset based on the provided message. +This is a blocking call. +Returns the committed offsets on success. +

+

+ func (*Consumer) + + CommitOffsets + + +

+ 
func (c *Consumer) CommitOffsets(offsets []TopicPartition) ([]TopicPartition, error)
+

+ CommitOffsets commits the provided list of offsets +This is a blocking call. +Returns the committed offsets on success. +

+

+ func (*Consumer) + + Committed + + +

+ 
func (c *Consumer) Committed(partitions []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error)
+

+ Committed retrieves committed offsets for the given set of partitions +

+

+ func (*Consumer) + + Events + + +

+ 
func (c *Consumer) Events() chan Event
+

+ Events returns the Events channel (if enabled) +

+

+ func (*Consumer) + + GetConsumerGroupMetadata + + +

+ 
func (c *Consumer) GetConsumerGroupMetadata() (*ConsumerGroupMetadata, error)
+

+ GetConsumerGroupMetadata returns the consumer's current group metadata. +This object should be passed to the transactional producer's +SendOffsetsToTransaction() API. +

+

+ func (*Consumer) + + GetMetadata + + +

+ 
func (c *Consumer) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error)
+

+ GetMetadata queries broker for cluster and topic metadata. +If topic is non-nil only information about that topic is returned, else if +allTopics is false only information about locally used topics is returned, +else information about all topics is returned. +GetMetadata is equivalent to listTopics, describeTopics and describeCluster in the Java API. +

+

+ func (*Consumer) + + GetRebalanceProtocol + + +

+ 
func (c *Consumer) GetRebalanceProtocol() string
+

+ GetRebalanceProtocol returns the current consumer group rebalance protocol, +which is either "EAGER" or "COOPERATIVE". +If the rebalance protocol is not known in the current state an empty string +is returned. +Should typically only be called during rebalancing. +

+

+ func (*Consumer) + + GetWatermarkOffsets + + +

+ 
func (c *Consumer) GetWatermarkOffsets(topic string, partition int32) (low, high int64, err error)
+

+ GetWatermarkOffsets returns the cached low and high offsets for the given topic +and partition. The high offset is populated on every fetch response or via calling QueryWatermarkOffsets. +The low offset is populated every statistics.interval.ms if that value is set. +OffsetInvalid will be returned if there is no cached offset for either value. +

+

+ func (*Consumer) + + IncrementalAssign + + +

+ 
func (c *Consumer) IncrementalAssign(partitions []TopicPartition) (err error)
+

+ IncrementalAssign adds the specified partitions to the current set of +partitions to consume. +

+

+ The .Offset field of each TopicPartition must either be set to an absolute +starting offset (>= 0), or one of the logical offsets (`kafka.OffsetEnd` etc), +but should typically be set to `kafka.OffsetStored` to have the consumer +use the committed offset as a start position, with a fallback to +`auto.offset.reset` if there is no committed offset. +

+

+ The new partitions must not be part of the current assignment. +

+

+ func (*Consumer) + + IncrementalUnassign + + +

+ 
func (c *Consumer) IncrementalUnassign(partitions []TopicPartition) (err error)
+

+ IncrementalUnassign removes the specified partitions from the current set of +partitions to consume. +

+

+ The .Offset field of the TopicPartition is ignored. +

+

+ The removed partitions must be part of the current assignment. +

+

+ func (*Consumer) + + Logs + + +

+ 
func (c *Consumer) Logs() chan LogEvent
+

+ Logs returns the log channel if enabled, or nil otherwise. +

+

+ func (*Consumer) + + OffsetsForTimes + + +

+ 
func (c *Consumer) OffsetsForTimes(times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error)
+

+ OffsetsForTimes looks up offsets by timestamp for the given partitions. +

+

+ The returned offset for each partition is the earliest offset whose +timestamp is greater than or equal to the given timestamp in the +corresponding partition. If the provided timestamp exceeds that of the +last message in the partition, a value of -1 will be returned. +

+

+ The timestamps to query are represented as `.Offset` in the `times` +argument and the looked up offsets are represented as `.Offset` in the returned +`offsets` list. +

+

+ The function will block for at most timeoutMs milliseconds. +

+

+ Duplicate Topic+Partitions are not supported. +Per-partition errors may be returned in the `.Error` field. +

+

+ func (*Consumer) + + Pause + + +

+ 
func (c *Consumer) Pause(partitions []TopicPartition) (err error)
+

+ Pause consumption for the provided list of partitions +

+

+ Note that messages already enqueued on the consumer's Event channel +(if `go.events.channel.enable` has been set) will NOT be purged by +this call, set `go.events.channel.size` accordingly. +

+

+ func (*Consumer) + + Poll + + +

+ 
func (c *Consumer) Poll(timeoutMs int) (event Event)
+

+ Poll the consumer for messages or events. +

+

+ Will block for at most timeoutMs milliseconds +

+

+ The following callbacks may be triggered: +

+ 
Subscribe()'s rebalanceCb
+
+

+ Returns nil on timeout, else an Event +

+

+ func (*Consumer) + + Position + + +

+ 
func (c *Consumer) Position(partitions []TopicPartition) (offsets []TopicPartition, err error)
+

+ Position returns the current consume position for the given partitions. +Typical use is to call Assignment() to get the partition list +and then pass it to Position() to get the current consume position for +each of the assigned partitions. +The consume position is the next message to read from the partition. +i.e., the offset of the last message seen by the application + 1. +

+

+ func (*Consumer) + + QueryWatermarkOffsets + + +

+ 
func (c *Consumer) QueryWatermarkOffsets(topic string, partition int32, timeoutMs int) (low, high int64, err error)
+

+ QueryWatermarkOffsets queries the broker for the low and high offsets for the given topic and partition. +

+

+ func (*Consumer) + + ReadMessage + + +

+ 
func (c *Consumer) ReadMessage(timeout time.Duration) (*Message, error)
+

+ ReadMessage polls the consumer for a message. +

+

+ This is a convenience API that wraps Poll() and only returns +messages or errors. All other event types are discarded. +

+

+ The call will block for at most `timeout` waiting for +a new message or error. `timeout` may be set to -1 for +indefinite wait. +

+

+ Timeout is returned as (nil, err) where err is `err.(kafka.Error).Code() == kafka.ErrTimedOut`. +

+

+ Messages are returned as (msg, nil), +while general errors are returned as (nil, err), +and partition-specific errors are returned as (msg, err) where +msg.TopicPartition provides partition-specific information (such as topic, partition and offset). +

+

+ All other event types, such as PartitionEOF, AssignedPartitions, etc, are silently discarded. +

+

+ func (*Consumer) + + Resume + + +

+ 
func (c *Consumer) Resume(partitions []TopicPartition) (err error)
+

+ Resume consumption for the provided list of partitions +

+

+ func (*Consumer) + + Seek + + +

+ 
func (c *Consumer) Seek(partition TopicPartition, timeoutMs int) error
+

+ Seek seeks the given topic partitions using the offset from the TopicPartition. +

+

+ If timeoutMs is not 0 the call will wait this long for the +seek to be performed. If the timeout is reached the internal state +will be unknown and this function returns ErrTimedOut. +If timeoutMs is 0 it will initiate the seek but return +immediately without any error reporting (e.g., async). +

+

+ Seek() may only be used for partitions already being consumed +(through Assign() or implicitly through a self-rebalanced Subscribe()). +To set the starting offset it is preferred to use Assign() and provide +a starting offset for each partition. +

+

+ Returns an error on failure or nil otherwise. +

+

+ func (*Consumer) + + SetOAuthBearerToken + + +

+ 
func (c *Consumer) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error
+

+ SetOAuthBearerToken sets the the data to be transmitted +to a broker during SASL/OAUTHBEARER authentication. It will return nil +on success, otherwise an error if: +1) the token data is invalid (meaning an expiration time in the past +or either a token value or an extension key or value that does not meet +the regular expression requirements as per + + https://tools.ietf.org/html/rfc7628#section-3.1 + + ); +2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +3) SASL/OAUTHBEARER is supported but is not configured as the client's +authentication mechanism. +

+

+ func (*Consumer) + + SetOAuthBearerTokenFailure + + +

+ 
func (c *Consumer) SetOAuthBearerTokenFailure(errstr string) error
+

+ SetOAuthBearerTokenFailure sets the error message describing why token +retrieval/setting failed; it also schedules a new token refresh event for 10 +seconds later so the attempt may be retried. It will return nil on +success, otherwise an error if: +1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +2) SASL/OAUTHBEARER is supported but is not configured as the client's +authentication mechanism. +

+

+ func (*Consumer) + + StoreOffsets + + +

+ 
func (c *Consumer) StoreOffsets(offsets []TopicPartition) (storedOffsets []TopicPartition, err error)
+

+ StoreOffsets stores the provided list of offsets that will be committed +to the offset store according to `auto.commit.interval.ms` or manual +offset-less Commit(). +

+

+ Returns the stored offsets on success. If at least one offset couldn't be stored, +an error and a list of offsets is returned. Each offset can be checked for +specific errors via its `.Error` member. +

+

+ func (*Consumer) + + String + + +

+ 
func (c *Consumer) String() string
+

+ Strings returns a human readable name for a Consumer instance +

+

+ func (*Consumer) + + Subscribe + + +

+ 
func (c *Consumer) Subscribe(topic string, rebalanceCb RebalanceCb) error
+

+ Subscribe to a single topic +This replaces the current subscription +

+

+ func (*Consumer) + + SubscribeTopics + + +

+ 
func (c *Consumer) SubscribeTopics(topics []string, rebalanceCb RebalanceCb) (err error)
+

+ SubscribeTopics subscribes to the provided list of topics. +This replaces the current subscription. +

+

+ func (*Consumer) + + Subscription + + +

+ 
func (c *Consumer) Subscription() (topics []string, err error)
+

+ Subscription returns the current subscription as set by Subscribe() +

+

+ func (*Consumer) + + Unassign + + +

+ 
func (c *Consumer) Unassign() (err error)
+

+ Unassign the current set of partitions to consume. +

+

+ func (*Consumer) + + Unsubscribe + + +

+ 
func (c *Consumer) Unsubscribe() (err error)
+

+ Unsubscribe from the current subscription, if any. +

+

+ type + + ConsumerGroupMetadata + + +

+

+ ConsumerGroupMetadata reflects the current consumer group member metadata. +

+ 
type ConsumerGroupMetadata struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + NewTestConsumerGroupMetadata + + +

+ 
func NewTestConsumerGroupMetadata(groupID string) (*ConsumerGroupMetadata, error)
+

+ NewTestConsumerGroupMetadata creates a new consumer group metadata instance +mainly for testing use. +Use GetConsumerGroupMetadata() to retrieve the real metadata. +

+

+ type + + CreatePartitionsAdminOption + + +

+

+ CreatePartitionsAdminOption - see setters. +

+

+ See SetAdminRequestTimeout, SetAdminOperationTimeout, SetAdminValidateOnly. +

+ 
type CreatePartitionsAdminOption interface {
+    // contains filtered or unexported methods
+}
+

+ type + + CreateTopicsAdminOption + + +

+

+ CreateTopicsAdminOption - see setters. +

+

+ See SetAdminRequestTimeout, SetAdminOperationTimeout, SetAdminValidateOnly. +

+ 
type CreateTopicsAdminOption interface {
+    // contains filtered or unexported methods
+}
+

+ type + + DeleteTopicsAdminOption + + +

+

+ DeleteTopicsAdminOption - see setters. +

+

+ See SetAdminRequestTimeout, SetAdminOperationTimeout. +

+ 
type DeleteTopicsAdminOption interface {
+    // contains filtered or unexported methods
+}
+

+ type + + DescribeConfigsAdminOption + + +

+

+ DescribeConfigsAdminOption - see setters. +

+

+ See SetAdminRequestTimeout. +

+ 
type DescribeConfigsAdminOption interface {
+    // contains filtered or unexported methods
+}
+

+ type + + Error + + +

+

+ Error provides a Kafka-specific error container +

+ 
type Error struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + NewError + + +

+ 
func NewError(code ErrorCode, str string, fatal bool) (err Error)
+

+ NewError creates a new Error. +

+

+ func (Error) + + Code + + +

+ 
func (e Error) Code() ErrorCode
+

+ Code returns the ErrorCode of an Error +

+

+ func (Error) + + Error + + +

+ 
func (e Error) Error() string
+

+ Error returns a human readable representation of an Error +Same as Error.String() +

+

+ func (Error) + + IsFatal + + +

+ 
func (e Error) IsFatal() bool
+

+ IsFatal returns true if the error is a fatal error. +A fatal error indicates the client instance is no longer operable and +should be terminated. Typical causes include non-recoverable +idempotent producer errors. +

+

+ func (Error) + + IsRetriable + + +

+ 
func (e Error) IsRetriable() bool
+

+ IsRetriable returns true if the operation that caused this error +may be retried. +This flag is currently only set by the Transactional producer API. +

+

+ func (Error) + + String + + +

+ 
func (e Error) String() string
+

+ String returns a human readable representation of an Error +

+

+ func (Error) + + TxnRequiresAbort + + +

+ 
func (e Error) TxnRequiresAbort() bool
+

+ TxnRequiresAbort returns true if the error is an abortable transaction error +that requires the application to abort the current transaction with +AbortTransaction() and start a new transaction with BeginTransaction() +if it wishes to proceed with transactional operations. +This flag is only set by the Transactional producer API. +

+

+ type + + ErrorCode + + +

+

+ ErrorCode is the integer representation of local and broker error codes +

+ 
type ErrorCode int
+ 
const (
+    // ErrBadMsg Local: Bad message format
+    ErrBadMsg ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__BAD_MSG)
+    // ErrBadCompression Local: Invalid compressed data
+    ErrBadCompression ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__BAD_COMPRESSION)
+    // ErrDestroy Local: Broker handle destroyed
+    ErrDestroy ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__DESTROY)
+    // ErrFail Local: Communication failure with broker
+    ErrFail ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FAIL)
+    // ErrTransport Local: Broker transport failure
+    ErrTransport ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__TRANSPORT)
+    // ErrCritSysResource Local: Critical system resource failure
+    ErrCritSysResource ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__CRIT_SYS_RESOURCE)
+    // ErrResolve Local: Host resolution failure
+    ErrResolve ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__RESOLVE)
+    // ErrMsgTimedOut Local: Message timed out
+    ErrMsgTimedOut ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__MSG_TIMED_OUT)
+    // ErrPartitionEOF Broker: No more messages
+    ErrPartitionEOF ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PARTITION_EOF)
+    // ErrUnknownPartition Local: Unknown partition
+    ErrUnknownPartition ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION)
+    // ErrFs Local: File or filesystem error
+    ErrFs ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FS)
+    // ErrUnknownTopic Local: Unknown topic
+    ErrUnknownTopic ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_TOPIC)
+    // ErrAllBrokersDown Local: All broker connections are down
+    ErrAllBrokersDown ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ALL_BROKERS_DOWN)
+    // ErrInvalidArg Local: Invalid argument or configuration
+    ErrInvalidArg ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INVALID_ARG)
+    // ErrTimedOut Local: Timed out
+    ErrTimedOut ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__TIMED_OUT)
+    // ErrQueueFull Local: Queue full
+    ErrQueueFull ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__QUEUE_FULL)
+    // ErrIsrInsuff Local: ISR count insufficient
+    ErrIsrInsuff ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ISR_INSUFF)
+    // ErrNodeUpdate Local: Broker node update
+    ErrNodeUpdate ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NODE_UPDATE)
+    // ErrSsl Local: SSL error
+    ErrSsl ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__SSL)
+    // ErrWaitCoord Local: Waiting for coordinator
+    ErrWaitCoord ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__WAIT_COORD)
+    // ErrUnknownGroup Local: Unknown group
+    ErrUnknownGroup ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_GROUP)
+    // ErrInProgress Local: Operation in progress
+    ErrInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__IN_PROGRESS)
+    // ErrPrevInProgress Local: Previous operation in progress
+    ErrPrevInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PREV_IN_PROGRESS)
+    // ErrExistingSubscription Local: Existing subscription
+    ErrExistingSubscription ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__EXISTING_SUBSCRIPTION)
+    // ErrAssignPartitions Local: Assign partitions
+    ErrAssignPartitions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS)
+    // ErrRevokePartitions Local: Revoke partitions
+    ErrRevokePartitions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS)
+    // ErrConflict Local: Conflicting use
+    ErrConflict ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__CONFLICT)
+    // ErrState Local: Erroneous state
+    ErrState ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__STATE)
+    // ErrUnknownProtocol Local: Unknown protocol
+    ErrUnknownProtocol ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_PROTOCOL)
+    // ErrNotImplemented Local: Not implemented
+    ErrNotImplemented ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED)
+    // ErrAuthentication Local: Authentication failure
+    ErrAuthentication ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__AUTHENTICATION)
+    // ErrNoOffset Local: No offset stored
+    ErrNoOffset ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NO_OFFSET)
+    // ErrOutdated Local: Outdated
+    ErrOutdated ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__OUTDATED)
+    // ErrTimedOutQueue Local: Timed out in queue
+    ErrTimedOutQueue ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__TIMED_OUT_QUEUE)
+    // ErrUnsupportedFeature Local: Required feature not supported by broker
+    ErrUnsupportedFeature ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNSUPPORTED_FEATURE)
+    // ErrWaitCache Local: Awaiting cache update
+    ErrWaitCache ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__WAIT_CACHE)
+    // ErrIntr Local: Operation interrupted
+    ErrIntr ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INTR)
+    // ErrKeySerialization Local: Key serialization error
+    ErrKeySerialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__KEY_SERIALIZATION)
+    // ErrValueSerialization Local: Value serialization error
+    ErrValueSerialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__VALUE_SERIALIZATION)
+    // ErrKeyDeserialization Local: Key deserialization error
+    ErrKeyDeserialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__KEY_DESERIALIZATION)
+    // ErrValueDeserialization Local: Value deserialization error
+    ErrValueDeserialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__VALUE_DESERIALIZATION)
+    // ErrPartial Local: Partial response
+    ErrPartial ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PARTIAL)
+    // ErrReadOnly Local: Read-only object
+    ErrReadOnly ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__READ_ONLY)
+    // ErrNoent Local: No such entry
+    ErrNoent ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOENT)
+    // ErrUnderflow Local: Read underflow
+    ErrUnderflow ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNDERFLOW)
+    // ErrInvalidType Local: Invalid type
+    ErrInvalidType ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INVALID_TYPE)
+    // ErrRetry Local: Retry operation
+    ErrRetry ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__RETRY)
+    // ErrPurgeQueue Local: Purged in queue
+    ErrPurgeQueue ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PURGE_QUEUE)
+    // ErrPurgeInflight Local: Purged in flight
+    ErrPurgeInflight ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PURGE_INFLIGHT)
+    // ErrFatal Local: Fatal error
+    ErrFatal ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FATAL)
+    // ErrInconsistent Local: Inconsistent state
+    ErrInconsistent ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INCONSISTENT)
+    // ErrGaplessGuarantee Local: Gap-less ordering would not be guaranteed if proceeding
+    ErrGaplessGuarantee ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__GAPLESS_GUARANTEE)
+    // ErrMaxPollExceeded Local: Maximum application poll interval (max.poll.interval.ms) exceeded
+    ErrMaxPollExceeded ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__MAX_POLL_EXCEEDED)
+    // ErrUnknownBroker Local: Unknown broker
+    ErrUnknownBroker ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_BROKER)
+    // ErrNotConfigured Local: Functionality not configured
+    ErrNotConfigured ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOT_CONFIGURED)
+    // ErrFenced Local: This instance has been fenced by a newer instance
+    ErrFenced ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FENCED)
+    // ErrApplication Local: Application generated error
+    ErrApplication ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__APPLICATION)
+    // ErrAssignmentLost Local: Group partition assignment lost
+    ErrAssignmentLost ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ASSIGNMENT_LOST)
+    // ErrNoop Local: No operation performed
+    ErrNoop ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOOP)
+    // ErrAutoOffsetReset Local: No offset to automatically reset to
+    ErrAutoOffsetReset ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__AUTO_OFFSET_RESET)
+    // ErrUnknown Unknown broker error
+    ErrUnknown ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN)
+    // ErrNoError Success
+    ErrNoError ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NO_ERROR)
+    // ErrOffsetOutOfRange Broker: Offset out of range
+    ErrOffsetOutOfRange ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OFFSET_OUT_OF_RANGE)
+    // ErrInvalidMsg Broker: Invalid message
+    ErrInvalidMsg ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_MSG)
+    // ErrUnknownTopicOrPart Broker: Unknown topic or partition
+    ErrUnknownTopicOrPart ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_TOPIC_OR_PART)
+    // ErrInvalidMsgSize Broker: Invalid message size
+    ErrInvalidMsgSize ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_MSG_SIZE)
+    // ErrLeaderNotAvailable Broker: Leader not available
+    ErrLeaderNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_LEADER_NOT_AVAILABLE)
+    // ErrNotLeaderForPartition Broker: Not leader for partition
+    ErrNotLeaderForPartition ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_LEADER_FOR_PARTITION)
+    // ErrRequestTimedOut Broker: Request timed out
+    ErrRequestTimedOut ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REQUEST_TIMED_OUT)
+    // ErrBrokerNotAvailable Broker: Broker not available
+    ErrBrokerNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_BROKER_NOT_AVAILABLE)
+    // ErrReplicaNotAvailable Broker: Replica not available
+    ErrReplicaNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REPLICA_NOT_AVAILABLE)
+    // ErrMsgSizeTooLarge Broker: Message size too large
+    ErrMsgSizeTooLarge ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_MSG_SIZE_TOO_LARGE)
+    // ErrStaleCtrlEpoch Broker: StaleControllerEpochCode
+    ErrStaleCtrlEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_STALE_CTRL_EPOCH)
+    // ErrOffsetMetadataTooLarge Broker: Offset metadata string too large
+    ErrOffsetMetadataTooLarge ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OFFSET_METADATA_TOO_LARGE)
+    // ErrNetworkException Broker: Broker disconnected before response received
+    ErrNetworkException ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NETWORK_EXCEPTION)
+    // ErrCoordinatorLoadInProgress Broker: Coordinator load in progress
+    ErrCoordinatorLoadInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_COORDINATOR_LOAD_IN_PROGRESS)
+    // ErrCoordinatorNotAvailable Broker: Coordinator not available
+    ErrCoordinatorNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_COORDINATOR_NOT_AVAILABLE)
+    // ErrNotCoordinator Broker: Not coordinator
+    ErrNotCoordinator ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_COORDINATOR)
+    // ErrTopicException Broker: Invalid topic
+    ErrTopicException ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_EXCEPTION)
+    // ErrRecordListTooLarge Broker: Message batch larger than configured server segment size
+    ErrRecordListTooLarge ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_RECORD_LIST_TOO_LARGE)
+    // ErrNotEnoughReplicas Broker: Not enough in-sync replicas
+    ErrNotEnoughReplicas ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_ENOUGH_REPLICAS)
+    // ErrNotEnoughReplicasAfterAppend Broker: Message(s) written to insufficient number of in-sync replicas
+    ErrNotEnoughReplicasAfterAppend ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_ENOUGH_REPLICAS_AFTER_APPEND)
+    // ErrInvalidRequiredAcks Broker: Invalid required acks value
+    ErrInvalidRequiredAcks ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REQUIRED_ACKS)
+    // ErrIllegalGeneration Broker: Specified group generation id is not valid
+    ErrIllegalGeneration ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ILLEGAL_GENERATION)
+    // ErrInconsistentGroupProtocol Broker: Inconsistent group protocol
+    ErrInconsistentGroupProtocol ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INCONSISTENT_GROUP_PROTOCOL)
+    // ErrInvalidGroupID Broker: Invalid group.id
+    ErrInvalidGroupID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_GROUP_ID)
+    // ErrUnknownMemberID Broker: Unknown member
+    ErrUnknownMemberID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_MEMBER_ID)
+    // ErrInvalidSessionTimeout Broker: Invalid session timeout
+    ErrInvalidSessionTimeout ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_SESSION_TIMEOUT)
+    // ErrRebalanceInProgress Broker: Group rebalance in progress
+    ErrRebalanceInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REBALANCE_IN_PROGRESS)
+    // ErrInvalidCommitOffsetSize Broker: Commit offset data size is not valid
+    ErrInvalidCommitOffsetSize ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_COMMIT_OFFSET_SIZE)
+    // ErrTopicAuthorizationFailed Broker: Topic authorization failed
+    ErrTopicAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_AUTHORIZATION_FAILED)
+    // ErrGroupAuthorizationFailed Broker: Group authorization failed
+    ErrGroupAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_AUTHORIZATION_FAILED)
+    // ErrClusterAuthorizationFailed Broker: Cluster authorization failed
+    ErrClusterAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_CLUSTER_AUTHORIZATION_FAILED)
+    // ErrInvalidTimestamp Broker: Invalid timestamp
+    ErrInvalidTimestamp ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_TIMESTAMP)
+    // ErrUnsupportedSaslMechanism Broker: Unsupported SASL mechanism
+    ErrUnsupportedSaslMechanism ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_SASL_MECHANISM)
+    // ErrIllegalSaslState Broker: Request not valid in current SASL state
+    ErrIllegalSaslState ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ILLEGAL_SASL_STATE)
+    // ErrUnsupportedVersion Broker: API version not supported
+    ErrUnsupportedVersion ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_VERSION)
+    // ErrTopicAlreadyExists Broker: Topic already exists
+    ErrTopicAlreadyExists ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_ALREADY_EXISTS)
+    // ErrInvalidPartitions Broker: Invalid number of partitions
+    ErrInvalidPartitions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PARTITIONS)
+    // ErrInvalidReplicationFactor Broker: Invalid replication factor
+    ErrInvalidReplicationFactor ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REPLICATION_FACTOR)
+    // ErrInvalidReplicaAssignment Broker: Invalid replica assignment
+    ErrInvalidReplicaAssignment ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REPLICA_ASSIGNMENT)
+    // ErrInvalidConfig Broker: Configuration is invalid
+    ErrInvalidConfig ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_CONFIG)
+    // ErrNotController Broker: Not controller for cluster
+    ErrNotController ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_CONTROLLER)
+    // ErrInvalidRequest Broker: Invalid request
+    ErrInvalidRequest ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REQUEST)
+    // ErrUnsupportedForMessageFormat Broker: Message format on broker does not support request
+    ErrUnsupportedForMessageFormat ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_FOR_MESSAGE_FORMAT)
+    // ErrPolicyViolation Broker: Policy violation
+    ErrPolicyViolation ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_POLICY_VIOLATION)
+    // ErrOutOfOrderSequenceNumber Broker: Broker received an out of order sequence number
+    ErrOutOfOrderSequenceNumber ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OUT_OF_ORDER_SEQUENCE_NUMBER)
+    // ErrDuplicateSequenceNumber Broker: Broker received a duplicate sequence number
+    ErrDuplicateSequenceNumber ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DUPLICATE_SEQUENCE_NUMBER)
+    // ErrInvalidProducerEpoch Broker: Producer attempted an operation with an old epoch
+    ErrInvalidProducerEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PRODUCER_EPOCH)
+    // ErrInvalidTxnState Broker: Producer attempted a transactional operation in an invalid state
+    ErrInvalidTxnState ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_TXN_STATE)
+    // ErrInvalidProducerIDMapping Broker: Producer attempted to use a producer id which is not currently assigned to its transactional id
+    ErrInvalidProducerIDMapping ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PRODUCER_ID_MAPPING)
+    // ErrInvalidTransactionTimeout Broker: Transaction timeout is larger than the maximum value allowed by the broker's max.transaction.timeout.ms
+    ErrInvalidTransactionTimeout ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_TRANSACTION_TIMEOUT)
+    // ErrConcurrentTransactions Broker: Producer attempted to update a transaction while another concurrent operation on the same transaction was ongoing
+    ErrConcurrentTransactions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_CONCURRENT_TRANSACTIONS)
+    // ErrTransactionCoordinatorFenced Broker: Indicates that the transaction coordinator sending a WriteTxnMarker is no longer the current coordinator for a given producer
+    ErrTransactionCoordinatorFenced ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TRANSACTION_COORDINATOR_FENCED)
+    // ErrTransactionalIDAuthorizationFailed Broker: Transactional Id authorization failed
+    ErrTransactionalIDAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TRANSACTIONAL_ID_AUTHORIZATION_FAILED)
+    // ErrSecurityDisabled Broker: Security features are disabled
+    ErrSecurityDisabled ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_SECURITY_DISABLED)
+    // ErrOperationNotAttempted Broker: Operation not attempted
+    ErrOperationNotAttempted ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OPERATION_NOT_ATTEMPTED)
+    // ErrKafkaStorageError Broker: Disk error when trying to access log file on disk
+    ErrKafkaStorageError ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_KAFKA_STORAGE_ERROR)
+    // ErrLogDirNotFound Broker: The user-specified log directory is not found in the broker config
+    ErrLogDirNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_LOG_DIR_NOT_FOUND)
+    // ErrSaslAuthenticationFailed Broker: SASL Authentication failed
+    ErrSaslAuthenticationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_SASL_AUTHENTICATION_FAILED)
+    // ErrUnknownProducerID Broker: Unknown Producer Id
+    ErrUnknownProducerID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_PRODUCER_ID)
+    // ErrReassignmentInProgress Broker: Partition reassignment is in progress
+    ErrReassignmentInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REASSIGNMENT_IN_PROGRESS)
+    // ErrDelegationTokenAuthDisabled Broker: Delegation Token feature is not enabled
+    ErrDelegationTokenAuthDisabled ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_AUTH_DISABLED)
+    // ErrDelegationTokenNotFound Broker: Delegation Token is not found on server
+    ErrDelegationTokenNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_NOT_FOUND)
+    // ErrDelegationTokenOwnerMismatch Broker: Specified Principal is not valid Owner/Renewer
+    ErrDelegationTokenOwnerMismatch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_OWNER_MISMATCH)
+    // ErrDelegationTokenRequestNotAllowed Broker: Delegation Token requests are not allowed on this connection
+    ErrDelegationTokenRequestNotAllowed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_REQUEST_NOT_ALLOWED)
+    // ErrDelegationTokenAuthorizationFailed Broker: Delegation Token authorization failed
+    ErrDelegationTokenAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_AUTHORIZATION_FAILED)
+    // ErrDelegationTokenExpired Broker: Delegation Token is expired
+    ErrDelegationTokenExpired ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_EXPIRED)
+    // ErrInvalidPrincipalType Broker: Supplied principalType is not supported
+    ErrInvalidPrincipalType ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PRINCIPAL_TYPE)
+    // ErrNonEmptyGroup Broker: The group is not empty
+    ErrNonEmptyGroup ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NON_EMPTY_GROUP)
+    // ErrGroupIDNotFound Broker: The group id does not exist
+    ErrGroupIDNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_ID_NOT_FOUND)
+    // ErrFetchSessionIDNotFound Broker: The fetch session ID was not found
+    ErrFetchSessionIDNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FETCH_SESSION_ID_NOT_FOUND)
+    // ErrInvalidFetchSessionEpoch Broker: The fetch session epoch is invalid
+    ErrInvalidFetchSessionEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_FETCH_SESSION_EPOCH)
+    // ErrListenerNotFound Broker: No matching listener
+    ErrListenerNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_LISTENER_NOT_FOUND)
+    // ErrTopicDeletionDisabled Broker: Topic deletion is disabled
+    ErrTopicDeletionDisabled ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_DELETION_DISABLED)
+    // ErrFencedLeaderEpoch Broker: Leader epoch is older than broker epoch
+    ErrFencedLeaderEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FENCED_LEADER_EPOCH)
+    // ErrUnknownLeaderEpoch Broker: Leader epoch is newer than broker epoch
+    ErrUnknownLeaderEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_LEADER_EPOCH)
+    // ErrUnsupportedCompressionType Broker: Unsupported compression type
+    ErrUnsupportedCompressionType ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_COMPRESSION_TYPE)
+    // ErrStaleBrokerEpoch Broker: Broker epoch has changed
+    ErrStaleBrokerEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_STALE_BROKER_EPOCH)
+    // ErrOffsetNotAvailable Broker: Leader high watermark is not caught up
+    ErrOffsetNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OFFSET_NOT_AVAILABLE)
+    // ErrMemberIDRequired Broker: Group member needs a valid member ID
+    ErrMemberIDRequired ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_MEMBER_ID_REQUIRED)
+    // ErrPreferredLeaderNotAvailable Broker: Preferred leader was not available
+    ErrPreferredLeaderNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_PREFERRED_LEADER_NOT_AVAILABLE)
+    // ErrGroupMaxSizeReached Broker: Consumer group has reached maximum size
+    ErrGroupMaxSizeReached ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_MAX_SIZE_REACHED)
+    // ErrFencedInstanceID Broker: Static consumer fenced by other consumer with same group.instance.id
+    ErrFencedInstanceID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FENCED_INSTANCE_ID)
+    // ErrEligibleLeadersNotAvailable Broker: Eligible partition leaders are not available
+    ErrEligibleLeadersNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ELIGIBLE_LEADERS_NOT_AVAILABLE)
+    // ErrElectionNotNeeded Broker: Leader election not needed for topic partition
+    ErrElectionNotNeeded ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ELECTION_NOT_NEEDED)
+    // ErrNoReassignmentInProgress Broker: No partition reassignment is in progress
+    ErrNoReassignmentInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NO_REASSIGNMENT_IN_PROGRESS)
+    // ErrGroupSubscribedToTopic Broker: Deleting offsets of a topic while the consumer group is subscribed to it
+    ErrGroupSubscribedToTopic ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_SUBSCRIBED_TO_TOPIC)
+    // ErrInvalidRecord Broker: Broker failed to validate record
+    ErrInvalidRecord ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_RECORD)
+    // ErrUnstableOffsetCommit Broker: There are unstable offsets that need to be cleared
+    ErrUnstableOffsetCommit ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSTABLE_OFFSET_COMMIT)
+    // ErrThrottlingQuotaExceeded Broker: Throttling quota has been exceeded
+    ErrThrottlingQuotaExceeded ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_THROTTLING_QUOTA_EXCEEDED)
+    // ErrProducerFenced Broker: There is a newer producer with the same transactionalId which fences the current one
+    ErrProducerFenced ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_PRODUCER_FENCED)
+    // ErrResourceNotFound Broker: Request illegally referred to resource that does not exist
+    ErrResourceNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_RESOURCE_NOT_FOUND)
+    // ErrDuplicateResource Broker: Request illegally referred to the same resource twice
+    ErrDuplicateResource ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DUPLICATE_RESOURCE)
+    // ErrUnacceptableCredential Broker: Requested credential would not meet criteria for acceptability
+    ErrUnacceptableCredential ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNACCEPTABLE_CREDENTIAL)
+    // ErrInconsistentVoterSet Broker: Indicates that the either the sender or recipient of a voter-only request is not one of the expected voters
+    ErrInconsistentVoterSet ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INCONSISTENT_VOTER_SET)
+    // ErrInvalidUpdateVersion Broker: Invalid update version
+    ErrInvalidUpdateVersion ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_UPDATE_VERSION)
+    // ErrFeatureUpdateFailed Broker: Unable to update finalized features due to server error
+    ErrFeatureUpdateFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FEATURE_UPDATE_FAILED)
+    // ErrPrincipalDeserializationFailure Broker: Request principal deserialization failed during forwarding
+    ErrPrincipalDeserializationFailure ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_PRINCIPAL_DESERIALIZATION_FAILURE)
+)
+

+ func (ErrorCode) + + String + + +

+ 
func (c ErrorCode) String() string
+

+ String returns a human readable representation of an error code +

+

+ type + + Event + + +

+

+ Event generic interface +

+ 
type Event interface {
+    // String returns a human-readable representation of the event
+    String() string
+}
+

+ type + + Handle + + +

+

+ Handle represents a generic client handle containing common parts for +both Producer and Consumer. +

+ 
type Handle interface {
+    // SetOAuthBearerToken sets the the data to be transmitted
+    // to a broker during SASL/OAUTHBEARER authentication. It will return nil
+    // on success, otherwise an error if:
+    // 1) the token data is invalid (meaning an expiration time in the past
+    // or either a token value or an extension key or value that does not meet
+    // the regular expression requirements as per
+    // https://tools.ietf.org/html/rfc7628#section-3.1);
+    // 2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build;
+    // 3) SASL/OAUTHBEARER is supported but is not configured as the client's
+    // authentication mechanism.
+    SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error
+
+    // SetOAuthBearerTokenFailure sets the error message describing why token
+    // retrieval/setting failed; it also schedules a new token refresh event for 10
+    // seconds later so the attempt may be retried. It will return nil on
+    // success, otherwise an error if:
+    // 1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build;
+    // 2) SASL/OAUTHBEARER is supported but is not configured as the client's
+    // authentication mechanism.
+    SetOAuthBearerTokenFailure(errstr string) error
+    // contains filtered or unexported methods
+}
+ +

+ Header represents a single Kafka message header. +

+

+ Message headers are made up of a list of Header elements, retaining their original insert +order and allowing for duplicate Keys. +

+

+ Key is a human readable string identifying the header. +Value is the key's binary value, Kafka does not put any restrictions on the format of +of the Value but it should be made relatively compact. +The value may be a byte array, empty, or nil. +

+

+ NOTE: Message headers are not available on producer delivery report messages. +

+ 
type Header struct {
+    Key   string // Header name (utf-8 string)
+    Value []byte // Header value (nil, empty, or binary)
+}
+
+

+ func (Header) + + String + + +

+ 
func (h Header) String() string
+

+ String returns the Header Key and data in a human representable possibly truncated form +suitable for displaying to the user. +

+

+ type + + LogEvent + + +

+

+ LogEvent represent the log from librdkafka internal log queue +

+ 
type LogEvent struct {
+    Name      string    // Name of client instance
+    Tag       string    // Log tag that provides context to the log Message (e.g., "METADATA" or "GRPCOORD")
+    Message   string    // Log message
+    Level     int       // Log syslog level, lower is more critical.
+    Timestamp time.Time // Log timestamp
+}
+
+

+ func (LogEvent) + + String + + +

+ 
func (logEvent LogEvent) String() string
+

+ type + + Message + + +

+

+ Message represents a Kafka message +

+ 
type Message struct {
+    TopicPartition TopicPartition
+    Value          []byte
+    Key            []byte
+    Timestamp      time.Time
+    TimestampType  TimestampType
+    Opaque         interface{}
+    Headers        []Header
+}
+
+

+ func (*Message) + + String + + +

+ 
func (m *Message) String() string
+

+ String returns a human readable representation of a Message. +Key and payload are not represented. +

+

+ type + + Metadata + + +

+

+ Metadata contains broker and topic metadata for all (matching) topics +

+ 
type Metadata struct {
+    Brokers []BrokerMetadata
+    Topics  map[string]TopicMetadata
+
+    OriginatingBroker BrokerMetadata
+}
+
+

+ type + + OAuthBearerToken + + +

+

+ OAuthBearerToken represents the data to be transmitted +to a broker during SASL/OAUTHBEARER authentication. +

+ 
type OAuthBearerToken struct {
+    // Token value, often (but not necessarily) a JWS compact serialization
+    // as per https://tools.ietf.org/html/rfc7515#section-3.1; it must meet
+    // the regular expression for a SASL/OAUTHBEARER value defined at
+    // https://tools.ietf.org/html/rfc7628#section-3.1
+    TokenValue string
+    // Metadata about the token indicating when it expires (local time);
+    // it must represent a time in the future
+    Expiration time.Time
+    // Metadata about the token indicating the Kafka principal name
+    // to which it applies (for example, "admin")
+    Principal string
+    // SASL extensions, if any, to be communicated to the broker during
+    // authentication (all keys and values of which must meet the regular
+    // expressions defined at https://tools.ietf.org/html/rfc7628#section-3.1,
+    // and it must not contain the reserved "auth" key)
+    Extensions map[string]string
+}
+
+

+ type + + OAuthBearerTokenRefresh + + +

+

+ OAuthBearerTokenRefresh indicates token refresh is required +

+ 
type OAuthBearerTokenRefresh struct {
+    // Config is the value of the sasl.oauthbearer.config property
+    Config string
+}
+
+

+ func (OAuthBearerTokenRefresh) + + String + + +

+ 
func (o OAuthBearerTokenRefresh) String() string
+

+ type + + Offset + + +

+

+ Offset type (int64) with support for canonical names +

+ 
type Offset int64
+

+ func + + NewOffset + + +

+ 
func NewOffset(offset interface{}) (Offset, error)
+

+ NewOffset creates a new Offset using the provided logical string, or an +absolute int64 offset value. +Logical offsets: "beginning", "earliest", "end", "latest", "unset", "invalid", "stored" +

+

+ func + + OffsetTail + + +

+ 
func OffsetTail(relativeOffset Offset) Offset
+

+ OffsetTail returns the logical offset relativeOffset from current end of partition +

+

+ func (*Offset) + + Set + + +

+ 
func (o *Offset) Set(offset interface{}) error
+

+ Set offset value, see NewOffset() +

+

+ func (Offset) + + String + + +

+ 
func (o Offset) String() string
+

+ type + + OffsetsCommitted + + +

+

+ OffsetsCommitted reports committed offsets +

+ 
type OffsetsCommitted struct {
+    Error   error
+    Offsets []TopicPartition
+}
+
+

+ func (OffsetsCommitted) + + String + + +

+ 
func (o OffsetsCommitted) String() string
+

+ type + + PartitionEOF + + +

+

+ PartitionEOF consumer reached end of partition +Needs to be explicitly enabled by setting the `enable.partition.eof` +configuration property to true. +

+ 
type PartitionEOF TopicPartition
+

+ func (PartitionEOF) + + String + + +

+ 
func (p PartitionEOF) String() string
+

+ type + + PartitionMetadata + + +

+

+ PartitionMetadata contains per-partition metadata +

+ 
type PartitionMetadata struct {
+    ID       int32
+    Error    Error
+    Leader   int32
+    Replicas []int32
+    Isrs     []int32
+}
+
+

+ type + + PartitionsSpecification + + +

+

+ PartitionsSpecification holds parameters for creating additional partitions for a topic. +PartitionsSpecification is analogous to NewPartitions in the Java Topic Admin API. +

+ 
type PartitionsSpecification struct {
+    // Topic to create more partitions for.
+    Topic string
+    // New partition count for topic, must be higher than current partition count.
+    IncreaseTo int
+    // (Optional) Explicit replica assignment. The outer array is
+    // indexed by the new partition index (i.e., 0 for the first added
+    // partition), while the inner per-partition array
+    // contains the replica broker ids. The first broker in each
+    // broker id list will be the preferred replica.
+    ReplicaAssignment [][]int32
+}
+
+

+ type + + Producer + + +

+

+ Producer implements a High-level Apache Kafka Producer instance +

+ 
type Producer struct {
+    // contains filtered or unexported fields
+}
+
+

+ func + + NewProducer + + +

+ 
func NewProducer(conf *ConfigMap) (*Producer, error)
+

+ NewProducer creates a new high-level Producer instance. +

+

+ conf is a *ConfigMap with standard librdkafka configuration properties. +

+

+ Supported special configuration properties (type, default): +

+ 
go.batch.producer (bool, false) - EXPERIMENTAL: Enable batch producer (for increased performance).
+                                  These batches do not relate to Kafka message batches in any way.
+                                  Note: timestamps and headers are not supported with this interface.
+go.delivery.reports (bool, true) - Forward per-message delivery reports to the
+                                   Events() channel.
+go.delivery.report.fields (string, "key,value") - Comma separated list of fields to enable for delivery reports.
+                                    Allowed values: all, none (or empty string), key, value, headers
+                                    Warning: There is a performance penalty to include headers in the delivery report.
+go.events.channel.size (int, 1000000) - Events().
+go.produce.channel.size (int, 1000000) - ProduceChannel() buffer size (in number of messages)
+go.logs.channel.enable (bool, false) - Forward log to Logs() channel.
+go.logs.channel (chan kafka.LogEvent, nil) - Forward logs to application-provided channel instead of Logs(). Requires go.logs.channel.enable=true.
+
+

+ func (*Producer) + + AbortTransaction + + +

+ 
func (p *Producer) AbortTransaction(ctx context.Context) error
+

+ AbortTransaction aborts the ongoing transaction. +

+

+ This function should also be used to recover from non-fatal abortable +transaction errors. +

+

+ Any outstanding messages will be purged and fail with +`ErrPurgeInflight` or `ErrPurgeQueue`. +

+

+ Parameters: +

+ 
* `ctx` - The maximum amount of time to block, or nil for indefinite.
+
+

+ Note: This function will block until all outstanding messages are purged +and the transaction abort request has been successfully +handled by the transaction coordinator, or until the `ctx` expires, +which ever comes first. On timeout the application may +call the function again. +

+

+ Note: Will automatically call `Purge()` and `Flush()` to ensure all queued +and in-flight messages are purged before attempting to abort the transaction. +The application MUST serve the `producer.Events()` channel for delivery +reports in a separate go-routine during this time. +

+

+ Returns nil on success or an error object on failure. +Check whether the returned error object permits retrying +by calling `err.(kafka.Error).IsRetriable()`, or whether a fatal error +has been raised by calling `err.(kafka.Error).IsFatal()`. +

+

+ func (*Producer) + + BeginTransaction + + +

+ 
func (p *Producer) BeginTransaction() error
+

+ BeginTransaction starts a new transaction. +

+

+ `InitTransactions()` must have been called successfully (once) +before this function is called. +

+

+ Any messages produced, offsets sent (`SendOffsetsToTransaction()`), +etc, after the successful return of this function will be part of +the transaction and committed or aborted atomatically. +

+

+ Finish the transaction by calling `CommitTransaction()` or +abort the transaction by calling `AbortTransaction()`. +

+

+ Returns nil on success or an error object on failure. +Check whether a fatal error has been raised by +calling `err.(kafka.Error).IsFatal()`. +

+

+ Note: With the transactional producer, `Produce()`, et.al, are only +allowed during an on-going transaction, as started with this function. +Any produce call outside an on-going transaction, or for a failed +transaction, will fail. +

+

+ func (*Producer) + + Close + + +

+ 
func (p *Producer) Close()
+

+ Close a Producer instance. +The Producer object or its channels are no longer usable after this call. +

+

+ func (*Producer) + + CommitTransaction + + +

+ 
func (p *Producer) CommitTransaction(ctx context.Context) error
+

+ CommitTransaction commits the current transaction. +

+

+ Any outstanding messages will be flushed (delivered) before actually +committing the transaction. +

+

+ If any of the outstanding messages fail permanently the current +transaction will enter the abortable error state and this +function will return an abortable error, in this case the application +must call `AbortTransaction()` before attempting a new +transaction with `BeginTransaction()`. +

+

+ Parameters: +

+ 
* `ctx` - The maximum amount of time to block, or nil for indefinite.
+
+

+ Note: This function will block until all outstanding messages are +delivered and the transaction commit request has been successfully +handled by the transaction coordinator, or until the `ctx` expires, +which ever comes first. On timeout the application may +call the function again. +

+

+ Note: Will automatically call `Flush()` to ensure all queued +messages are delivered before attempting to commit the transaction. +The application MUST serve the `producer.Events()` channel for delivery +reports in a separate go-routine during this time. +

+

+ Returns nil on success or an error object on failure. +Check whether the returned error object permits retrying +by calling `err.(kafka.Error).IsRetriable()`, or whether an abortable +or fatal error has been raised by calling +`err.(kafka.Error).TxnRequiresAbort()` or `err.(kafka.Error).IsFatal()` +respectively. +

+

+ func (*Producer) + + Events + + +

+ 
func (p *Producer) Events() chan Event
+

+ Events returns the Events channel (read) +

+

+ func (*Producer) + + Flush + + +

+ 
func (p *Producer) Flush(timeoutMs int) int
+

+ Flush and wait for outstanding messages and requests to complete delivery. +Includes messages on ProduceChannel. +Runs until value reaches zero or on timeoutMs. +Returns the number of outstanding events still un-flushed. +

+

+ func (*Producer) + + GetFatalError + + +

+ 
func (p *Producer) GetFatalError() error
+

+ GetFatalError returns an Error object if the client instance has raised a fatal error, else nil. +

+

+ func (*Producer) + + GetMetadata + + +

+ 
func (p *Producer) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error)
+

+ GetMetadata queries broker for cluster and topic metadata. +If topic is non-nil only information about that topic is returned, else if +allTopics is false only information about locally used topics is returned, +else information about all topics is returned. +GetMetadata is equivalent to listTopics, describeTopics and describeCluster in the Java API. +

+

+ func (*Producer) + + InitTransactions + + +

+ 
func (p *Producer) InitTransactions(ctx context.Context) error
+

+ InitTransactions Initializes transactions for the producer instance. +

+

+ This function ensures any transactions initiated by previous instances +of the producer with the same `transactional.id` are completed. +If the previous instance failed with a transaction in progress the +previous transaction will be aborted. +This function needs to be called before any other transactional or +produce functions are called when the `transactional.id` is configured. +

+

+ If the last transaction had begun completion (following transaction commit) +but not yet finished, this function will await the previous transaction's +completion. +

+

+ When any previous transactions have been fenced this function +will acquire the internal producer id and epoch, used in all future +transactional messages issued by this producer instance. +

+

+ Upon successful return from this function the application has to perform at +least one of the following operations within `transaction.timeout.ms` to +avoid timing out the transaction on the broker: +

+ 
* `Produce()` (et.al)
+* `SendOffsetsToTransaction()`
+* `CommitTransaction()`
+* `AbortTransaction()`
+
+

+ Parameters: +

+ 
* `ctx` - The maximum time to block, or nil for indefinite.
+          On timeout the operation may continue in the background,
+          depending on state, and it is okay to call `InitTransactions()`
+          again.
+
+

+ Returns nil on success or an error on failure. +Check whether the returned error object permits retrying +by calling `err.(kafka.Error).IsRetriable()`, or whether a fatal +error has been raised by calling `err.(kafka.Error).IsFatal()`. +

+

+ func (*Producer) + + Len + + +

+ 
func (p *Producer) Len() int
+

+ Len returns the number of messages and requests waiting to be transmitted to the broker +as well as delivery reports queued for the application. +Includes messages on ProduceChannel. +

+

+ func (*Producer) + + Logs + + +

+ 
func (p *Producer) Logs() chan LogEvent
+

+ Logs returns the Log channel (if enabled), else nil +

+

+ func (*Producer) + + OffsetsForTimes + + +

+ 
func (p *Producer) OffsetsForTimes(times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error)
+

+ OffsetsForTimes looks up offsets by timestamp for the given partitions. +

+

+ The returned offset for each partition is the earliest offset whose +timestamp is greater than or equal to the given timestamp in the +corresponding partition. If the provided timestamp exceeds that of the +last message in the partition, a value of -1 will be returned. +

+

+ The timestamps to query are represented as `.Offset` in the `times` +argument and the looked up offsets are represented as `.Offset` in the returned +`offsets` list. +

+

+ The function will block for at most timeoutMs milliseconds. +

+

+ Duplicate Topic+Partitions are not supported. +Per-partition errors may be returned in the `.Error` field. +

+

+ func (*Producer) + + Produce + + +

+ 
func (p *Producer) Produce(msg *Message, deliveryChan chan Event) error
+

+ Produce single message. +This is an asynchronous call that enqueues the message on the internal +transmit queue, thus returning immediately. +The delivery report will be sent on the provided deliveryChan if specified, +or on the Producer object's Events() channel if not. +msg.Timestamp requires librdkafka >= 0.9.4 (else returns ErrNotImplemented), +api.version.request=true, and broker >= 0.10.0.0. +msg.Headers requires librdkafka >= 0.11.4 (else returns ErrNotImplemented), +api.version.request=true, and broker >= 0.11.0.0. +Returns an error if message could not be enqueued. +

+

+ func (*Producer) + + ProduceChannel + + +

+ 
func (p *Producer) ProduceChannel() chan *Message
+

+ ProduceChannel returns the produce *Message channel (write) +

+

+ func (*Producer) + + Purge + + +

+ 
func (p *Producer) Purge(flags int) error
+

+ Purge messages currently handled by this producer instance. +

+

+ flags is a combination of PurgeQueue, PurgeInFlight and PurgeNonBlocking. +

+

+ The application will need to call Poll(), Flush() or read the Events() channel +after this call to serve delivery reports for the purged messages. +

+

+ Messages purged from internal queues fail with the delivery report +error code set to ErrPurgeQueue, while purged messages that +are in-flight to or from the broker will fail with the error code set to +ErrPurgeInflight. +

+

+ Warning: Purging messages that are in-flight to or from the broker +will ignore any sub-sequent acknowledgement for these messages +received from the broker, effectively making it impossible +for the application to know if the messages were successfully +produced or not. This may result in duplicate messages if the +application retries these messages at a later time. +

+

+ Note: This call may block for a short time while background thread +queues are purged. +

+

+ Returns nil on success, ErrInvalidArg if the purge flags are invalid or unknown. +

+

+ func (*Producer) + + QueryWatermarkOffsets + + +

+ 
func (p *Producer) QueryWatermarkOffsets(topic string, partition int32, timeoutMs int) (low, high int64, err error)
+

+ QueryWatermarkOffsets returns the broker's low and high offsets for the given topic +and partition. +

+

+ func (*Producer) + + SendOffsetsToTransaction + + +

+ 
func (p *Producer) SendOffsetsToTransaction(ctx context.Context, offsets []TopicPartition, consumerMetadata *ConsumerGroupMetadata) error
+

+ SendOffsetsToTransaction sends a list of topic partition offsets to the +consumer group coordinator for `consumerMetadata`, and marks the offsets +as part part of the current transaction. +These offsets will be considered committed only if the transaction is +committed successfully. +

+

+ The offsets should be the next message your application will consume, +i.e., the last processed message's offset + 1 for each partition. +Either track the offsets manually during processing or use +`consumer.Position()` (on the consumer) to get the current offsets for +the partitions assigned to the consumer. +

+

+ Use this method at the end of a consume-transform-produce loop prior +to committing the transaction with `CommitTransaction()`. +

+

+ Parameters: +

+ 
* `ctx` - The maximum amount of time to block, or nil for indefinite.
+* `offsets` - List of offsets to commit to the consumer group upon
+              successful commit of the transaction. Offsets should be
+              the next message to consume, e.g., last processed message + 1.
+* `consumerMetadata` - The current consumer group metadata as returned by
+              `consumer.GetConsumerGroupMetadata()` on the consumer
+              instance the provided offsets were consumed from.
+
+

+ Note: The consumer must disable auto commits (set `enable.auto.commit` to false on the consumer). +

+

+ Note: Logical and invalid offsets (e.g., OffsetInvalid) in +`offsets` will be ignored. If there are no valid offsets in +`offsets` the function will return nil and no action will be taken. +

+

+ Returns nil on success or an error object on failure. +Check whether the returned error object permits retrying +by calling `err.(kafka.Error).IsRetriable()`, or whether an abortable +or fatal error has been raised by calling +`err.(kafka.Error).TxnRequiresAbort()` or `err.(kafka.Error).IsFatal()` +respectively. +

+

+ func (*Producer) + + SetOAuthBearerToken + + +

+ 
func (p *Producer) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error
+

+ SetOAuthBearerToken sets the the data to be transmitted +to a broker during SASL/OAUTHBEARER authentication. It will return nil +on success, otherwise an error if: +1) the token data is invalid (meaning an expiration time in the past +or either a token value or an extension key or value that does not meet +the regular expression requirements as per + + https://tools.ietf.org/html/rfc7628#section-3.1 + + ); +2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +3) SASL/OAUTHBEARER is supported but is not configured as the client's +authentication mechanism. +

+

+ func (*Producer) + + SetOAuthBearerTokenFailure + + +

+ 
func (p *Producer) SetOAuthBearerTokenFailure(errstr string) error
+

+ SetOAuthBearerTokenFailure sets the error message describing why token +retrieval/setting failed; it also schedules a new token refresh event for 10 +seconds later so the attempt may be retried. It will return nil on +success, otherwise an error if: +1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +2) SASL/OAUTHBEARER is supported but is not configured as the client's +authentication mechanism. +

+

+ func (*Producer) + + String + + +

+ 
func (p *Producer) String() string
+

+ String returns a human readable name for a Producer instance +

+

+ func (*Producer) + + TestFatalError + + +

+ 
func (p *Producer) TestFatalError(code ErrorCode, str string) ErrorCode
+

+ TestFatalError triggers a fatal error in the underlying client. +This is to be used strictly for testing purposes. +

+

+ type + + RebalanceCb + + +

+

+ RebalanceCb provides a per-Subscribe*() rebalance event callback. +The passed Event will be either AssignedPartitions or RevokedPartitions +

+ 
type RebalanceCb func(*Consumer, Event) error
+

+ type + + ResourceType + + +

+

+ ResourceType represents an Apache Kafka resource type +

+ 
type ResourceType int
+

+ func + + ResourceTypeFromString + + +

+ 
func ResourceTypeFromString(typeString string) (ResourceType, error)
+

+ ResourceTypeFromString translates a resource type name/string to +a ResourceType value. +

+

+ func (ResourceType) + + String + + +

+ 
func (t ResourceType) String() string
+

+ String returns the human-readable representation of a ResourceType +

+

+ type + + RevokedPartitions + + +

+

+ RevokedPartitions consumer group rebalance event: revoked partition set +

+ 
type RevokedPartitions struct {
+    Partitions []TopicPartition
+}
+
+

+ func (RevokedPartitions) + + String + + +

+ 
func (e RevokedPartitions) String() string
+

+ type + + Stats + + +

+

+ Stats statistics event +

+ 
type Stats struct {
+    // contains filtered or unexported fields
+}
+
+

+ func (Stats) + + String + + +

+ 
func (e Stats) String() string
+

+ type + + TimestampType + + +

+

+ TimestampType is a the Message timestamp type or source +

+ 
type TimestampType int
+

+ func (TimestampType) + + String + + +

+ 
func (t TimestampType) String() string
+

+ type + + TopicMetadata + + +

+

+ TopicMetadata contains per-topic metadata +

+ 
type TopicMetadata struct {
+    Topic      string
+    Partitions []PartitionMetadata
+    Error      Error
+}
+
+

+ type + + TopicPartition + + +

+

+ TopicPartition is a generic placeholder for a Topic+Partition and optionally Offset. +

+ 
type TopicPartition struct {
+    Topic     *string
+    Partition int32
+    Offset    Offset
+    Metadata  *string
+    Error     error
+}
+
+

+ func (TopicPartition) + + String + + +

+ 
func (p TopicPartition) String() string
+

+ type + + TopicPartitions + + +

+

+ TopicPartitions is a slice of TopicPartitions that also implements +the sort interface +

+ 
type TopicPartitions []TopicPartition
+

+ func (TopicPartitions) + + Len + + +

+ 
func (tps TopicPartitions) Len() int
+

+ func (TopicPartitions) + + Less + + +

+ 
func (tps TopicPartitions) Less(i, j int) bool
+

+ func (TopicPartitions) + + Swap + + +

+ 
func (tps TopicPartitions) Swap(i, j int)
+

+ type + + TopicResult + + +

+

+ TopicResult provides per-topic operation result (error) information. +

+ 
type TopicResult struct {
+    // Topic name
+    Topic string
+    // Error, if any, of result. Check with `Error.Code() != ErrNoError`.
+    Error Error
+}
+
+

+ func (TopicResult) + + String + + +

+ 
func (t TopicResult) String() string
+

+ String returns a human-readable representation of a TopicResult. +

+

+ type + + TopicSpecification + + +

+

+ TopicSpecification holds parameters for creating a new topic. +TopicSpecification is analogous to NewTopic in the Java Topic Admin API. +

+ 
type TopicSpecification struct {
+    // Topic name to create.
+    Topic string
+    // Number of partitions in topic.
+    NumPartitions int
+    // Default replication factor for the topic's partitions, or zero
+    // if an explicit ReplicaAssignment is set.
+    ReplicationFactor int
+    // (Optional) Explicit replica assignment. The outer array is
+    // indexed by the partition number, while the inner per-partition array
+    // contains the replica broker ids. The first broker in each
+    // broker id list will be the preferred replica.
+    ReplicaAssignment [][]int32
+    // Topic configuration.
+    Config map[string]string
+}
+
+ +
+ +
+ + + + diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_darwin.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_darwin.go new file mode 100644 index 0000000..5b6812f --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_darwin.go @@ -0,0 +1,13 @@ +// +build !dynamic + + +// This file was auto-generated by librdkafka_vendor/bundle-import.sh, DO NOT EDIT. + +package kafka + +// #cgo CFLAGS: -DUSE_VENDORED_LIBRDKAFKA -DLIBRDKAFKA_STATICLIB +// #cgo LDFLAGS: ${SRCDIR}/librdkafka_vendor/librdkafka_darwin.a -lm -lsasl2 -ldl -lpthread +import "C" + +// LibrdkafkaLinkInfo explains how librdkafka was linked to the Go client +const LibrdkafkaLinkInfo = "static darwin from librdkafka-static-bundle-v1.8.2.tgz" diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_dynamic.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_dynamic.go new file mode 100644 index 0000000..bb7da9b --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_dynamic.go @@ -0,0 +1,9 @@ +// +build dynamic + +package kafka + +// #cgo pkg-config: rdkafka +import "C" + +// LibrdkafkaLinkInfo explains how librdkafka was linked to the Go client +const LibrdkafkaLinkInfo = "dynamically linked to librdkafka" diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_glibc_linux.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_glibc_linux.go new file mode 100644 index 0000000..a69638d --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_glibc_linux.go @@ -0,0 +1,13 @@ +// +build !dynamic +// +build !musl + +// This file was auto-generated by librdkafka_vendor/bundle-import.sh, DO NOT EDIT. + +package kafka + +// #cgo CFLAGS: -DUSE_VENDORED_LIBRDKAFKA -DLIBRDKAFKA_STATICLIB +// #cgo LDFLAGS: ${SRCDIR}/librdkafka_vendor/librdkafka_glibc_linux.a -lm -ldl -lpthread -lrt +import "C" + +// LibrdkafkaLinkInfo explains how librdkafka was linked to the Go client +const LibrdkafkaLinkInfo = "static glibc_linux from librdkafka-static-bundle-v1.8.2.tgz" diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_musl_linux.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_musl_linux.go new file mode 100644 index 0000000..ed2bee5 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_musl_linux.go @@ -0,0 +1,13 @@ +// +build !dynamic +// +build musl + +// This file was auto-generated by librdkafka_vendor/bundle-import.sh, DO NOT EDIT. + +package kafka + +// #cgo CFLAGS: -DUSE_VENDORED_LIBRDKAFKA -DLIBRDKAFKA_STATICLIB +// #cgo LDFLAGS: ${SRCDIR}/librdkafka_vendor/librdkafka_musl_linux.a -lm -ldl -lpthread -lrt -lpthread -lrt +import "C" + +// LibrdkafkaLinkInfo explains how librdkafka was linked to the Go client +const LibrdkafkaLinkInfo = "static musl_linux from librdkafka-static-bundle-v1.8.2.tgz" diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_windows.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_windows.go new file mode 100644 index 0000000..1c15c55 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/build_windows.go @@ -0,0 +1,13 @@ +// +build !dynamic + + +// This file was auto-generated by librdkafka_vendor/bundle-import.sh, DO NOT EDIT. + +package kafka + +// #cgo CFLAGS: -DUSE_VENDORED_LIBRDKAFKA -DLIBRDKAFKA_STATICLIB +// #cgo LDFLAGS: ${SRCDIR}/librdkafka_vendor/librdkafka_windows.a -lws2_32 -lsecur32 -lcrypt32 +import "C" + +// LibrdkafkaLinkInfo explains how librdkafka was linked to the Go client +const LibrdkafkaLinkInfo = "static windows from librdkafka-static-bundle-v1.8.2.tgz" diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/config.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/config.go new file mode 100644 index 0000000..f0c8ecf --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/config.go @@ -0,0 +1,280 @@ +package kafka + +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "fmt" + "reflect" + "strings" + "unsafe" +) + +/* +#include +#include "select_rdkafka.h" +*/ +import "C" + +// ConfigValue supports the following types: +// bool, int, string, any type with the standard String() interface +type ConfigValue interface{} + +// ConfigMap is a map containing standard librdkafka configuration properties as documented in: +// https://github.com/edenhill/librdkafka/tree/master/CONFIGURATION.md +// +// The special property "default.topic.config" (optional) is a ConfigMap +// containing default topic configuration properties. +// +// The use of "default.topic.config" is deprecated, +// topic configuration properties shall be specified in the standard ConfigMap. +// For backwards compatibility, "default.topic.config" (if supplied) +// takes precedence. +type ConfigMap map[string]ConfigValue + +// SetKey sets configuration property key to value. +// +// For user convenience a key prefixed with {topic}. will be +// set on the "default.topic.config" sub-map, this use is deprecated. +func (m ConfigMap) SetKey(key string, value ConfigValue) error { + if strings.HasPrefix(key, "{topic}.") { + _, found := m["default.topic.config"] + if !found { + m["default.topic.config"] = ConfigMap{} + } + m["default.topic.config"].(ConfigMap)[strings.TrimPrefix(key, "{topic}.")] = value + } else { + m[key] = value + } + + return nil +} + +// Set implements flag.Set (command line argument parser) as a convenience +// for `-X key=value` config. +func (m ConfigMap) Set(kv string) error { + i := strings.Index(kv, "=") + if i == -1 { + return newErrorFromString(ErrInvalidArg, "Expected key=value") + } + + k := kv[:i] + v := kv[i+1:] + + return m.SetKey(k, v) +} + +func value2string(v ConfigValue) (ret string, errstr string) { + + switch x := v.(type) { + case bool: + if x { + ret = "true" + } else { + ret = "false" + } + case int: + ret = fmt.Sprintf("%d", x) + case string: + ret = x + case fmt.Stringer: + ret = x.String() + default: + return "", fmt.Sprintf("Invalid value type %T", v) + } + + return ret, "" +} + +// rdkAnyconf abstracts rd_kafka_conf_t and rd_kafka_topic_conf_t +// into a common interface. +type rdkAnyconf interface { + set(cKey *C.char, cVal *C.char, cErrstr *C.char, errstrSize int) C.rd_kafka_conf_res_t +} + +func anyconfSet(anyconf rdkAnyconf, key string, val ConfigValue) (err error) { + value, errstr := value2string(val) + if errstr != "" { + return newErrorFromString(ErrInvalidArg, fmt.Sprintf("%s for key %s (expected string,bool,int,ConfigMap)", errstr, key)) + } + cKey := C.CString(key) + defer C.free(unsafe.Pointer(cKey)) + cVal := C.CString(value) + defer C.free(unsafe.Pointer(cVal)) + cErrstr := (*C.char)(C.malloc(C.size_t(128))) + defer C.free(unsafe.Pointer(cErrstr)) + + if anyconf.set(cKey, cVal, cErrstr, 128) != C.RD_KAFKA_CONF_OK { + return newErrorFromCString(C.RD_KAFKA_RESP_ERR__INVALID_ARG, cErrstr) + } + + return nil +} + +// we need these typedefs to workaround a crash in golint +// when parsing the set() methods below +type rdkConf C.rd_kafka_conf_t +type rdkTopicConf C.rd_kafka_topic_conf_t + +func (cConf *rdkConf) set(cKey *C.char, cVal *C.char, cErrstr *C.char, errstrSize int) C.rd_kafka_conf_res_t { + return C.rd_kafka_conf_set((*C.rd_kafka_conf_t)(cConf), cKey, cVal, cErrstr, C.size_t(errstrSize)) +} + +func (ctopicConf *rdkTopicConf) set(cKey *C.char, cVal *C.char, cErrstr *C.char, errstrSize int) C.rd_kafka_conf_res_t { + return C.rd_kafka_topic_conf_set((*C.rd_kafka_topic_conf_t)(ctopicConf), cKey, cVal, cErrstr, C.size_t(errstrSize)) +} + +func configConvertAnyconf(m ConfigMap, anyconf rdkAnyconf) (err error) { + // set plugins first, any plugin-specific configuration depends on + // the plugin to have already been set + pluginPaths, ok := m["plugin.library.paths"] + if ok { + err = anyconfSet(anyconf, "plugin.library.paths", pluginPaths) + if err != nil { + return err + } + } + for k, v := range m { + if k == "plugin.library.paths" { + continue + } + switch v.(type) { + case ConfigMap: + /* Special sub-ConfigMap, only used for default.topic.config */ + + if k != "default.topic.config" { + return newErrorFromString(ErrInvalidArg, fmt.Sprintf("Invalid type for key %s", k)) + } + + var cTopicConf = C.rd_kafka_topic_conf_new() + + err = configConvertAnyconf(v.(ConfigMap), + (*rdkTopicConf)(cTopicConf)) + if err != nil { + C.rd_kafka_topic_conf_destroy(cTopicConf) + return err + } + + C.rd_kafka_conf_set_default_topic_conf( + (*C.rd_kafka_conf_t)(anyconf.(*rdkConf)), + (*C.rd_kafka_topic_conf_t)((*rdkTopicConf)(cTopicConf))) + + default: + err = anyconfSet(anyconf, k, v) + if err != nil { + return err + } + } + } + + return nil +} + +// convert ConfigMap to C rd_kafka_conf_t * +func (m ConfigMap) convert() (cConf *C.rd_kafka_conf_t, err error) { + cConf = C.rd_kafka_conf_new() + + // Set the client.software.name and .version (use librdkafka version). + _, librdkafkaVersion := LibraryVersion() + anyconfSet((*rdkConf)(cConf), "client.software.name", "confluent-kafka-go") + anyconfSet((*rdkConf)(cConf), "client.software.version", librdkafkaVersion) + + err = configConvertAnyconf(m, (*rdkConf)(cConf)) + if err != nil { + C.rd_kafka_conf_destroy(cConf) + return nil, err + } + return cConf, nil +} + +// get finds key in the configmap and returns its value. +// If the key is not found defval is returned. +// If the key is found but the type is mismatched an error is returned. +func (m ConfigMap) get(key string, defval ConfigValue) (ConfigValue, error) { + if strings.HasPrefix(key, "{topic}.") { + defconfCv, found := m["default.topic.config"] + if !found { + return defval, nil + } + return defconfCv.(ConfigMap).get(strings.TrimPrefix(key, "{topic}."), defval) + } + + v, ok := m[key] + if !ok { + return defval, nil + } + + if defval != nil && reflect.TypeOf(defval) != reflect.TypeOf(v) { + return nil, newErrorFromString(ErrInvalidArg, fmt.Sprintf("%s expects type %T, not %T", key, defval, v)) + } + + return v, nil +} + +// extract performs a get() and if found deletes the key. +func (m ConfigMap) extract(key string, defval ConfigValue) (ConfigValue, error) { + + v, err := m.get(key, defval) + if err != nil { + return nil, err + } + + delete(m, key) + + return v, nil +} + +// extractLogConfig extracts generic go.logs.* configuration properties. +func (m ConfigMap) extractLogConfig() (logsChanEnable bool, logsChan chan LogEvent, err error) { + v, err := m.extract("go.logs.channel.enable", false) + if err != nil { + return + } + + logsChanEnable = v.(bool) + + v, err = m.extract("go.logs.channel", nil) + if err != nil { + return + } + + if v != nil { + logsChan = v.(chan LogEvent) + } + + if logsChanEnable { + // Tell librdkafka to forward logs to the log queue + m.Set("log.queue=true") + } + + return +} + +func (m ConfigMap) clone() ConfigMap { + m2 := make(ConfigMap) + for k, v := range m { + m2[k] = v + } + return m2 +} + +// Get finds the given key in the ConfigMap and returns its value. +// If the key is not found `defval` is returned. +// If the key is found but the type does not match that of `defval` (unless nil) +// an ErrInvalidArg error is returned. +func (m ConfigMap) Get(key string, defval ConfigValue) (ConfigValue, error) { + return m.get(key, defval) +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/consumer.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/consumer.go new file mode 100644 index 0000000..ef48330 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/consumer.go @@ -0,0 +1,923 @@ +package kafka + +/** + * Copyright 2016-2020 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "fmt" + "math" + "time" + "unsafe" +) + +/* +#include +#include "select_rdkafka.h" + + +static rd_kafka_topic_partition_t *_c_rdkafka_topic_partition_list_entry(rd_kafka_topic_partition_list_t *rktparlist, int idx) { + return idx < rktparlist->cnt ? &rktparlist->elems[idx] : NULL; +} +*/ +import "C" + +// RebalanceCb provides a per-Subscribe*() rebalance event callback. +// The passed Event will be either AssignedPartitions or RevokedPartitions +type RebalanceCb func(*Consumer, Event) error + +// Consumer implements a High-level Apache Kafka Consumer instance +type Consumer struct { + events chan Event + handle handle + eventsChanEnable bool + readerTermChan chan bool + rebalanceCb RebalanceCb + appReassigned bool + appRebalanceEnable bool // Config setting +} + +// Strings returns a human readable name for a Consumer instance +func (c *Consumer) String() string { + return c.handle.String() +} + +// getHandle implements the Handle interface +func (c *Consumer) gethandle() *handle { + return &c.handle +} + +// Subscribe to a single topic +// This replaces the current subscription +func (c *Consumer) Subscribe(topic string, rebalanceCb RebalanceCb) error { + return c.SubscribeTopics([]string{topic}, rebalanceCb) +} + +// SubscribeTopics subscribes to the provided list of topics. +// This replaces the current subscription. +func (c *Consumer) SubscribeTopics(topics []string, rebalanceCb RebalanceCb) (err error) { + ctopics := C.rd_kafka_topic_partition_list_new(C.int(len(topics))) + defer C.rd_kafka_topic_partition_list_destroy(ctopics) + + for _, topic := range topics { + ctopic := C.CString(topic) + defer C.free(unsafe.Pointer(ctopic)) + C.rd_kafka_topic_partition_list_add(ctopics, ctopic, C.RD_KAFKA_PARTITION_UA) + } + + e := C.rd_kafka_subscribe(c.handle.rk, ctopics) + if e != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(e) + } + + c.rebalanceCb = rebalanceCb + + return nil +} + +// Unsubscribe from the current subscription, if any. +func (c *Consumer) Unsubscribe() (err error) { + C.rd_kafka_unsubscribe(c.handle.rk) + return nil +} + +// Assign an atomic set of partitions to consume. +// +// The .Offset field of each TopicPartition must either be set to an absolute +// starting offset (>= 0), or one of the logical offsets (`kafka.OffsetEnd` etc), +// but should typically be set to `kafka.OffsetStored` to have the consumer +// use the committed offset as a start position, with a fallback to +// `auto.offset.reset` if there is no committed offset. +// +// This replaces the current assignment. +func (c *Consumer) Assign(partitions []TopicPartition) (err error) { + c.appReassigned = true + + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + + e := C.rd_kafka_assign(c.handle.rk, cparts) + if e != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(e) + } + + return nil +} + +// Unassign the current set of partitions to consume. +func (c *Consumer) Unassign() (err error) { + c.appReassigned = true + + e := C.rd_kafka_assign(c.handle.rk, nil) + if e != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(e) + } + + return nil +} + +// IncrementalAssign adds the specified partitions to the current set of +// partitions to consume. +// +// The .Offset field of each TopicPartition must either be set to an absolute +// starting offset (>= 0), or one of the logical offsets (`kafka.OffsetEnd` etc), +// but should typically be set to `kafka.OffsetStored` to have the consumer +// use the committed offset as a start position, with a fallback to +// `auto.offset.reset` if there is no committed offset. +// +// The new partitions must not be part of the current assignment. +func (c *Consumer) IncrementalAssign(partitions []TopicPartition) (err error) { + c.appReassigned = true + + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + + cError := C.rd_kafka_incremental_assign(c.handle.rk, cparts) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} + +// IncrementalUnassign removes the specified partitions from the current set of +// partitions to consume. +// +// The .Offset field of the TopicPartition is ignored. +// +// The removed partitions must be part of the current assignment. +func (c *Consumer) IncrementalUnassign(partitions []TopicPartition) (err error) { + c.appReassigned = true + + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + + cError := C.rd_kafka_incremental_unassign(c.handle.rk, cparts) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} + +// GetRebalanceProtocol returns the current consumer group rebalance protocol, +// which is either "EAGER" or "COOPERATIVE". +// If the rebalance protocol is not known in the current state an empty string +// is returned. +// Should typically only be called during rebalancing. +func (c *Consumer) GetRebalanceProtocol() string { + cStr := C.rd_kafka_rebalance_protocol(c.handle.rk) + if cStr == nil { + return "" + } + + return C.GoString(cStr) +} + +// AssignmentLost returns true if current partition assignment has been lost. +// This method is only applicable for use with a subscribing consumer when +// handling a rebalance event or callback. +// Partitions that have been lost may already be owned by other members in the +// group and therefore commiting offsets, for example, may fail. +func (c *Consumer) AssignmentLost() bool { + return cint2bool(C.rd_kafka_assignment_lost(c.handle.rk)) +} + +// commit offsets for specified offsets. +// If offsets is nil the currently assigned partitions' offsets are committed. +// This is a blocking call, caller will need to wrap in go-routine to +// get async or throw-away behaviour. +func (c *Consumer) commit(offsets []TopicPartition) (committedOffsets []TopicPartition, err error) { + var rkqu *C.rd_kafka_queue_t + + rkqu = C.rd_kafka_queue_new(c.handle.rk) + defer C.rd_kafka_queue_destroy(rkqu) + + var coffsets *C.rd_kafka_topic_partition_list_t + if offsets != nil { + coffsets = newCPartsFromTopicPartitions(offsets) + defer C.rd_kafka_topic_partition_list_destroy(coffsets) + } + + cErr := C.rd_kafka_commit_queue(c.handle.rk, coffsets, rkqu, nil, nil) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cErr) + } + + rkev := C.rd_kafka_queue_poll(rkqu, C.int(-1)) + if rkev == nil { + // shouldn't happen + return nil, newError(C.RD_KAFKA_RESP_ERR__DESTROY) + } + defer C.rd_kafka_event_destroy(rkev) + + if C.rd_kafka_event_type(rkev) != C.RD_KAFKA_EVENT_OFFSET_COMMIT { + panic(fmt.Sprintf("Expected OFFSET_COMMIT, got %s", + C.GoString(C.rd_kafka_event_name(rkev)))) + } + + cErr = C.rd_kafka_event_error(rkev) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newErrorFromCString(cErr, C.rd_kafka_event_error_string(rkev)) + } + + cRetoffsets := C.rd_kafka_event_topic_partition_list(rkev) + if cRetoffsets == nil { + // no offsets, no error + return nil, nil + } + committedOffsets = newTopicPartitionsFromCparts(cRetoffsets) + + return committedOffsets, nil +} + +// Commit offsets for currently assigned partitions +// This is a blocking call. +// Returns the committed offsets on success. +func (c *Consumer) Commit() ([]TopicPartition, error) { + return c.commit(nil) +} + +// CommitMessage commits offset based on the provided message. +// This is a blocking call. +// Returns the committed offsets on success. +func (c *Consumer) CommitMessage(m *Message) ([]TopicPartition, error) { + if m.TopicPartition.Error != nil { + return nil, newErrorFromString(ErrInvalidArg, "Can't commit errored message") + } + offsets := []TopicPartition{m.TopicPartition} + offsets[0].Offset++ + return c.commit(offsets) +} + +// CommitOffsets commits the provided list of offsets +// This is a blocking call. +// Returns the committed offsets on success. +func (c *Consumer) CommitOffsets(offsets []TopicPartition) ([]TopicPartition, error) { + return c.commit(offsets) +} + +// StoreOffsets stores the provided list of offsets that will be committed +// to the offset store according to `auto.commit.interval.ms` or manual +// offset-less Commit(). +// +// Returns the stored offsets on success. If at least one offset couldn't be stored, +// an error and a list of offsets is returned. Each offset can be checked for +// specific errors via its `.Error` member. +func (c *Consumer) StoreOffsets(offsets []TopicPartition) (storedOffsets []TopicPartition, err error) { + coffsets := newCPartsFromTopicPartitions(offsets) + defer C.rd_kafka_topic_partition_list_destroy(coffsets) + + cErr := C.rd_kafka_offsets_store(c.handle.rk, coffsets) + + // coffsets might be annotated with an error + storedOffsets = newTopicPartitionsFromCparts(coffsets) + + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return storedOffsets, newError(cErr) + } + + return storedOffsets, nil +} + +// StoreMessage stores offset based on the provided message. +// This is a convenience method that uses StoreOffsets to do the actual work. +func (c *Consumer) StoreMessage(m *Message) (storedOffsets []TopicPartition, err error) { + if m.TopicPartition.Error != nil { + return nil, newErrorFromString(ErrInvalidArg, "Can't store errored message") + } + if m.TopicPartition.Offset < 0 { + return nil, newErrorFromString(ErrInvalidArg, "Can't store message with offset less than 0") + } + offsets := []TopicPartition{m.TopicPartition} + offsets[0].Offset++ + return c.StoreOffsets(offsets) +} + +// Seek seeks the given topic partitions using the offset from the TopicPartition. +// +// If timeoutMs is not 0 the call will wait this long for the +// seek to be performed. If the timeout is reached the internal state +// will be unknown and this function returns ErrTimedOut. +// If timeoutMs is 0 it will initiate the seek but return +// immediately without any error reporting (e.g., async). +// +// Seek() may only be used for partitions already being consumed +// (through Assign() or implicitly through a self-rebalanced Subscribe()). +// To set the starting offset it is preferred to use Assign() and provide +// a starting offset for each partition. +// +// Returns an error on failure or nil otherwise. +func (c *Consumer) Seek(partition TopicPartition, timeoutMs int) error { + rkt := c.handle.getRkt(*partition.Topic) + cErr := C.rd_kafka_seek(rkt, + C.int32_t(partition.Partition), + C.int64_t(partition.Offset), + C.int(timeoutMs)) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(cErr) + } + return nil +} + +// Poll the consumer for messages or events. +// +// Will block for at most timeoutMs milliseconds +// +// The following callbacks may be triggered: +// Subscribe()'s rebalanceCb +// +// Returns nil on timeout, else an Event +func (c *Consumer) Poll(timeoutMs int) (event Event) { + ev, _ := c.handle.eventPoll(nil, timeoutMs, 1, nil) + return ev +} + +// Events returns the Events channel (if enabled) +func (c *Consumer) Events() chan Event { + return c.events +} + +// Logs returns the log channel if enabled, or nil otherwise. +func (c *Consumer) Logs() chan LogEvent { + return c.handle.logs +} + +// ReadMessage polls the consumer for a message. +// +// This is a convenience API that wraps Poll() and only returns +// messages or errors. All other event types are discarded. +// +// The call will block for at most `timeout` waiting for +// a new message or error. `timeout` may be set to -1 for +// indefinite wait. +// +// Timeout is returned as (nil, err) where err is `err.(kafka.Error).Code() == kafka.ErrTimedOut`. +// +// Messages are returned as (msg, nil), +// while general errors are returned as (nil, err), +// and partition-specific errors are returned as (msg, err) where +// msg.TopicPartition provides partition-specific information (such as topic, partition and offset). +// +// All other event types, such as PartitionEOF, AssignedPartitions, etc, are silently discarded. +// +func (c *Consumer) ReadMessage(timeout time.Duration) (*Message, error) { + + var absTimeout time.Time + var timeoutMs int + + if timeout > 0 { + absTimeout = time.Now().Add(timeout) + timeoutMs = (int)(timeout.Seconds() * 1000.0) + } else { + timeoutMs = (int)(timeout) + } + + for { + ev := c.Poll(timeoutMs) + + switch e := ev.(type) { + case *Message: + if e.TopicPartition.Error != nil { + return e, e.TopicPartition.Error + } + return e, nil + case Error: + return nil, e + default: + // Ignore other event types + } + + if timeout > 0 { + // Calculate remaining time + timeoutMs = int(math.Max(0.0, absTimeout.Sub(time.Now()).Seconds()*1000.0)) + } + + if timeoutMs == 0 && ev == nil { + return nil, newError(C.RD_KAFKA_RESP_ERR__TIMED_OUT) + } + + } + +} + +// Close Consumer instance. +// The object is no longer usable after this call. +func (c *Consumer) Close() (err error) { + + // Wait for consumerReader() or pollLogEvents to terminate (by closing readerTermChan) + close(c.readerTermChan) + c.handle.waitGroup.Wait() + if c.eventsChanEnable { + close(c.events) + } + + // librdkafka's rd_kafka_consumer_close() will block + // and trigger the rebalance_cb() if one is set, if not, which is the + // case with the Go client since it registers EVENTs rather than callbacks, + // librdkafka will shortcut the rebalance_cb and do a forced unassign. + // But we can't have that since the application might need the final RevokePartitions + // before shutting down. So we trigger an Unsubscribe() first, wait for that to + // propagate (in the Poll loop below), and then close the consumer. + c.Unsubscribe() + + // Poll for rebalance events + for { + c.Poll(10 * 1000) + if int(C.rd_kafka_queue_length(c.handle.rkq)) == 0 { + break + } + } + + // Destroy our queue + C.rd_kafka_queue_destroy(c.handle.rkq) + c.handle.rkq = nil + + // Close the consumer + C.rd_kafka_consumer_close(c.handle.rk) + + c.handle.cleanup() + + C.rd_kafka_destroy(c.handle.rk) + + return nil +} + +// NewConsumer creates a new high-level Consumer instance. +// +// conf is a *ConfigMap with standard librdkafka configuration properties. +// +// Supported special configuration properties: +// go.application.rebalance.enable (bool, false) - Forward rebalancing responsibility to application via the Events() channel. +// If set to true the app must handle the AssignedPartitions and +// RevokedPartitions events and call Assign() and Unassign() +// respectively. +// go.events.channel.enable (bool, false) - [deprecated] Enable the Events() channel. Messages and events will be pushed on the Events() channel and the Poll() interface will be disabled. +// go.events.channel.size (int, 1000) - Events() channel size +// go.logs.channel.enable (bool, false) - Forward log to Logs() channel. +// go.logs.channel (chan kafka.LogEvent, nil) - Forward logs to application-provided channel instead of Logs(). Requires go.logs.channel.enable=true. +// +// WARNING: Due to the buffering nature of channels (and queues in general) the +// use of the events channel risks receiving outdated events and +// messages. Minimizing go.events.channel.size reduces the risk +// and number of outdated events and messages but does not eliminate +// the factor completely. With a channel size of 1 at most one +// event or message may be outdated. +func NewConsumer(conf *ConfigMap) (*Consumer, error) { + + err := versionCheck() + if err != nil { + return nil, err + } + + // before we do anything with the configuration, create a copy such that + // the original is not mutated. + confCopy := conf.clone() + + groupid, _ := confCopy.get("group.id", nil) + if groupid == nil { + // without a group.id the underlying cgrp subsystem in librdkafka wont get started + // and without it there is no way to consume assigned partitions. + // So for now require the group.id, this might change in the future. + return nil, newErrorFromString(ErrInvalidArg, "Required property group.id not set") + } + + c := &Consumer{} + + v, err := confCopy.extract("go.application.rebalance.enable", false) + if err != nil { + return nil, err + } + c.appRebalanceEnable = v.(bool) + + v, err = confCopy.extract("go.events.channel.enable", false) + if err != nil { + return nil, err + } + c.eventsChanEnable = v.(bool) + + v, err = confCopy.extract("go.events.channel.size", 1000) + if err != nil { + return nil, err + } + eventsChanSize := v.(int) + + logsChanEnable, logsChan, err := confCopy.extractLogConfig() + if err != nil { + return nil, err + } + + cConf, err := confCopy.convert() + if err != nil { + return nil, err + } + cErrstr := (*C.char)(C.malloc(C.size_t(256))) + defer C.free(unsafe.Pointer(cErrstr)) + + C.rd_kafka_conf_set_events(cConf, C.RD_KAFKA_EVENT_REBALANCE|C.RD_KAFKA_EVENT_OFFSET_COMMIT|C.RD_KAFKA_EVENT_STATS|C.RD_KAFKA_EVENT_ERROR|C.RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH) + + c.handle.rk = C.rd_kafka_new(C.RD_KAFKA_CONSUMER, cConf, cErrstr, 256) + if c.handle.rk == nil { + return nil, newErrorFromCString(C.RD_KAFKA_RESP_ERR__INVALID_ARG, cErrstr) + } + + C.rd_kafka_poll_set_consumer(c.handle.rk) + + c.handle.c = c + c.handle.setup() + c.readerTermChan = make(chan bool) + c.handle.rkq = C.rd_kafka_queue_get_consumer(c.handle.rk) + if c.handle.rkq == nil { + // no cgrp (no group.id configured), revert to main queue. + c.handle.rkq = C.rd_kafka_queue_get_main(c.handle.rk) + } + + if logsChanEnable { + c.handle.setupLogQueue(logsChan, c.readerTermChan) + } + + if c.eventsChanEnable { + c.events = make(chan Event, eventsChanSize) + /* Start rdkafka consumer queue reader -> events writer goroutine */ + c.handle.waitGroup.Add(1) + go func() { + consumerReader(c, c.readerTermChan) + c.handle.waitGroup.Done() + }() + } + + return c, nil +} + +// consumerReader reads messages and events from the librdkafka consumer queue +// and posts them on the consumer channel. +// Runs until termChan closes +func consumerReader(c *Consumer, termChan chan bool) { + for { + select { + case _ = <-termChan: + return + default: + _, term := c.handle.eventPoll(c.events, 100, 1000, termChan) + if term { + return + } + + } + } +} + +// GetMetadata queries broker for cluster and topic metadata. +// If topic is non-nil only information about that topic is returned, else if +// allTopics is false only information about locally used topics is returned, +// else information about all topics is returned. +// GetMetadata is equivalent to listTopics, describeTopics and describeCluster in the Java API. +func (c *Consumer) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error) { + return getMetadata(c, topic, allTopics, timeoutMs) +} + +// QueryWatermarkOffsets queries the broker for the low and high offsets for the given topic and partition. +func (c *Consumer) QueryWatermarkOffsets(topic string, partition int32, timeoutMs int) (low, high int64, err error) { + return queryWatermarkOffsets(c, topic, partition, timeoutMs) +} + +// GetWatermarkOffsets returns the cached low and high offsets for the given topic +// and partition. The high offset is populated on every fetch response or via calling QueryWatermarkOffsets. +// The low offset is populated every statistics.interval.ms if that value is set. +// OffsetInvalid will be returned if there is no cached offset for either value. +func (c *Consumer) GetWatermarkOffsets(topic string, partition int32) (low, high int64, err error) { + return getWatermarkOffsets(c, topic, partition) +} + +// OffsetsForTimes looks up offsets by timestamp for the given partitions. +// +// The returned offset for each partition is the earliest offset whose +// timestamp is greater than or equal to the given timestamp in the +// corresponding partition. If the provided timestamp exceeds that of the +// last message in the partition, a value of -1 will be returned. +// +// The timestamps to query are represented as `.Offset` in the `times` +// argument and the looked up offsets are represented as `.Offset` in the returned +// `offsets` list. +// +// The function will block for at most timeoutMs milliseconds. +// +// Duplicate Topic+Partitions are not supported. +// Per-partition errors may be returned in the `.Error` field. +func (c *Consumer) OffsetsForTimes(times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) { + return offsetsForTimes(c, times, timeoutMs) +} + +// Subscription returns the current subscription as set by Subscribe() +func (c *Consumer) Subscription() (topics []string, err error) { + var cTopics *C.rd_kafka_topic_partition_list_t + + cErr := C.rd_kafka_subscription(c.handle.rk, &cTopics) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cErr) + } + defer C.rd_kafka_topic_partition_list_destroy(cTopics) + + topicCnt := int(cTopics.cnt) + topics = make([]string, topicCnt) + for i := 0; i < topicCnt; i++ { + crktpar := C._c_rdkafka_topic_partition_list_entry(cTopics, + C.int(i)) + topics[i] = C.GoString(crktpar.topic) + } + + return topics, nil +} + +// Assignment returns the current partition assignments +func (c *Consumer) Assignment() (partitions []TopicPartition, err error) { + var cParts *C.rd_kafka_topic_partition_list_t + + cErr := C.rd_kafka_assignment(c.handle.rk, &cParts) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cErr) + } + defer C.rd_kafka_topic_partition_list_destroy(cParts) + + partitions = newTopicPartitionsFromCparts(cParts) + + return partitions, nil +} + +// Committed retrieves committed offsets for the given set of partitions +func (c *Consumer) Committed(partitions []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) { + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + cerr := C.rd_kafka_committed(c.handle.rk, cparts, C.int(timeoutMs)) + if cerr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cerr) + } + + return newTopicPartitionsFromCparts(cparts), nil +} + +// Position returns the current consume position for the given partitions. +// Typical use is to call Assignment() to get the partition list +// and then pass it to Position() to get the current consume position for +// each of the assigned partitions. +// The consume position is the next message to read from the partition. +// i.e., the offset of the last message seen by the application + 1. +func (c *Consumer) Position(partitions []TopicPartition) (offsets []TopicPartition, err error) { + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + cerr := C.rd_kafka_position(c.handle.rk, cparts) + if cerr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cerr) + } + + return newTopicPartitionsFromCparts(cparts), nil +} + +// Pause consumption for the provided list of partitions +// +// Note that messages already enqueued on the consumer's Event channel +// (if `go.events.channel.enable` has been set) will NOT be purged by +// this call, set `go.events.channel.size` accordingly. +func (c *Consumer) Pause(partitions []TopicPartition) (err error) { + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + cerr := C.rd_kafka_pause_partitions(c.handle.rk, cparts) + if cerr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(cerr) + } + return nil +} + +// Resume consumption for the provided list of partitions +func (c *Consumer) Resume(partitions []TopicPartition) (err error) { + cparts := newCPartsFromTopicPartitions(partitions) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + cerr := C.rd_kafka_resume_partitions(c.handle.rk, cparts) + if cerr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(cerr) + } + return nil +} + +// SetOAuthBearerToken sets the the data to be transmitted +// to a broker during SASL/OAUTHBEARER authentication. It will return nil +// on success, otherwise an error if: +// 1) the token data is invalid (meaning an expiration time in the past +// or either a token value or an extension key or value that does not meet +// the regular expression requirements as per +// https://tools.ietf.org/html/rfc7628#section-3.1); +// 2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +// 3) SASL/OAUTHBEARER is supported but is not configured as the client's +// authentication mechanism. +func (c *Consumer) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error { + return c.handle.setOAuthBearerToken(oauthBearerToken) +} + +// SetOAuthBearerTokenFailure sets the error message describing why token +// retrieval/setting failed; it also schedules a new token refresh event for 10 +// seconds later so the attempt may be retried. It will return nil on +// success, otherwise an error if: +// 1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +// 2) SASL/OAUTHBEARER is supported but is not configured as the client's +// authentication mechanism. +func (c *Consumer) SetOAuthBearerTokenFailure(errstr string) error { + return c.handle.setOAuthBearerTokenFailure(errstr) +} + +// ConsumerGroupMetadata reflects the current consumer group member metadata. +type ConsumerGroupMetadata struct { + serialized []byte +} + +// serializeConsumerGroupMetadata converts a C metadata object to its +// binary representation so we don't have to hold on to the C object, +// which would require an explicit .Close(). +func serializeConsumerGroupMetadata(cgmd *C.rd_kafka_consumer_group_metadata_t) ([]byte, error) { + var cBuffer *C.void + var cSize C.size_t + cError := C.rd_kafka_consumer_group_metadata_write(cgmd, + (*unsafe.Pointer)(unsafe.Pointer(&cBuffer)), &cSize) + if cError != nil { + return nil, newErrorFromCErrorDestroy(cError) + } + defer C.rd_kafka_mem_free(nil, unsafe.Pointer(cBuffer)) + + return C.GoBytes(unsafe.Pointer(cBuffer), C.int(cSize)), nil +} + +// deserializeConsumerGroupMetadata converts a serialized metadata object +// back to a C object. +func deserializeConsumerGroupMetadata(serialized []byte) (*C.rd_kafka_consumer_group_metadata_t, error) { + var cgmd *C.rd_kafka_consumer_group_metadata_t + + cSerialized := C.CBytes(serialized) + defer C.free(cSerialized) + + cError := C.rd_kafka_consumer_group_metadata_read( + &cgmd, cSerialized, C.size_t(len(serialized))) + if cError != nil { + return nil, newErrorFromCErrorDestroy(cError) + } + + return cgmd, nil +} + +// GetConsumerGroupMetadata returns the consumer's current group metadata. +// This object should be passed to the transactional producer's +// SendOffsetsToTransaction() API. +func (c *Consumer) GetConsumerGroupMetadata() (*ConsumerGroupMetadata, error) { + cgmd := C.rd_kafka_consumer_group_metadata(c.handle.rk) + if cgmd == nil { + return nil, NewError(ErrState, "Consumer group metadata not available", false) + } + defer C.rd_kafka_consumer_group_metadata_destroy(cgmd) + + serialized, err := serializeConsumerGroupMetadata(cgmd) + if err != nil { + return nil, err + } + + return &ConsumerGroupMetadata{serialized}, nil +} + +// NewTestConsumerGroupMetadata creates a new consumer group metadata instance +// mainly for testing use. +// Use GetConsumerGroupMetadata() to retrieve the real metadata. +func NewTestConsumerGroupMetadata(groupID string) (*ConsumerGroupMetadata, error) { + cGroupID := C.CString(groupID) + defer C.free(unsafe.Pointer(cGroupID)) + + cgmd := C.rd_kafka_consumer_group_metadata_new(cGroupID) + if cgmd == nil { + return nil, NewError(ErrInvalidArg, "Failed to create metadata object", false) + } + + defer C.rd_kafka_consumer_group_metadata_destroy(cgmd) + serialized, err := serializeConsumerGroupMetadata(cgmd) + if err != nil { + return nil, err + } + + return &ConsumerGroupMetadata{serialized}, nil +} + +// cEventToRebalanceEvent returns an Event (AssignedPartitions or RevokedPartitions) +// based on the specified rkev. +func cEventToRebalanceEvent(rkev *C.rd_kafka_event_t) Event { + if C.rd_kafka_event_error(rkev) == C.RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS { + var ev AssignedPartitions + ev.Partitions = newTopicPartitionsFromCparts(C.rd_kafka_event_topic_partition_list(rkev)) + return ev + } else if C.rd_kafka_event_error(rkev) == C.RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS { + var ev RevokedPartitions + ev.Partitions = newTopicPartitionsFromCparts(C.rd_kafka_event_topic_partition_list(rkev)) + return ev + } else { + panic(fmt.Sprintf("Unable to create rebalance event from C type %s", + C.GoString(C.rd_kafka_err2name(C.rd_kafka_event_error(rkev))))) + } + +} + +// handleRebalanceEvent handles a assign/rebalance rebalance event. +// +// If the app provided a RebalanceCb to Subscribe*() or +// has go.application.rebalance.enable=true we create an event +// and forward it to the application thru the RebalanceCb or the +// Events channel respectively. +// Since librdkafka requires the rebalance event to be "acked" by +// the application (by calling *assign()) to synchronize state we keep track +// of if the application performed *Assign() or *Unassign(), but this only +// works for the non-channel case. For the channel case we assume the +// application calls *Assign() or *Unassign(). +// Failure to do so will "hang" the consumer, e.g., it wont start consuming +// and it wont close cleanly, so this error case should be visible +// immediately to the application developer. +// +// In the polling case (not channel based consumer) the rebalance event +// is returned in retval, else nil is returned. + +func (c *Consumer) handleRebalanceEvent(channel chan Event, rkev *C.rd_kafka_event_t) (retval Event) { + + var ev Event + + if c.rebalanceCb != nil || c.appRebalanceEnable { + // Application has a rebalance callback or has enabled + // rebalances on the events channel, create the appropriate Event. + ev = cEventToRebalanceEvent(rkev) + + } + + if channel != nil && c.appRebalanceEnable && c.rebalanceCb == nil { + // Channel-based consumer with rebalancing enabled, + // return the rebalance event and rely on the application + // to call *Assign() / *Unassign(). + return ev + } + + // Call the application's rebalance callback, if any. + if c.rebalanceCb != nil { + // Mark .appReassigned as false to keep track of whether the + // application called *Assign() / *Unassign(). + c.appReassigned = false + + c.rebalanceCb(c, ev) + + if c.appReassigned { + // Rebalance event handled by application. + return nil + } + } + + // Either there was no rebalance callback, or the application + // did not call *Assign / *Unassign, so we need to do it. + + isCooperative := c.GetRebalanceProtocol() == "COOPERATIVE" + var cError *C.rd_kafka_error_t + var cErr C.rd_kafka_resp_err_t + + if C.rd_kafka_event_error(rkev) == C.RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS { + // Assign partitions + if isCooperative { + cError = C.rd_kafka_incremental_assign( + c.handle.rk, + C.rd_kafka_event_topic_partition_list(rkev)) + } else { + cErr = C.rd_kafka_assign( + c.handle.rk, + C.rd_kafka_event_topic_partition_list(rkev)) + } + } else { + // Revoke partitions + + if isCooperative { + cError = C.rd_kafka_incremental_unassign( + c.handle.rk, + C.rd_kafka_event_topic_partition_list(rkev)) + } else { + cErr = C.rd_kafka_assign(c.handle.rk, nil) + } + } + + // If the *assign() call returned error, forward it to the + // the consumer's Events() channel for visibility. + if cError != nil { + c.events <- newErrorFromCErrorDestroy(cError) + } else if cErr != 0 { + c.events <- newError(cErr) + } + + return nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/context.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/context.go new file mode 100644 index 0000000..85709be --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/context.go @@ -0,0 +1,31 @@ +/** + * Copyright 2019 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import ( + "context" + "time" +) + +// Timeout returns the remaining time after which work done on behalf of this context should be +// canceled, or ok==false if no deadline/timeout is set. +func timeout(ctx context.Context) (timeout time.Duration, ok bool) { + if deadline, ok := ctx.Deadline(); ok { + return deadline.Sub(time.Now()), true + } + return 0, false +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/error.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/error.go new file mode 100644 index 0000000..1827f43 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/error.go @@ -0,0 +1,154 @@ +package kafka + +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +// Automatically generate error codes from librdkafka +// See README for instructions +//go:generate $GOPATH/bin/go_rdkafka_generr generated_errors.go + +/* +#include +#include "select_rdkafka.h" +*/ +import "C" + +import ( + "fmt" + "unsafe" +) + +// Error provides a Kafka-specific error container +type Error struct { + code ErrorCode + str string + fatal bool + retriable bool + txnRequiresAbort bool +} + +func newError(code C.rd_kafka_resp_err_t) (err Error) { + return Error{code: ErrorCode(code)} +} + +// NewError creates a new Error. +func NewError(code ErrorCode, str string, fatal bool) (err Error) { + return Error{code: code, str: str, fatal: fatal} +} + +func newErrorFromString(code ErrorCode, str string) (err Error) { + return Error{code: code, str: str} +} + +func newErrorFromCString(code C.rd_kafka_resp_err_t, cstr *C.char) (err Error) { + var str string + if cstr != nil { + str = C.GoString(cstr) + } else { + str = "" + } + return Error{code: ErrorCode(code), str: str} +} + +func newCErrorFromString(code C.rd_kafka_resp_err_t, str string) (err Error) { + return newErrorFromString(ErrorCode(code), str) +} + +// newErrorFromCError creates a new Error instance and destroys +// the passed cError. +func newErrorFromCErrorDestroy(cError *C.rd_kafka_error_t) Error { + defer C.rd_kafka_error_destroy(cError) + return Error{ + code: ErrorCode(C.rd_kafka_error_code(cError)), + str: C.GoString(C.rd_kafka_error_string(cError)), + fatal: cint2bool(C.rd_kafka_error_is_fatal(cError)), + retriable: cint2bool(C.rd_kafka_error_is_retriable(cError)), + txnRequiresAbort: cint2bool(C.rd_kafka_error_txn_requires_abort(cError)), + } +} + +// Error returns a human readable representation of an Error +// Same as Error.String() +func (e Error) Error() string { + return e.String() +} + +// String returns a human readable representation of an Error +func (e Error) String() string { + var errstr string + if len(e.str) > 0 { + errstr = e.str + } else { + errstr = e.code.String() + } + + if e.IsFatal() { + return fmt.Sprintf("Fatal error: %s", errstr) + } + + return errstr +} + +// Code returns the ErrorCode of an Error +func (e Error) Code() ErrorCode { + return e.code +} + +// IsFatal returns true if the error is a fatal error. +// A fatal error indicates the client instance is no longer operable and +// should be terminated. Typical causes include non-recoverable +// idempotent producer errors. +func (e Error) IsFatal() bool { + return e.fatal +} + +// IsRetriable returns true if the operation that caused this error +// may be retried. +// This flag is currently only set by the Transactional producer API. +func (e Error) IsRetriable() bool { + return e.retriable +} + +// TxnRequiresAbort returns true if the error is an abortable transaction error +// that requires the application to abort the current transaction with +// AbortTransaction() and start a new transaction with BeginTransaction() +// if it wishes to proceed with transactional operations. +// This flag is only set by the Transactional producer API. +func (e Error) TxnRequiresAbort() bool { + return e.txnRequiresAbort +} + +// getFatalError returns an Error object if the client instance has raised a fatal error, else nil. +func getFatalError(H Handle) error { + cErrstr := (*C.char)(C.malloc(C.size_t(512))) + defer C.free(unsafe.Pointer(cErrstr)) + + cErr := C.rd_kafka_fatal_error(H.gethandle().rk, cErrstr, 512) + if int(cErr) == 0 { + return nil + } + + err := newErrorFromCString(cErr, cErrstr) + err.fatal = true + + return err +} + +// testFatalError triggers a fatal error in the underlying client. +// This is to be used strictly for testing purposes. +func testFatalError(H Handle, code ErrorCode, str string) ErrorCode { + return ErrorCode(C.rd_kafka_test_fatal_error(H.gethandle().rk, C.rd_kafka_resp_err_t(code), C.CString(str))) +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/error_gen.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/error_gen.go new file mode 100644 index 0000000..4ceecb3 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/error_gen.go @@ -0,0 +1,112 @@ +package kafka + +/** + * Copyright 2020 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +// Automatically generate error codes from librdkafka +// See README for instructions +//go:generate $GOPATH/bin/go_rdkafka_generr generated_errors.go + +/* +#include +#include "select_rdkafka.h" + +static const char *errdesc_to_string (const struct rd_kafka_err_desc *ed, int idx) { + return ed[idx].name; +} + +static const char *errdesc_to_desc (const struct rd_kafka_err_desc *ed, int idx) { + return ed[idx].desc; +} +*/ +import "C" + +import ( + "fmt" + "os" + "strings" + "time" +) + +// camelCase transforms a snake_case string to camelCase. +func camelCase(s string) string { + ret := "" + for _, v := range strings.Split(s, "_") { + if len(v) == 0 { + continue + } + ret += strings.ToUpper((string)(v[0])) + strings.ToLower(v[1:]) + } + return ret +} + +// WriteErrorCodes writes Go error code constants to file from the +// librdkafka error codes. +// This function is not intended for public use. +func WriteErrorCodes(f *os.File) { + f.WriteString("package kafka\n") + now := time.Now() + f.WriteString(fmt.Sprintf("// Copyright 2016-%d Confluent Inc.\n", now.Year())) + f.WriteString(fmt.Sprintf("// AUTOMATICALLY GENERATED ON %v USING librdkafka %s\n", + now, C.GoString(C.rd_kafka_version_str()))) + + var errdescs *C.struct_rd_kafka_err_desc + var csize C.size_t + C.rd_kafka_get_err_descs(&errdescs, &csize) + + f.WriteString(` +/* +#include "select_rdkafka.h" +*/ +import "C" + +// ErrorCode is the integer representation of local and broker error codes +type ErrorCode int + +// String returns a human readable representation of an error code +func (c ErrorCode) String() string { + return C.GoString(C.rd_kafka_err2str(C.rd_kafka_resp_err_t(c))) +} + +const ( +`) + + for i := 0; i < int(csize); i++ { + orig := C.GoString(C.errdesc_to_string(errdescs, C.int(i))) + if len(orig) == 0 { + continue + } + desc := C.GoString(C.errdesc_to_desc(errdescs, C.int(i))) + if len(desc) == 0 { + continue + } + + errname := "Err" + camelCase(orig) + + // Special handling to please golint + // Eof -> EOF + // Id -> ID + errname = strings.Replace(errname, "Eof", "EOF", -1) + errname = strings.Replace(errname, "Id", "ID", -1) + + f.WriteString(fmt.Sprintf(" // %s %s\n", errname, desc)) + f.WriteString(fmt.Sprintf(" %s ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_%s)\n", + errname, orig)) + } + + f.WriteString(")\n") + +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/event.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/event.go new file mode 100644 index 0000000..6357ad8 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/event.go @@ -0,0 +1,316 @@ +package kafka + +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "fmt" + "os" + "unsafe" +) + +/* +#include +#include "select_rdkafka.h" +#include "glue_rdkafka.h" + + +void chdrs_to_tmphdrs (glue_msg_t *gMsg) { + size_t i = 0; + const char *name; + const void *val; + size_t size; + rd_kafka_headers_t *chdrs; + + if (rd_kafka_message_headers(gMsg->msg, &chdrs)) { + gMsg->tmphdrs = NULL; + gMsg->tmphdrsCnt = 0; + return; + } + + gMsg->tmphdrsCnt = rd_kafka_header_cnt(chdrs); + gMsg->tmphdrs = malloc(sizeof(*gMsg->tmphdrs) * gMsg->tmphdrsCnt); + + while (!rd_kafka_header_get_all(chdrs, i, + &gMsg->tmphdrs[i].key, + &gMsg->tmphdrs[i].val, + (size_t *)&gMsg->tmphdrs[i].size)) + i++; +} + +rd_kafka_event_t *_rk_queue_poll (rd_kafka_queue_t *rkq, int timeoutMs, + rd_kafka_event_type_t *evtype, + glue_msg_t *gMsg, + rd_kafka_event_t *prev_rkev) { + rd_kafka_event_t *rkev; + + if (prev_rkev) + rd_kafka_event_destroy(prev_rkev); + + rkev = rd_kafka_queue_poll(rkq, timeoutMs); + *evtype = rd_kafka_event_type(rkev); + + if (*evtype == RD_KAFKA_EVENT_FETCH) { + gMsg->msg = (rd_kafka_message_t *)rd_kafka_event_message_next(rkev); + gMsg->ts = rd_kafka_message_timestamp(gMsg->msg, &gMsg->tstype); + + if (gMsg->want_hdrs) + chdrs_to_tmphdrs(gMsg); + } + + return rkev; +} +*/ +import "C" + +func chdrsToTmphdrs(gMsg *C.glue_msg_t) { + C.chdrs_to_tmphdrs(gMsg) +} + +// Event generic interface +type Event interface { + // String returns a human-readable representation of the event + String() string +} + +// Specific event types + +// Stats statistics event +type Stats struct { + statsJSON string +} + +func (e Stats) String() string { + return e.statsJSON +} + +// AssignedPartitions consumer group rebalance event: assigned partition set +type AssignedPartitions struct { + Partitions []TopicPartition +} + +func (e AssignedPartitions) String() string { + return fmt.Sprintf("AssignedPartitions: %v", e.Partitions) +} + +// RevokedPartitions consumer group rebalance event: revoked partition set +type RevokedPartitions struct { + Partitions []TopicPartition +} + +func (e RevokedPartitions) String() string { + return fmt.Sprintf("RevokedPartitions: %v", e.Partitions) +} + +// PartitionEOF consumer reached end of partition +// Needs to be explicitly enabled by setting the `enable.partition.eof` +// configuration property to true. +type PartitionEOF TopicPartition + +func (p PartitionEOF) String() string { + return fmt.Sprintf("EOF at %s", TopicPartition(p)) +} + +// OffsetsCommitted reports committed offsets +type OffsetsCommitted struct { + Error error + Offsets []TopicPartition +} + +func (o OffsetsCommitted) String() string { + return fmt.Sprintf("OffsetsCommitted (%v, %v)", o.Error, o.Offsets) +} + +// OAuthBearerTokenRefresh indicates token refresh is required +type OAuthBearerTokenRefresh struct { + // Config is the value of the sasl.oauthbearer.config property + Config string +} + +func (o OAuthBearerTokenRefresh) String() string { + return "OAuthBearerTokenRefresh" +} + +// eventPoll polls an event from the handler's C rd_kafka_queue_t, +// translates it into an Event type and then sends on `channel` if non-nil, else returns the Event. +// term_chan is an optional channel to monitor along with producing to channel +// to indicate that `channel` is being terminated. +// returns (event Event, terminate Bool) tuple, where Terminate indicates +// if termChan received a termination event. +func (h *handle) eventPoll(channel chan Event, timeoutMs int, maxEvents int, termChan chan bool) (Event, bool) { + + var prevRkev *C.rd_kafka_event_t + term := false + + var retval Event + + if channel == nil { + maxEvents = 1 + } +out: + for evcnt := 0; evcnt < maxEvents; evcnt++ { + var evtype C.rd_kafka_event_type_t + var gMsg C.glue_msg_t + gMsg.want_hdrs = C.int8_t(bool2cint(h.msgFields.Headers)) + rkev := C._rk_queue_poll(h.rkq, C.int(timeoutMs), &evtype, &gMsg, prevRkev) + prevRkev = rkev + timeoutMs = 0 + + retval = nil + + switch evtype { + case C.RD_KAFKA_EVENT_FETCH: + // Consumer fetch event, new message. + // Extracted into temporary gMsg for optimization + retval = h.newMessageFromGlueMsg(&gMsg) + + case C.RD_KAFKA_EVENT_REBALANCE: + // Consumer rebalance event + retval = h.c.handleRebalanceEvent(channel, rkev) + + case C.RD_KAFKA_EVENT_ERROR: + // Error event + cErr := C.rd_kafka_event_error(rkev) + if cErr == C.RD_KAFKA_RESP_ERR__PARTITION_EOF { + crktpar := C.rd_kafka_event_topic_partition(rkev) + if crktpar == nil { + break + } + + defer C.rd_kafka_topic_partition_destroy(crktpar) + var peof PartitionEOF + setupTopicPartitionFromCrktpar((*TopicPartition)(&peof), crktpar) + + retval = peof + + } else if int(C.rd_kafka_event_error_is_fatal(rkev)) != 0 { + // A fatal error has been raised. + // Extract the actual error from the client + // instance and return a new Error with + // fatal set to true. + cFatalErrstrSize := C.size_t(512) + cFatalErrstr := (*C.char)(C.malloc(cFatalErrstrSize)) + defer C.free(unsafe.Pointer(cFatalErrstr)) + cFatalErr := C.rd_kafka_fatal_error(h.rk, cFatalErrstr, cFatalErrstrSize) + fatalErr := newErrorFromCString(cFatalErr, cFatalErrstr) + fatalErr.fatal = true + retval = fatalErr + + } else { + retval = newErrorFromCString(cErr, C.rd_kafka_event_error_string(rkev)) + } + + case C.RD_KAFKA_EVENT_STATS: + retval = &Stats{C.GoString(C.rd_kafka_event_stats(rkev))} + + case C.RD_KAFKA_EVENT_DR: + // Producer Delivery Report event + // Each such event contains delivery reports for all + // messages in the produced batch. + // Forward delivery reports to per-message's response channel + // or to the global Producer.Events channel, or none. + rkmessages := make([]*C.rd_kafka_message_t, int(C.rd_kafka_event_message_count(rkev))) + + cnt := int(C.rd_kafka_event_message_array(rkev, (**C.rd_kafka_message_t)(unsafe.Pointer(&rkmessages[0])), C.size_t(len(rkmessages)))) + + for _, rkmessage := range rkmessages[:cnt] { + msg := h.newMessageFromC(rkmessage) + var ch *chan Event + + if rkmessage._private != nil { + // Find cgoif by id + cg, found := h.cgoGet((int)((uintptr)(rkmessage._private))) + if found { + cdr := cg.(cgoDr) + + if cdr.deliveryChan != nil { + ch = &cdr.deliveryChan + } + msg.Opaque = cdr.opaque + } + } + + if ch == nil && h.fwdDr { + ch = &channel + } + + if ch != nil { + select { + case *ch <- msg: + case <-termChan: + retval = nil + term = true + break out + } + + } else { + retval = msg + break out + } + } + + case C.RD_KAFKA_EVENT_OFFSET_COMMIT: + // Offsets committed + cErr := C.rd_kafka_event_error(rkev) + coffsets := C.rd_kafka_event_topic_partition_list(rkev) + var offsets []TopicPartition + if coffsets != nil { + offsets = newTopicPartitionsFromCparts(coffsets) + } + + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + retval = OffsetsCommitted{newErrorFromCString(cErr, C.rd_kafka_event_error_string(rkev)), offsets} + } else { + retval = OffsetsCommitted{nil, offsets} + } + + case C.RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH: + ev := OAuthBearerTokenRefresh{C.GoString(C.rd_kafka_event_config_string(rkev))} + retval = ev + + case C.RD_KAFKA_EVENT_NONE: + // poll timed out: no events available + break out + + default: + if rkev != nil { + fmt.Fprintf(os.Stderr, "Ignored event %s\n", + C.GoString(C.rd_kafka_event_name(rkev))) + } + + } + + if retval != nil { + if channel != nil { + select { + case channel <- retval: + case <-termChan: + retval = nil + term = true + break out + } + } else { + break out + } + } + } + + if prevRkev != nil { + C.rd_kafka_event_destroy(prevRkev) + } + + return retval, term +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/generated_errors.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/generated_errors.go new file mode 100644 index 0000000..fb828f9 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/generated_errors.go @@ -0,0 +1,337 @@ +package kafka +// Copyright 2016-2021 Confluent Inc. +// AUTOMATICALLY GENERATED ON 2021-12-08 12:44:39.243338672 +0100 CET m=+0.000248284 USING librdkafka 1.8.2 + +/* +#include "select_rdkafka.h" +*/ +import "C" + +// ErrorCode is the integer representation of local and broker error codes +type ErrorCode int + +// String returns a human readable representation of an error code +func (c ErrorCode) String() string { + return C.GoString(C.rd_kafka_err2str(C.rd_kafka_resp_err_t(c))) +} + +const ( + // ErrBadMsg Local: Bad message format + ErrBadMsg ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__BAD_MSG) + // ErrBadCompression Local: Invalid compressed data + ErrBadCompression ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__BAD_COMPRESSION) + // ErrDestroy Local: Broker handle destroyed + ErrDestroy ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__DESTROY) + // ErrFail Local: Communication failure with broker + ErrFail ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FAIL) + // ErrTransport Local: Broker transport failure + ErrTransport ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__TRANSPORT) + // ErrCritSysResource Local: Critical system resource failure + ErrCritSysResource ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__CRIT_SYS_RESOURCE) + // ErrResolve Local: Host resolution failure + ErrResolve ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__RESOLVE) + // ErrMsgTimedOut Local: Message timed out + ErrMsgTimedOut ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__MSG_TIMED_OUT) + // ErrPartitionEOF Broker: No more messages + ErrPartitionEOF ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PARTITION_EOF) + // ErrUnknownPartition Local: Unknown partition + ErrUnknownPartition ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION) + // ErrFs Local: File or filesystem error + ErrFs ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FS) + // ErrUnknownTopic Local: Unknown topic + ErrUnknownTopic ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_TOPIC) + // ErrAllBrokersDown Local: All broker connections are down + ErrAllBrokersDown ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ALL_BROKERS_DOWN) + // ErrInvalidArg Local: Invalid argument or configuration + ErrInvalidArg ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INVALID_ARG) + // ErrTimedOut Local: Timed out + ErrTimedOut ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__TIMED_OUT) + // ErrQueueFull Local: Queue full + ErrQueueFull ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__QUEUE_FULL) + // ErrIsrInsuff Local: ISR count insufficient + ErrIsrInsuff ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ISR_INSUFF) + // ErrNodeUpdate Local: Broker node update + ErrNodeUpdate ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NODE_UPDATE) + // ErrSsl Local: SSL error + ErrSsl ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__SSL) + // ErrWaitCoord Local: Waiting for coordinator + ErrWaitCoord ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__WAIT_COORD) + // ErrUnknownGroup Local: Unknown group + ErrUnknownGroup ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_GROUP) + // ErrInProgress Local: Operation in progress + ErrInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__IN_PROGRESS) + // ErrPrevInProgress Local: Previous operation in progress + ErrPrevInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PREV_IN_PROGRESS) + // ErrExistingSubscription Local: Existing subscription + ErrExistingSubscription ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__EXISTING_SUBSCRIPTION) + // ErrAssignPartitions Local: Assign partitions + ErrAssignPartitions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS) + // ErrRevokePartitions Local: Revoke partitions + ErrRevokePartitions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS) + // ErrConflict Local: Conflicting use + ErrConflict ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__CONFLICT) + // ErrState Local: Erroneous state + ErrState ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__STATE) + // ErrUnknownProtocol Local: Unknown protocol + ErrUnknownProtocol ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_PROTOCOL) + // ErrNotImplemented Local: Not implemented + ErrNotImplemented ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED) + // ErrAuthentication Local: Authentication failure + ErrAuthentication ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__AUTHENTICATION) + // ErrNoOffset Local: No offset stored + ErrNoOffset ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NO_OFFSET) + // ErrOutdated Local: Outdated + ErrOutdated ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__OUTDATED) + // ErrTimedOutQueue Local: Timed out in queue + ErrTimedOutQueue ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__TIMED_OUT_QUEUE) + // ErrUnsupportedFeature Local: Required feature not supported by broker + ErrUnsupportedFeature ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNSUPPORTED_FEATURE) + // ErrWaitCache Local: Awaiting cache update + ErrWaitCache ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__WAIT_CACHE) + // ErrIntr Local: Operation interrupted + ErrIntr ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INTR) + // ErrKeySerialization Local: Key serialization error + ErrKeySerialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__KEY_SERIALIZATION) + // ErrValueSerialization Local: Value serialization error + ErrValueSerialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__VALUE_SERIALIZATION) + // ErrKeyDeserialization Local: Key deserialization error + ErrKeyDeserialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__KEY_DESERIALIZATION) + // ErrValueDeserialization Local: Value deserialization error + ErrValueDeserialization ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__VALUE_DESERIALIZATION) + // ErrPartial Local: Partial response + ErrPartial ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PARTIAL) + // ErrReadOnly Local: Read-only object + ErrReadOnly ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__READ_ONLY) + // ErrNoent Local: No such entry + ErrNoent ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOENT) + // ErrUnderflow Local: Read underflow + ErrUnderflow ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNDERFLOW) + // ErrInvalidType Local: Invalid type + ErrInvalidType ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INVALID_TYPE) + // ErrRetry Local: Retry operation + ErrRetry ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__RETRY) + // ErrPurgeQueue Local: Purged in queue + ErrPurgeQueue ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PURGE_QUEUE) + // ErrPurgeInflight Local: Purged in flight + ErrPurgeInflight ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__PURGE_INFLIGHT) + // ErrFatal Local: Fatal error + ErrFatal ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FATAL) + // ErrInconsistent Local: Inconsistent state + ErrInconsistent ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__INCONSISTENT) + // ErrGaplessGuarantee Local: Gap-less ordering would not be guaranteed if proceeding + ErrGaplessGuarantee ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__GAPLESS_GUARANTEE) + // ErrMaxPollExceeded Local: Maximum application poll interval (max.poll.interval.ms) exceeded + ErrMaxPollExceeded ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__MAX_POLL_EXCEEDED) + // ErrUnknownBroker Local: Unknown broker + ErrUnknownBroker ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__UNKNOWN_BROKER) + // ErrNotConfigured Local: Functionality not configured + ErrNotConfigured ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOT_CONFIGURED) + // ErrFenced Local: This instance has been fenced by a newer instance + ErrFenced ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__FENCED) + // ErrApplication Local: Application generated error + ErrApplication ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__APPLICATION) + // ErrAssignmentLost Local: Group partition assignment lost + ErrAssignmentLost ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__ASSIGNMENT_LOST) + // ErrNoop Local: No operation performed + ErrNoop ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__NOOP) + // ErrAutoOffsetReset Local: No offset to automatically reset to + ErrAutoOffsetReset ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR__AUTO_OFFSET_RESET) + // ErrUnknown Unknown broker error + ErrUnknown ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN) + // ErrNoError Success + ErrNoError ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NO_ERROR) + // ErrOffsetOutOfRange Broker: Offset out of range + ErrOffsetOutOfRange ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OFFSET_OUT_OF_RANGE) + // ErrInvalidMsg Broker: Invalid message + ErrInvalidMsg ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_MSG) + // ErrUnknownTopicOrPart Broker: Unknown topic or partition + ErrUnknownTopicOrPart ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_TOPIC_OR_PART) + // ErrInvalidMsgSize Broker: Invalid message size + ErrInvalidMsgSize ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_MSG_SIZE) + // ErrLeaderNotAvailable Broker: Leader not available + ErrLeaderNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_LEADER_NOT_AVAILABLE) + // ErrNotLeaderForPartition Broker: Not leader for partition + ErrNotLeaderForPartition ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_LEADER_FOR_PARTITION) + // ErrRequestTimedOut Broker: Request timed out + ErrRequestTimedOut ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REQUEST_TIMED_OUT) + // ErrBrokerNotAvailable Broker: Broker not available + ErrBrokerNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_BROKER_NOT_AVAILABLE) + // ErrReplicaNotAvailable Broker: Replica not available + ErrReplicaNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REPLICA_NOT_AVAILABLE) + // ErrMsgSizeTooLarge Broker: Message size too large + ErrMsgSizeTooLarge ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_MSG_SIZE_TOO_LARGE) + // ErrStaleCtrlEpoch Broker: StaleControllerEpochCode + ErrStaleCtrlEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_STALE_CTRL_EPOCH) + // ErrOffsetMetadataTooLarge Broker: Offset metadata string too large + ErrOffsetMetadataTooLarge ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OFFSET_METADATA_TOO_LARGE) + // ErrNetworkException Broker: Broker disconnected before response received + ErrNetworkException ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NETWORK_EXCEPTION) + // ErrCoordinatorLoadInProgress Broker: Coordinator load in progress + ErrCoordinatorLoadInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_COORDINATOR_LOAD_IN_PROGRESS) + // ErrCoordinatorNotAvailable Broker: Coordinator not available + ErrCoordinatorNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_COORDINATOR_NOT_AVAILABLE) + // ErrNotCoordinator Broker: Not coordinator + ErrNotCoordinator ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_COORDINATOR) + // ErrTopicException Broker: Invalid topic + ErrTopicException ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_EXCEPTION) + // ErrRecordListTooLarge Broker: Message batch larger than configured server segment size + ErrRecordListTooLarge ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_RECORD_LIST_TOO_LARGE) + // ErrNotEnoughReplicas Broker: Not enough in-sync replicas + ErrNotEnoughReplicas ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_ENOUGH_REPLICAS) + // ErrNotEnoughReplicasAfterAppend Broker: Message(s) written to insufficient number of in-sync replicas + ErrNotEnoughReplicasAfterAppend ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_ENOUGH_REPLICAS_AFTER_APPEND) + // ErrInvalidRequiredAcks Broker: Invalid required acks value + ErrInvalidRequiredAcks ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REQUIRED_ACKS) + // ErrIllegalGeneration Broker: Specified group generation id is not valid + ErrIllegalGeneration ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ILLEGAL_GENERATION) + // ErrInconsistentGroupProtocol Broker: Inconsistent group protocol + ErrInconsistentGroupProtocol ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INCONSISTENT_GROUP_PROTOCOL) + // ErrInvalidGroupID Broker: Invalid group.id + ErrInvalidGroupID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_GROUP_ID) + // ErrUnknownMemberID Broker: Unknown member + ErrUnknownMemberID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_MEMBER_ID) + // ErrInvalidSessionTimeout Broker: Invalid session timeout + ErrInvalidSessionTimeout ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_SESSION_TIMEOUT) + // ErrRebalanceInProgress Broker: Group rebalance in progress + ErrRebalanceInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REBALANCE_IN_PROGRESS) + // ErrInvalidCommitOffsetSize Broker: Commit offset data size is not valid + ErrInvalidCommitOffsetSize ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_COMMIT_OFFSET_SIZE) + // ErrTopicAuthorizationFailed Broker: Topic authorization failed + ErrTopicAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_AUTHORIZATION_FAILED) + // ErrGroupAuthorizationFailed Broker: Group authorization failed + ErrGroupAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_AUTHORIZATION_FAILED) + // ErrClusterAuthorizationFailed Broker: Cluster authorization failed + ErrClusterAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_CLUSTER_AUTHORIZATION_FAILED) + // ErrInvalidTimestamp Broker: Invalid timestamp + ErrInvalidTimestamp ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_TIMESTAMP) + // ErrUnsupportedSaslMechanism Broker: Unsupported SASL mechanism + ErrUnsupportedSaslMechanism ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_SASL_MECHANISM) + // ErrIllegalSaslState Broker: Request not valid in current SASL state + ErrIllegalSaslState ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ILLEGAL_SASL_STATE) + // ErrUnsupportedVersion Broker: API version not supported + ErrUnsupportedVersion ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_VERSION) + // ErrTopicAlreadyExists Broker: Topic already exists + ErrTopicAlreadyExists ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_ALREADY_EXISTS) + // ErrInvalidPartitions Broker: Invalid number of partitions + ErrInvalidPartitions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PARTITIONS) + // ErrInvalidReplicationFactor Broker: Invalid replication factor + ErrInvalidReplicationFactor ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REPLICATION_FACTOR) + // ErrInvalidReplicaAssignment Broker: Invalid replica assignment + ErrInvalidReplicaAssignment ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REPLICA_ASSIGNMENT) + // ErrInvalidConfig Broker: Configuration is invalid + ErrInvalidConfig ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_CONFIG) + // ErrNotController Broker: Not controller for cluster + ErrNotController ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NOT_CONTROLLER) + // ErrInvalidRequest Broker: Invalid request + ErrInvalidRequest ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_REQUEST) + // ErrUnsupportedForMessageFormat Broker: Message format on broker does not support request + ErrUnsupportedForMessageFormat ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_FOR_MESSAGE_FORMAT) + // ErrPolicyViolation Broker: Policy violation + ErrPolicyViolation ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_POLICY_VIOLATION) + // ErrOutOfOrderSequenceNumber Broker: Broker received an out of order sequence number + ErrOutOfOrderSequenceNumber ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OUT_OF_ORDER_SEQUENCE_NUMBER) + // ErrDuplicateSequenceNumber Broker: Broker received a duplicate sequence number + ErrDuplicateSequenceNumber ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DUPLICATE_SEQUENCE_NUMBER) + // ErrInvalidProducerEpoch Broker: Producer attempted an operation with an old epoch + ErrInvalidProducerEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PRODUCER_EPOCH) + // ErrInvalidTxnState Broker: Producer attempted a transactional operation in an invalid state + ErrInvalidTxnState ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_TXN_STATE) + // ErrInvalidProducerIDMapping Broker: Producer attempted to use a producer id which is not currently assigned to its transactional id + ErrInvalidProducerIDMapping ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PRODUCER_ID_MAPPING) + // ErrInvalidTransactionTimeout Broker: Transaction timeout is larger than the maximum value allowed by the broker's max.transaction.timeout.ms + ErrInvalidTransactionTimeout ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_TRANSACTION_TIMEOUT) + // ErrConcurrentTransactions Broker: Producer attempted to update a transaction while another concurrent operation on the same transaction was ongoing + ErrConcurrentTransactions ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_CONCURRENT_TRANSACTIONS) + // ErrTransactionCoordinatorFenced Broker: Indicates that the transaction coordinator sending a WriteTxnMarker is no longer the current coordinator for a given producer + ErrTransactionCoordinatorFenced ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TRANSACTION_COORDINATOR_FENCED) + // ErrTransactionalIDAuthorizationFailed Broker: Transactional Id authorization failed + ErrTransactionalIDAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TRANSACTIONAL_ID_AUTHORIZATION_FAILED) + // ErrSecurityDisabled Broker: Security features are disabled + ErrSecurityDisabled ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_SECURITY_DISABLED) + // ErrOperationNotAttempted Broker: Operation not attempted + ErrOperationNotAttempted ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OPERATION_NOT_ATTEMPTED) + // ErrKafkaStorageError Broker: Disk error when trying to access log file on disk + ErrKafkaStorageError ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_KAFKA_STORAGE_ERROR) + // ErrLogDirNotFound Broker: The user-specified log directory is not found in the broker config + ErrLogDirNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_LOG_DIR_NOT_FOUND) + // ErrSaslAuthenticationFailed Broker: SASL Authentication failed + ErrSaslAuthenticationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_SASL_AUTHENTICATION_FAILED) + // ErrUnknownProducerID Broker: Unknown Producer Id + ErrUnknownProducerID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_PRODUCER_ID) + // ErrReassignmentInProgress Broker: Partition reassignment is in progress + ErrReassignmentInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_REASSIGNMENT_IN_PROGRESS) + // ErrDelegationTokenAuthDisabled Broker: Delegation Token feature is not enabled + ErrDelegationTokenAuthDisabled ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_AUTH_DISABLED) + // ErrDelegationTokenNotFound Broker: Delegation Token is not found on server + ErrDelegationTokenNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_NOT_FOUND) + // ErrDelegationTokenOwnerMismatch Broker: Specified Principal is not valid Owner/Renewer + ErrDelegationTokenOwnerMismatch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_OWNER_MISMATCH) + // ErrDelegationTokenRequestNotAllowed Broker: Delegation Token requests are not allowed on this connection + ErrDelegationTokenRequestNotAllowed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_REQUEST_NOT_ALLOWED) + // ErrDelegationTokenAuthorizationFailed Broker: Delegation Token authorization failed + ErrDelegationTokenAuthorizationFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_AUTHORIZATION_FAILED) + // ErrDelegationTokenExpired Broker: Delegation Token is expired + ErrDelegationTokenExpired ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_EXPIRED) + // ErrInvalidPrincipalType Broker: Supplied principalType is not supported + ErrInvalidPrincipalType ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_PRINCIPAL_TYPE) + // ErrNonEmptyGroup Broker: The group is not empty + ErrNonEmptyGroup ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NON_EMPTY_GROUP) + // ErrGroupIDNotFound Broker: The group id does not exist + ErrGroupIDNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_ID_NOT_FOUND) + // ErrFetchSessionIDNotFound Broker: The fetch session ID was not found + ErrFetchSessionIDNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FETCH_SESSION_ID_NOT_FOUND) + // ErrInvalidFetchSessionEpoch Broker: The fetch session epoch is invalid + ErrInvalidFetchSessionEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_FETCH_SESSION_EPOCH) + // ErrListenerNotFound Broker: No matching listener + ErrListenerNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_LISTENER_NOT_FOUND) + // ErrTopicDeletionDisabled Broker: Topic deletion is disabled + ErrTopicDeletionDisabled ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_TOPIC_DELETION_DISABLED) + // ErrFencedLeaderEpoch Broker: Leader epoch is older than broker epoch + ErrFencedLeaderEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FENCED_LEADER_EPOCH) + // ErrUnknownLeaderEpoch Broker: Leader epoch is newer than broker epoch + ErrUnknownLeaderEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNKNOWN_LEADER_EPOCH) + // ErrUnsupportedCompressionType Broker: Unsupported compression type + ErrUnsupportedCompressionType ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSUPPORTED_COMPRESSION_TYPE) + // ErrStaleBrokerEpoch Broker: Broker epoch has changed + ErrStaleBrokerEpoch ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_STALE_BROKER_EPOCH) + // ErrOffsetNotAvailable Broker: Leader high watermark is not caught up + ErrOffsetNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_OFFSET_NOT_AVAILABLE) + // ErrMemberIDRequired Broker: Group member needs a valid member ID + ErrMemberIDRequired ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_MEMBER_ID_REQUIRED) + // ErrPreferredLeaderNotAvailable Broker: Preferred leader was not available + ErrPreferredLeaderNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_PREFERRED_LEADER_NOT_AVAILABLE) + // ErrGroupMaxSizeReached Broker: Consumer group has reached maximum size + ErrGroupMaxSizeReached ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_MAX_SIZE_REACHED) + // ErrFencedInstanceID Broker: Static consumer fenced by other consumer with same group.instance.id + ErrFencedInstanceID ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FENCED_INSTANCE_ID) + // ErrEligibleLeadersNotAvailable Broker: Eligible partition leaders are not available + ErrEligibleLeadersNotAvailable ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ELIGIBLE_LEADERS_NOT_AVAILABLE) + // ErrElectionNotNeeded Broker: Leader election not needed for topic partition + ErrElectionNotNeeded ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_ELECTION_NOT_NEEDED) + // ErrNoReassignmentInProgress Broker: No partition reassignment is in progress + ErrNoReassignmentInProgress ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_NO_REASSIGNMENT_IN_PROGRESS) + // ErrGroupSubscribedToTopic Broker: Deleting offsets of a topic while the consumer group is subscribed to it + ErrGroupSubscribedToTopic ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_GROUP_SUBSCRIBED_TO_TOPIC) + // ErrInvalidRecord Broker: Broker failed to validate record + ErrInvalidRecord ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_RECORD) + // ErrUnstableOffsetCommit Broker: There are unstable offsets that need to be cleared + ErrUnstableOffsetCommit ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNSTABLE_OFFSET_COMMIT) + // ErrThrottlingQuotaExceeded Broker: Throttling quota has been exceeded + ErrThrottlingQuotaExceeded ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_THROTTLING_QUOTA_EXCEEDED) + // ErrProducerFenced Broker: There is a newer producer with the same transactionalId which fences the current one + ErrProducerFenced ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_PRODUCER_FENCED) + // ErrResourceNotFound Broker: Request illegally referred to resource that does not exist + ErrResourceNotFound ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_RESOURCE_NOT_FOUND) + // ErrDuplicateResource Broker: Request illegally referred to the same resource twice + ErrDuplicateResource ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_DUPLICATE_RESOURCE) + // ErrUnacceptableCredential Broker: Requested credential would not meet criteria for acceptability + ErrUnacceptableCredential ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_UNACCEPTABLE_CREDENTIAL) + // ErrInconsistentVoterSet Broker: Indicates that the either the sender or recipient of a voter-only request is not one of the expected voters + ErrInconsistentVoterSet ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INCONSISTENT_VOTER_SET) + // ErrInvalidUpdateVersion Broker: Invalid update version + ErrInvalidUpdateVersion ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_INVALID_UPDATE_VERSION) + // ErrFeatureUpdateFailed Broker: Unable to update finalized features due to server error + ErrFeatureUpdateFailed ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_FEATURE_UPDATE_FAILED) + // ErrPrincipalDeserializationFailure Broker: Request principal deserialization failed during forwarding + ErrPrincipalDeserializationFailure ErrorCode = ErrorCode(C.RD_KAFKA_RESP_ERR_PRINCIPAL_DESERIALIZATION_FAILURE) +) diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/glue_rdkafka.h b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/glue_rdkafka.h new file mode 100644 index 0000000..bbff507 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/glue_rdkafka.h @@ -0,0 +1,48 @@ +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#pragma once + + +/** + * Glue between Go, Cgo and librdkafka + */ + + +/** + * Temporary C to Go header representation + */ +typedef struct tmphdr_s { + const char *key; + const void *val; // producer: malloc()ed by Go code if size > 0 + // consumer: owned by librdkafka + ssize_t size; +} tmphdr_t; + + + +/** + * @struct This is a glue struct used by the C code in this client to + * effectively map fields from a librdkafka rd_kafka_message_t + * to something usable in Go with as few CGo calls as possible. + */ +typedef struct glue_msg_s { + rd_kafka_message_t *msg; + rd_kafka_timestamp_type_t tstype; + int64_t ts; + tmphdr_t *tmphdrs; + size_t tmphdrsCnt; + int8_t want_hdrs; /**< If true, copy headers */ +} glue_msg_t; diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/handle.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/handle.go new file mode 100644 index 0000000..161a395 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/handle.go @@ -0,0 +1,379 @@ +package kafka + +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "fmt" + "strings" + "sync" + "time" + "unsafe" +) + +/* +#include "select_rdkafka.h" +#include +*/ +import "C" + +// OAuthBearerToken represents the data to be transmitted +// to a broker during SASL/OAUTHBEARER authentication. +type OAuthBearerToken struct { + // Token value, often (but not necessarily) a JWS compact serialization + // as per https://tools.ietf.org/html/rfc7515#section-3.1; it must meet + // the regular expression for a SASL/OAUTHBEARER value defined at + // https://tools.ietf.org/html/rfc7628#section-3.1 + TokenValue string + // Metadata about the token indicating when it expires (local time); + // it must represent a time in the future + Expiration time.Time + // Metadata about the token indicating the Kafka principal name + // to which it applies (for example, "admin") + Principal string + // SASL extensions, if any, to be communicated to the broker during + // authentication (all keys and values of which must meet the regular + // expressions defined at https://tools.ietf.org/html/rfc7628#section-3.1, + // and it must not contain the reserved "auth" key) + Extensions map[string]string +} + +// Handle represents a generic client handle containing common parts for +// both Producer and Consumer. +type Handle interface { + // SetOAuthBearerToken sets the the data to be transmitted + // to a broker during SASL/OAUTHBEARER authentication. It will return nil + // on success, otherwise an error if: + // 1) the token data is invalid (meaning an expiration time in the past + // or either a token value or an extension key or value that does not meet + // the regular expression requirements as per + // https://tools.ietf.org/html/rfc7628#section-3.1); + // 2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; + // 3) SASL/OAUTHBEARER is supported but is not configured as the client's + // authentication mechanism. + SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error + + // SetOAuthBearerTokenFailure sets the error message describing why token + // retrieval/setting failed; it also schedules a new token refresh event for 10 + // seconds later so the attempt may be retried. It will return nil on + // success, otherwise an error if: + // 1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; + // 2) SASL/OAUTHBEARER is supported but is not configured as the client's + // authentication mechanism. + SetOAuthBearerTokenFailure(errstr string) error + + // gethandle() returns the internal handle struct pointer + gethandle() *handle +} + +// Common instance handle for both Producer and Consumer +type handle struct { + rk *C.rd_kafka_t + rkq *C.rd_kafka_queue_t + + // Forward logs from librdkafka log queue to logs channel. + logs chan LogEvent + logq *C.rd_kafka_queue_t + closeLogsChan bool + + // Topic <-> rkt caches + rktCacheLock sync.Mutex + // topic name -> rkt cache + rktCache map[string]*C.rd_kafka_topic_t + // rkt -> topic name cache + rktNameCache map[*C.rd_kafka_topic_t]string + + // Cached instance name to avoid CGo call in String() + name string + + // + // cgo map + // Maps C callbacks based on cgoid back to its Go object + cgoLock sync.Mutex + cgoidNext uintptr + cgomap map[int]cgoif + + // + // producer + // + p *Producer + + // Forward delivery reports on Producer.Events channel + fwdDr bool + + // Enabled message fields for delivery reports and consumed messages. + msgFields *messageFields + + // + // consumer + // + c *Consumer + + // WaitGroup to wait for spawned go-routines to finish. + waitGroup sync.WaitGroup +} + +func (h *handle) String() string { + return h.name +} + +func (h *handle) setup() { + h.rktCache = make(map[string]*C.rd_kafka_topic_t) + h.rktNameCache = make(map[*C.rd_kafka_topic_t]string) + h.cgomap = make(map[int]cgoif) + h.name = C.GoString(C.rd_kafka_name(h.rk)) + if h.msgFields == nil { + h.msgFields = newMessageFields() + } +} + +func (h *handle) cleanup() { + if h.logs != nil { + C.rd_kafka_queue_destroy(h.logq) + if h.closeLogsChan { + close(h.logs) + } + } + + for _, crkt := range h.rktCache { + C.rd_kafka_topic_destroy(crkt) + } + + if h.rkq != nil { + C.rd_kafka_queue_destroy(h.rkq) + } +} + +func (h *handle) setupLogQueue(logsChan chan LogEvent, termChan chan bool) { + if logsChan == nil { + logsChan = make(chan LogEvent, 10000) + h.closeLogsChan = true + } + + h.logs = logsChan + + // Let librdkafka forward logs to our log queue instead of the main queue + h.logq = C.rd_kafka_queue_new(h.rk) + C.rd_kafka_set_log_queue(h.rk, h.logq) + + // Start a polling goroutine to consume the log queue + h.waitGroup.Add(1) + go func() { + h.pollLogEvents(h.logs, 100, termChan) + h.waitGroup.Done() + }() + +} + +// getRkt0 finds or creates and returns a C topic_t object from the local cache. +func (h *handle) getRkt0(topic string, ctopic *C.char, doLock bool) (crkt *C.rd_kafka_topic_t) { + if doLock { + h.rktCacheLock.Lock() + defer h.rktCacheLock.Unlock() + } + crkt, ok := h.rktCache[topic] + if ok { + return crkt + } + + if ctopic == nil { + ctopic = C.CString(topic) + defer C.free(unsafe.Pointer(ctopic)) + } + + crkt = C.rd_kafka_topic_new(h.rk, ctopic, nil) + if crkt == nil { + panic(fmt.Sprintf("Unable to create new C topic \"%s\": %s", + topic, C.GoString(C.rd_kafka_err2str(C.rd_kafka_last_error())))) + } + + h.rktCache[topic] = crkt + h.rktNameCache[crkt] = topic + + return crkt +} + +// getRkt finds or creates and returns a C topic_t object from the local cache. +func (h *handle) getRkt(topic string) (crkt *C.rd_kafka_topic_t) { + return h.getRkt0(topic, nil, true) +} + +// getTopicNameFromRkt returns the topic name for a C topic_t object, preferably +// using the local cache to avoid a cgo call. +func (h *handle) getTopicNameFromRkt(crkt *C.rd_kafka_topic_t) (topic string) { + h.rktCacheLock.Lock() + defer h.rktCacheLock.Unlock() + + topic, ok := h.rktNameCache[crkt] + if ok { + return topic + } + + // we need our own copy/refcount of the crkt + ctopic := C.rd_kafka_topic_name(crkt) + topic = C.GoString(ctopic) + + crkt = h.getRkt0(topic, ctopic, false /* dont lock */) + + return topic +} + +// cgoif is a generic interface for holding Go state passed as opaque +// value to the C code. +// Since pointers to complex Go types cannot be passed to C we instead create +// a cgoif object, generate a unique id that is added to the cgomap, +// and then pass that id to the C code. When the C code callback is called we +// use the id to look up the cgoif object in the cgomap. +type cgoif interface{} + +// delivery report cgoif container +type cgoDr struct { + deliveryChan chan Event + opaque interface{} +} + +// cgoPut adds object cg to the handle's cgo map and returns a +// unique id for the added entry. +// Thread-safe. +// FIXME: the uniquity of the id is questionable over time. +func (h *handle) cgoPut(cg cgoif) (cgoid int) { + h.cgoLock.Lock() + defer h.cgoLock.Unlock() + + h.cgoidNext++ + if h.cgoidNext == 0 { + h.cgoidNext++ + } + cgoid = (int)(h.cgoidNext) + h.cgomap[cgoid] = cg + return cgoid +} + +// cgoGet looks up cgoid in the cgo map, deletes the reference from the map +// and returns the object, if found. Else returns nil, false. +// Thread-safe. +func (h *handle) cgoGet(cgoid int) (cg cgoif, found bool) { + if cgoid == 0 { + return nil, false + } + + h.cgoLock.Lock() + defer h.cgoLock.Unlock() + cg, found = h.cgomap[cgoid] + if found { + delete(h.cgomap, cgoid) + } + + return cg, found +} + +// setOauthBearerToken - see rd_kafka_oauthbearer_set_token() +func (h *handle) setOAuthBearerToken(oauthBearerToken OAuthBearerToken) error { + cTokenValue := C.CString(oauthBearerToken.TokenValue) + defer C.free(unsafe.Pointer(cTokenValue)) + + cPrincipal := C.CString(oauthBearerToken.Principal) + defer C.free(unsafe.Pointer(cPrincipal)) + + cErrstrSize := C.size_t(512) + cErrstr := (*C.char)(C.malloc(cErrstrSize)) + defer C.free(unsafe.Pointer(cErrstr)) + + cExtensions := make([]*C.char, 2*len(oauthBearerToken.Extensions)) + extensionSize := 0 + for key, value := range oauthBearerToken.Extensions { + cExtensions[extensionSize] = C.CString(key) + defer C.free(unsafe.Pointer(cExtensions[extensionSize])) + extensionSize++ + cExtensions[extensionSize] = C.CString(value) + defer C.free(unsafe.Pointer(cExtensions[extensionSize])) + extensionSize++ + } + + var cExtensionsToUse **C.char + if extensionSize > 0 { + cExtensionsToUse = (**C.char)(unsafe.Pointer(&cExtensions[0])) + } + + cErr := C.rd_kafka_oauthbearer_set_token(h.rk, cTokenValue, + C.int64_t(oauthBearerToken.Expiration.UnixNano()/(1000*1000)), cPrincipal, + cExtensionsToUse, C.size_t(extensionSize), cErrstr, cErrstrSize) + if cErr == C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil + } + return newErrorFromCString(cErr, cErrstr) +} + +// setOauthBearerTokenFailure - see rd_kafka_oauthbearer_set_token_failure() +func (h *handle) setOAuthBearerTokenFailure(errstr string) error { + cerrstr := C.CString(errstr) + defer C.free(unsafe.Pointer(cerrstr)) + cErr := C.rd_kafka_oauthbearer_set_token_failure(h.rk, cerrstr) + if cErr == C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil + } + return newError(cErr) +} + +// messageFields controls which fields are made available for producer delivery reports & consumed messages. +// true values indicate that the field should be included +type messageFields struct { + Key bool + Value bool + Headers bool +} + +// disableAll disable all fields +func (mf *messageFields) disableAll() { + mf.Key = false + mf.Value = false + mf.Headers = false +} + +// newMessageFields returns a new messageFields with all fields enabled +func newMessageFields() *messageFields { + return &messageFields{ + Key: true, + Value: true, + Headers: true, + } +} + +// newMessageFieldsFrom constructs a new messageFields from the given configuration value +func newMessageFieldsFrom(v ConfigValue) (*messageFields, error) { + msgFields := newMessageFields() + switch v { + case "all": + // nothing to do + case "", "none": + msgFields.disableAll() + default: + msgFields.disableAll() + for _, value := range strings.Split(v.(string), ",") { + switch value { + case "key": + msgFields.Key = true + case "value": + msgFields.Value = true + case "headers": + msgFields.Headers = true + default: + return nil, fmt.Errorf("unknown message field: %s", value) + } + } + } + return msgFields, nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/header.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/header.go new file mode 100644 index 0000000..f7a73a4 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/header.go @@ -0,0 +1,67 @@ +package kafka + +/** + * Copyright 2018 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "fmt" + "strconv" +) + +/* +#include +#include "select_rdkafka.h" +#include "glue_rdkafka.h" +*/ +import "C" + +// Header represents a single Kafka message header. +// +// Message headers are made up of a list of Header elements, retaining their original insert +// order and allowing for duplicate Keys. +// +// Key is a human readable string identifying the header. +// Value is the key's binary value, Kafka does not put any restrictions on the format of +// of the Value but it should be made relatively compact. +// The value may be a byte array, empty, or nil. +// +// NOTE: Message headers are not available on producer delivery report messages. +type Header struct { + Key string // Header name (utf-8 string) + Value []byte // Header value (nil, empty, or binary) +} + +// String returns the Header Key and data in a human representable possibly truncated form +// suitable for displaying to the user. +func (h Header) String() string { + if h.Value == nil { + return fmt.Sprintf("%s=nil", h.Key) + } + + valueLen := len(h.Value) + if valueLen == 0 { + return fmt.Sprintf("%s=", h.Key) + } + + truncSize := valueLen + trunc := "" + if valueLen > 50+15 { + truncSize = 50 + trunc = fmt.Sprintf("(%d more bytes)", valueLen-truncSize) + } + + return fmt.Sprintf("%s=%s%s", h.Key, strconv.Quote(string(h.Value[:truncSize])), trunc) +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/kafka.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/kafka.go new file mode 100644 index 0000000..20dc30a --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/kafka.go @@ -0,0 +1,375 @@ +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +// Package kafka provides high-level Apache Kafka producer and consumers +// using bindings on-top of the librdkafka C library. +// +// +// High-level Consumer +// +// * Decide if you want to read messages and events by calling `.Poll()` or +// the deprecated option of using the `.Events()` channel. (If you want to use +// `.Events()` channel then set `"go.events.channel.enable": true`). +// +// * Create a Consumer with `kafka.NewConsumer()` providing at +// least the `bootstrap.servers` and `group.id` configuration properties. +// +// * Call `.Subscribe()` or (`.SubscribeTopics()` to subscribe to multiple topics) +// to join the group with the specified subscription set. +// Subscriptions are atomic, calling `.Subscribe*()` again will leave +// the group and rejoin with the new set of topics. +// +// * Start reading events and messages from either the `.Events` channel +// or by calling `.Poll()`. +// +// * When the group has rebalanced each client member is assigned a +// (sub-)set of topic+partitions. +// By default the consumer will start fetching messages for its assigned +// partitions at this point, but your application may enable rebalance +// events to get an insight into what the assigned partitions where +// as well as set the initial offsets. To do this you need to pass +// `"go.application.rebalance.enable": true` to the `NewConsumer()` call +// mentioned above. You will (eventually) see a `kafka.AssignedPartitions` event +// with the assigned partition set. You can optionally modify the initial +// offsets (they'll default to stored offsets and if there are no previously stored +// offsets it will fall back to `"auto.offset.reset"` +// which defaults to the `latest` message) and then call `.Assign(partitions)` +// to start consuming. If you don't need to modify the initial offsets you will +// not need to call `.Assign()`, the client will do so automatically for you if +// you dont, unless you are using the channel-based consumer in which case +// you MUST call `.Assign()` when receiving the `AssignedPartitions` and +// `RevokedPartitions` events. +// +// * As messages are fetched they will be made available on either the +// `.Events` channel or by calling `.Poll()`, look for event type `*kafka.Message`. +// +// * Handle messages, events and errors to your liking. +// +// * When you are done consuming call `.Close()` to commit final offsets +// and leave the consumer group. +// +// +// +// Producer +// +// * Create a Producer with `kafka.NewProducer()` providing at least +// the `bootstrap.servers` configuration properties. +// +// * Messages may now be produced either by sending a `*kafka.Message` +// on the `.ProduceChannel` or by calling `.Produce()`. +// +// * Producing is an asynchronous operation so the client notifies the application +// of per-message produce success or failure through something called delivery reports. +// Delivery reports are by default emitted on the `.Events()` channel as `*kafka.Message` +// and you should check `msg.TopicPartition.Error` for `nil` to find out if the message +// was succesfully delivered or not. +// It is also possible to direct delivery reports to alternate channels +// by providing a non-nil `chan Event` channel to `.Produce()`. +// If no delivery reports are wanted they can be completely disabled by +// setting configuration property `"go.delivery.reports": false`. +// +// * When you are done producing messages you will need to make sure all messages +// are indeed delivered to the broker (or failed), remember that this is +// an asynchronous client so some of your messages may be lingering in internal +// channels or tranmission queues. +// To do this you can either keep track of the messages you've produced +// and wait for their corresponding delivery reports, or call the convenience +// function `.Flush()` that will block until all message deliveries are done +// or the provided timeout elapses. +// +// * Finally call `.Close()` to decommission the producer. +// +// +// Transactional producer API +// +// The transactional producer operates on top of the idempotent producer, +// and provides full exactly-once semantics (EOS) for Apache Kafka when used +// with the transaction aware consumer (`isolation.level=read_committed`). +// +// A producer instance is configured for transactions by setting the +// `transactional.id` to an identifier unique for the application. This +// id will be used to fence stale transactions from previous instances of +// the application, typically following an outage or crash. +// +// After creating the transactional producer instance using `NewProducer()` +// the transactional state must be initialized by calling +// `InitTransactions()`. This is a blocking call that will +// acquire a runtime producer id from the transaction coordinator broker +// as well as abort any stale transactions and fence any still running producer +// instances with the same `transactional.id`. +// +// Once transactions are initialized the application may begin a new +// transaction by calling `BeginTransaction()`. +// A producer instance may only have one single on-going transaction. +// +// Any messages produced after the transaction has been started will +// belong to the ongoing transaction and will be committed or aborted +// atomically. +// It is not permitted to produce messages outside a transaction +// boundary, e.g., before `BeginTransaction()` or after `CommitTransaction()`, +// `AbortTransaction()` or if the current transaction has failed. +// +// If consumed messages are used as input to the transaction, the consumer +// instance must be configured with `enable.auto.commit` set to `false`. +// To commit the consumed offsets along with the transaction pass the +// list of consumed partitions and the last offset processed + 1 to +// `SendOffsetsToTransaction()` prior to committing the transaction. +// This allows an aborted transaction to be restarted using the previously +// committed offsets. +// +// To commit the produced messages, and any consumed offsets, to the +// current transaction, call `CommitTransaction()`. +// This call will block until the transaction has been fully committed or +// failed (typically due to fencing by a newer producer instance). +// +// Alternatively, if processing fails, or an abortable transaction error is +// raised, the transaction needs to be aborted by calling +// `AbortTransaction()` which marks any produced messages and +// offset commits as aborted. +// +// After the current transaction has been committed or aborted a new +// transaction may be started by calling `BeginTransaction()` again. +// +// Retriable errors: +// Some error cases allow the attempted operation to be retried, this is +// indicated by the error object having the retriable flag set which can +// be detected by calling `err.(kafka.Error).IsRetriable()`. +// When this flag is set the application may retry the operation immediately +// or preferably after a shorter grace period (to avoid busy-looping). +// Retriable errors include timeouts, broker transport failures, etc. +// +// Abortable errors: +// An ongoing transaction may fail permanently due to various errors, +// such as transaction coordinator becoming unavailable, write failures to the +// Apache Kafka log, under-replicated partitions, etc. +// At this point the producer application must abort the current transaction +// using `AbortTransaction()` and optionally start a new transaction +// by calling `BeginTransaction()`. +// Whether an error is abortable or not is detected by calling +// `err.(kafka.Error).TxnRequiresAbort()` on the returned error object. +// +// Fatal errors: +// While the underlying idempotent producer will typically only raise +// fatal errors for unrecoverable cluster errors where the idempotency +// guarantees can't be maintained, most of these are treated as abortable by +// the transactional producer since transactions may be aborted and retried +// in their entirety; +// The transactional producer on the other hand introduces a set of additional +// fatal errors which the application needs to handle by shutting down the +// producer and terminate. There is no way for a producer instance to recover +// from fatal errors. +// Whether an error is fatal or not is detected by calling +// `err.(kafka.Error).IsFatal()` on the returned error object or by checking +// the global `GetFatalError()`. +// +// Handling of other errors: +// For errors that have neither retriable, abortable or the fatal flag set +// it is not always obvious how to handle them. While some of these errors +// may be indicative of bugs in the application code, such as when +// an invalid parameter is passed to a method, other errors might originate +// from the broker and be passed thru as-is to the application. +// The general recommendation is to treat these errors, that have +// neither the retriable or abortable flags set, as fatal. +// +// Error handling example: +// retry: +// +// err := producer.CommitTransaction(...) +// if err == nil { +// return nil +// } else if err.(kafka.Error).TxnRequiresAbort() { +// do_abort_transaction_and_reset_inputs() +// } else if err.(kafka.Error).IsRetriable() { +// goto retry +// } else { // treat all other errors as fatal errors +// panic(err) +// } +// +// +// Events +// +// Apart from emitting messages and delivery reports the client also communicates +// with the application through a number of different event types. +// An application may choose to handle or ignore these events. +// +// Consumer events +// +// * `*kafka.Message` - a fetched message. +// +// * `AssignedPartitions` - The assigned partition set for this client following a rebalance. +// Requires `go.application.rebalance.enable` +// +// * `RevokedPartitions` - The counter part to `AssignedPartitions` following a rebalance. +// `AssignedPartitions` and `RevokedPartitions` are symmetrical. +// Requires `go.application.rebalance.enable` +// +// * `PartitionEOF` - Consumer has reached the end of a partition. +// NOTE: The consumer will keep trying to fetch new messages for the partition. +// +// * `OffsetsCommitted` - Offset commit results (when `enable.auto.commit` is enabled). +// +// +// Producer events +// +// * `*kafka.Message` - delivery report for produced message. +// Check `.TopicPartition.Error` for delivery result. +// +// +// Generic events for both Consumer and Producer +// +// * `KafkaError` - client (error codes are prefixed with _) or broker error. +// These errors are normally just informational since the +// client will try its best to automatically recover (eventually). +// +// * `OAuthBearerTokenRefresh` - retrieval of a new SASL/OAUTHBEARER token is required. +// This event only occurs with sasl.mechanism=OAUTHBEARER. +// Be sure to invoke SetOAuthBearerToken() on the Producer/Consumer/AdminClient +// instance when a successful token retrieval is completed, otherwise be sure to +// invoke SetOAuthBearerTokenFailure() to indicate that retrieval failed (or +// if setting the token failed, which could happen if an extension doesn't meet +// the required regular expression); invoking SetOAuthBearerTokenFailure() will +// schedule a new event for 10 seconds later so another retrieval can be attempted. +// +// +// Hint: If your application registers a signal notification +// (signal.Notify) makes sure the signals channel is buffered to avoid +// possible complications with blocking Poll() calls. +// +// Note: The Confluent Kafka Go client is safe for concurrent use. +package kafka + +import ( + "fmt" + // Make sure librdkafka_vendor/ sub-directory is included in vendor pulls. + _ "github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor" + "unsafe" +) + +/* +#include +#include +#include "select_rdkafka.h" + +static rd_kafka_topic_partition_t *_c_rdkafka_topic_partition_list_entry(rd_kafka_topic_partition_list_t *rktparlist, int idx) { + return idx < rktparlist->cnt ? &rktparlist->elems[idx] : NULL; +} +*/ +import "C" + +// PartitionAny represents any partition (for partitioning), +// or unspecified value (for all other cases) +const PartitionAny = int32(C.RD_KAFKA_PARTITION_UA) + +// TopicPartition is a generic placeholder for a Topic+Partition and optionally Offset. +type TopicPartition struct { + Topic *string + Partition int32 + Offset Offset + Metadata *string + Error error +} + +func (p TopicPartition) String() string { + topic := "" + if p.Topic != nil { + topic = *p.Topic + } + if p.Error != nil { + return fmt.Sprintf("%s[%d]@%s(%s)", + topic, p.Partition, p.Offset, p.Error) + } + return fmt.Sprintf("%s[%d]@%s", + topic, p.Partition, p.Offset) +} + +// TopicPartitions is a slice of TopicPartitions that also implements +// the sort interface +type TopicPartitions []TopicPartition + +func (tps TopicPartitions) Len() int { + return len(tps) +} + +func (tps TopicPartitions) Less(i, j int) bool { + if *tps[i].Topic < *tps[j].Topic { + return true + } else if *tps[i].Topic > *tps[j].Topic { + return false + } + return tps[i].Partition < tps[j].Partition +} + +func (tps TopicPartitions) Swap(i, j int) { + tps[i], tps[j] = tps[j], tps[i] +} + +// new_cparts_from_TopicPartitions creates a new C rd_kafka_topic_partition_list_t +// from a TopicPartition array. +func newCPartsFromTopicPartitions(partitions []TopicPartition) (cparts *C.rd_kafka_topic_partition_list_t) { + cparts = C.rd_kafka_topic_partition_list_new(C.int(len(partitions))) + for _, part := range partitions { + ctopic := C.CString(*part.Topic) + defer C.free(unsafe.Pointer(ctopic)) + rktpar := C.rd_kafka_topic_partition_list_add(cparts, ctopic, C.int32_t(part.Partition)) + rktpar.offset = C.int64_t(part.Offset) + + if part.Metadata != nil { + cmetadata := C.CString(*part.Metadata) + rktpar.metadata = unsafe.Pointer(cmetadata) + rktpar.metadata_size = C.size_t(len(*part.Metadata)) + } + } + + return cparts +} + +func setupTopicPartitionFromCrktpar(partition *TopicPartition, crktpar *C.rd_kafka_topic_partition_t) { + + topic := C.GoString(crktpar.topic) + partition.Topic = &topic + partition.Partition = int32(crktpar.partition) + partition.Offset = Offset(crktpar.offset) + if crktpar.metadata_size > 0 { + size := C.int(crktpar.metadata_size) + cstr := (*C.char)(unsafe.Pointer(crktpar.metadata)) + metadata := C.GoStringN(cstr, size) + partition.Metadata = &metadata + } + if crktpar.err != C.RD_KAFKA_RESP_ERR_NO_ERROR { + partition.Error = newError(crktpar.err) + } +} + +func newTopicPartitionsFromCparts(cparts *C.rd_kafka_topic_partition_list_t) (partitions []TopicPartition) { + + partcnt := int(cparts.cnt) + + partitions = make([]TopicPartition, partcnt) + for i := 0; i < partcnt; i++ { + crktpar := C._c_rdkafka_topic_partition_list_entry(cparts, C.int(i)) + setupTopicPartitionFromCrktpar(&partitions[i], crktpar) + } + + return partitions +} + +// LibraryVersion returns the underlying librdkafka library version as a +// (version_int, version_str) tuple. +func LibraryVersion() (int, string) { + ver := (int)(C.rd_kafka_version()) + verstr := C.GoString(C.rd_kafka_version_str()) + return ver, verstr +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/.gitignore b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/.gitignore new file mode 100644 index 0000000..a2f65a7 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/.gitignore @@ -0,0 +1,3 @@ +*.tar.gz +*.tgz +tmp* diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/LICENSES.txt b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/LICENSES.txt new file mode 100644 index 0000000..f2aa57d --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/LICENSES.txt @@ -0,0 +1,366 @@ +LICENSE +-------------------------------------------------------------- +librdkafka - Apache Kafka C driver library + +Copyright (c) 2012-2020, Magnus Edenhill +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + +1. Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. +2. Redistributions in binary form must reproduce the above copyright notice, + this list of conditions and the following disclaimer in the documentation + and/or other materials provided with the distribution. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + + +LICENSE.crc32c +-------------------------------------------------------------- +# For src/crc32c.c copied (with modifications) from +# http://stackoverflow.com/a/17646775/1821055 + +/* crc32c.c -- compute CRC-32C using the Intel crc32 instruction + * Copyright (C) 2013 Mark Adler + * Version 1.1 1 Aug 2013 Mark Adler + */ + +/* + This software is provided 'as-is', without any express or implied + warranty. In no event will the author be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + Mark Adler + madler@alumni.caltech.edu + */ + + +LICENSE.fnv1a +-------------------------------------------------------------- +parts of src/rdfnv1a.c: http://www.isthe.com/chongo/src/fnv/hash_32a.c + + +Please do not copyright this code. This code is in the public domain. + +LANDON CURT NOLL DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO +EVENT SHALL LANDON CURT NOLL BE LIABLE FOR ANY SPECIAL, INDIRECT OR +CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF +USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR +OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR +PERFORMANCE OF THIS SOFTWARE. + +By: + chongo /\oo/\ + http://www.isthe.com/chongo/ + +Share and Enjoy! :-) + + +LICENSE.hdrhistogram +-------------------------------------------------------------- +This license covers src/rdhdrhistogram.c which is a C port of +Coda Hale's Golang HdrHistogram https://github.com/codahale/hdrhistogram +at revision 3a0bb77429bd3a61596f5e8a3172445844342120 + +----------------------------------------------------------------------------- + +The MIT License (MIT) + +Copyright (c) 2014 Coda Hale + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in +all copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN +THE SOFTWARE + + +LICENSE.lz4 +-------------------------------------------------------------- +src/rdxxhash.[ch] src/lz4*.[ch]: git@github.com:lz4/lz4.git e2827775ee80d2ef985858727575df31fc60f1f3 + +LZ4 Library +Copyright (c) 2011-2016, Yann Collet +All rights reserved. + +Redistribution and use in source and binary forms, with or without modification, +are permitted provided that the following conditions are met: + +* Redistributions of source code must retain the above copyright notice, this + list of conditions and the following disclaimer. + +* Redistributions in binary form must reproduce the above copyright notice, this + list of conditions and the following disclaimer in the documentation and/or + other materials provided with the distribution. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND +ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED +WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE +DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR +ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES +(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; +LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON +ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + +LICENSE.murmur2 +-------------------------------------------------------------- +parts of src/rdmurmur2.c: git@github.com:abrandoned/murmur2.git + + +MurMurHash2 Library +//----------------------------------------------------------------------------- +// MurmurHash2 was written by Austin Appleby, and is placed in the public +// domain. The author hereby disclaims copyright to this source code. + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. + + +LICENSE.pycrc +-------------------------------------------------------------- +The following license applies to the files rdcrc32.c and rdcrc32.h which +have been generated by the pycrc tool. +============================================================================ + +Copyright (c) 2006-2012, Thomas Pircher + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in +all copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN +THE SOFTWARE. + + +LICENSE.queue +-------------------------------------------------------------- +For sys/queue.h: + + * Copyright (c) 1991, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)queue.h 8.5 (Berkeley) 8/20/94 + * $FreeBSD$ + +LICENSE.regexp +-------------------------------------------------------------- +regexp.c and regexp.h from https://github.com/ccxvii/minilibs sha 875c33568b5a4aa4fb3dd0c52ea98f7f0e5ca684 + +" +These libraries are in the public domain (or the equivalent where that is not possible). You can do anything you want with them. You have no legal obligation to do anything else, although I appreciate attribution. +" + + +LICENSE.snappy +-------------------------------------------------------------- +###################################################################### +# LICENSE.snappy covers files: snappy.c, snappy.h, snappy_compat.h # +# originally retrieved from http://github.com/andikleen/snappy-c # +# git revision 8015f2d28739b9a6076ebaa6c53fe27bc238d219 # +###################################################################### + +The snappy-c code is under the same license as the original snappy source + +Copyright 2011 Intel Corporation All Rights Reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Intel Corporation nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + + +LICENSE.tinycthread +-------------------------------------------------------------- +From https://github.com/tinycthread/tinycthread/README.txt c57166cd510ffb5022dd5f127489b131b61441b9 + +License +------- + +Copyright (c) 2012 Marcus Geelnard + 2013-2014 Evan Nemerson + +This software is provided 'as-is', without any express or implied +warranty. In no event will the authors be held liable for any damages +arising from the use of this software. + +Permission is granted to anyone to use this software for any purpose, +including commercial applications, and to alter it and redistribute it +freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + + 3. This notice may not be removed or altered from any source + distribution. + + +LICENSE.wingetopt +-------------------------------------------------------------- +For the files wingetopt.c wingetopt.h downloaded from https://github.com/alex85k/wingetopt + +/* + * Copyright (c) 2002 Todd C. Miller + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + * + * Sponsored in part by the Defense Advanced Research Projects + * Agency (DARPA) and Air Force Research Laboratory, Air Force + * Materiel Command, USAF, under agreement number F39502-99-1-0512. + */ +/*- + * Copyright (c) 2000 The NetBSD Foundation, Inc. + * All rights reserved. + * + * This code is derived from software contributed to The NetBSD Foundation + * by Dieter Baron and Thomas Klausner. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS + * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED + * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGE. + */ + + diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/README.md b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/README.md new file mode 100644 index 0000000..b13dfd2 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/README.md @@ -0,0 +1,24 @@ +# Bundling prebuilt librdkafka + +confluent-kafka-go bundles prebuilt statically linked +versions of librdkafka for the following platforms: + + * MacOSX x64 (aka Darwin) + * Linux glibc x64 (Ubuntu, CentOS, etc) + * Linux musl x64 (Alpine) + +## Import static librdkafka bundle + +First create the static librdkafka bundle following the instructions in +librdkafka's packaging/nuget/README.md. + +Then import the new version by using the import.sh script here, this script +will create a branch, import the bundle, create a commit and push the +branch to Github for PR review. This PR must be manually opened, reviewed +and then finally merged (make sure to merge it, DO NOT squash or rebase). + + $ ./import.sh ~/path/to/librdkafka-static-bundle-v1.4.0.tgz + +This will copy the static library and the rdkafka.h header file +to this directory, as well as generate a new ../build_..go file +for this platform + variant. diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/bundle-import.sh b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/bundle-import.sh new file mode 100644 index 0000000..2f7aeda --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/bundle-import.sh @@ -0,0 +1,113 @@ +#!/bin/bash +# +# Updates the bundled prebuilt librdkafka libraries to specified version. +# + +set -e + + +usage() { + echo "Usage: $0 librdkafka-static-bundle-.tgz" + echo "" + echo "This tool must be run from the TOPDIR/kafka/librdkafka_vendor directory" + exit 1 +} + + +parse_dynlibs() { + # Parse dynamic libraries from pkg-config file, + # both the ones specified with Libs: but also through Requires: + local pc=$1 + local libs= + local req= + local n= + for req in $(grep ^Requires: $pc | sed -e 's/^Requires://'); do + n=$(pkg-config --libs $req) + if [[ $n == -l* ]]; then + libs="${libs} $n" + fi + done + for n in $(grep ^Libs: $pc); do + if [[ $n == -l* ]]; then + libs="${libs} $n" + fi + done + + echo "$libs" +} + +setup_build() { + # Copies static library from the temp directory into final location, + # extracts dynamic lib list from the pkg-config file, + # and generates the build_..go file + local btype=$1 + local apath=$2 + local pc=$3 + local srcinfo=$4 + local build_tag= + local gpath="../build_${btype}.go" + local dpath="librdkafka_${btype}.a" + + if [[ $btype == glibc_linux ]]; then + build_tag="// +build !musl" + elif [[ $btype == musl_linux ]]; then + build_tag="// +build musl" + fi + + local dynlibs=$(parse_dynlibs $pc) + + echo "Copying $apath to $dpath" + cp "$apath" "$dpath" + + echo "Generating $gpath (extra build tag: $build_tag)" + + cat >$gpath <.tgz" + echo "" + echo "This tool must be run from the TOPDIR/kafka/librdkafka directory" + echo "" + echo "Options:" + echo " --devel - Development use: No branch checks and does not push to github" + exit 1 +} + +error_cleanup() { + echo "Error occurred, cleaning up" + git checkout $currbranch + git branch -D $import_branch + exit 1 +} + +devel=0 +if [[ $1 == --devel ]]; then + devel=1 + shift +fi + +bundle="$1" +[[ -f $bundle ]] || usage + +# Parse the librdkafka version from the bundle +bundlename=$(basename $bundle) +version=${bundlename#librdkafka-static-bundle-} +version=${version%.tgz} + +if [[ -z $version ]]; then + echo "Error: Could not parse version from bundle $bundle" + exit 1 +fi + +# Verify branch state +curr_branch=$(git symbolic-ref HEAD 2>/dev/null | cut -d"/" -f 3-) +uncommitted=$(git status --untracked-files=no --porcelain) + +if [[ $devel != 1 ]] && ( [[ $curr_branch != master ]] || [[ ! -z $uncommitted ]] ); then + echo "Error: This script must be run on an up-to-date, clean, master branch" + if [[ ! -z $uncommitted ]]; then + echo "Uncommitted files:" + echo "$uncommitted" + fi + exit 1 +fi + + +# Create import branch, import bundle, commit. +import_branch="import_$version" + +exists=$(git branch -rlq | grep "/$import_branch\$" || true) +if [[ ! -z $exists ]]; then + echo "Error: This version branch already seems to exist: $exists: already imorted?" + [[ $devel != 1 ]] && exit 1 +fi + +echo "Checking for existing commits that match this version (should be none)" +git log --oneline | grep "^librdkafka static bundle $version\$" && exit 1 + + +echo "Creating import branch $import_branch" +git checkout -b $import_branch + +echo "Importing bundle $bundle" +./bundle-import.sh "$bundle" || error_cleanup + +echo "Committing $version" +git commit -a -m "librdkafka static bundle $version" || error_cleanup + +echo "Updating error codes and docs" +pushd ../../ +make -f mk/Makefile docs || error_cleanup +git commit -a -m "Documentation and error code update for librdkafka $version" \ + || error_cleanup +popd + +if [[ $devel != 1 ]]; then + echo "Pushing branch" + git push origin $import_branch || error_cleanup +fi + +git checkout $curr_branch + +if [[ $devel != 1 ]]; then + git branch -D $import_branch +fi + +echo "" +echo "############## IMPORT OF $version COMPLETE ##############" +if [[ $devel != 1 ]]; then + echo "Branch $import_branch has been pushed." + echo "Create a PR, have it reviewed and then merge it (do NOT squash or rebase)." +fi + diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka.go new file mode 100644 index 0000000..52e1ad7 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka.go @@ -0,0 +1,21 @@ +/** + * Copyright 2020 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package librdkafka + +// LibrdkafkaGoSubdir is a dummy variable needed to export something so the +// file is not empty. +var LibrdkafkaGoSubdir = true diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_darwin.a b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_darwin.a new file mode 100644 index 0000000..6294180 Binary files /dev/null and b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_darwin.a differ diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_glibc_linux.a b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_glibc_linux.a new file mode 100644 index 0000000..c736b84 Binary files /dev/null and b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_glibc_linux.a differ diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_musl_linux.a b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_musl_linux.a new file mode 100644 index 0000000..f876505 Binary files /dev/null and b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_musl_linux.a differ diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_windows.a b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_windows.a new file mode 100644 index 0000000..7c5909f Binary files /dev/null and b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/librdkafka_windows.a differ diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/rdkafka.h b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/rdkafka.h new file mode 100644 index 0000000..b85ba90 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/librdkafka_vendor/rdkafka.h @@ -0,0 +1,7781 @@ +/* + * librdkafka - Apache Kafka C library + * + * Copyright (c) 2012-2020 Magnus Edenhill + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are met: + * + * 1. Redistributions of source code must retain the above copyright notice, + * this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright notice, + * this list of conditions and the following disclaimer in the documentation + * and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE + * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGE. + */ + +/** + * @file rdkafka.h + * @brief Apache Kafka C/C++ consumer and producer client library. + * + * rdkafka.h contains the public API for librdkafka. + * The API is documented in this file as comments prefixing the function, type, + * enum, define, etc. + * + * @sa For the C++ interface see rdkafkacpp.h + * + * @tableofcontents + */ + + +/* @cond NO_DOC */ +#ifndef _RDKAFKA_H_ +#define _RDKAFKA_H_ + +#include +#include +#include + +#ifdef __cplusplus +extern "C" { +#if 0 +} /* Restore indent */ +#endif +#endif + +#ifdef _WIN32 +#include +#ifndef WIN32_MEAN_AND_LEAN +#define WIN32_MEAN_AND_LEAN +#endif +#include /* for sockaddr, .. */ +#ifndef _SSIZE_T_DEFINED +#define _SSIZE_T_DEFINED +typedef SSIZE_T ssize_t; +#endif +#define RD_UNUSED +#define RD_INLINE __inline +#define RD_DEPRECATED __declspec(deprecated) +#define RD_FORMAT(...) +#undef RD_EXPORT +#ifdef LIBRDKAFKA_STATICLIB +#define RD_EXPORT +#else +#ifdef LIBRDKAFKA_EXPORTS +#define RD_EXPORT __declspec(dllexport) +#else +#define RD_EXPORT __declspec(dllimport) +#endif +#ifndef LIBRDKAFKA_TYPECHECKS +#define LIBRDKAFKA_TYPECHECKS 0 +#endif +#endif + +#else +#include /* for sockaddr, .. */ + +#define RD_UNUSED __attribute__((unused)) +#define RD_INLINE inline +#define RD_EXPORT +#define RD_DEPRECATED __attribute__((deprecated)) + +#if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__) +#define RD_FORMAT(...) __attribute__((format (__VA_ARGS__))) +#else +#define RD_FORMAT(...) +#endif + +#ifndef LIBRDKAFKA_TYPECHECKS +#define LIBRDKAFKA_TYPECHECKS 1 +#endif +#endif + + +/** + * @brief Type-checking macros + * Compile-time checking that \p ARG is of type \p TYPE. + * @returns \p RET + */ +#if LIBRDKAFKA_TYPECHECKS +#define _LRK_TYPECHECK(RET,TYPE,ARG) \ + ({ if (0) { TYPE __t RD_UNUSED = (ARG); } RET; }) + +#define _LRK_TYPECHECK2(RET,TYPE,ARG,TYPE2,ARG2) \ + ({ \ + if (0) { \ + TYPE __t RD_UNUSED = (ARG); \ + TYPE2 __t2 RD_UNUSED = (ARG2); \ + } \ + RET; }) + +#define _LRK_TYPECHECK3(RET,TYPE,ARG,TYPE2,ARG2,TYPE3,ARG3) \ + ({ \ + if (0) { \ + TYPE __t RD_UNUSED = (ARG); \ + TYPE2 __t2 RD_UNUSED = (ARG2); \ + TYPE3 __t3 RD_UNUSED = (ARG3); \ + } \ + RET; }) +#else +#define _LRK_TYPECHECK(RET,TYPE,ARG) (RET) +#define _LRK_TYPECHECK2(RET,TYPE,ARG,TYPE2,ARG2) (RET) +#define _LRK_TYPECHECK3(RET,TYPE,ARG,TYPE2,ARG2,TYPE3,ARG3) (RET) +#endif + +/* @endcond */ + + +/** + * @name librdkafka version + * @{ + * + * + */ + +/** + * @brief librdkafka version + * + * Interpreted as hex \c MM.mm.rr.xx: + * - MM = Major + * - mm = minor + * - rr = revision + * - xx = pre-release id (0xff is the final release) + * + * E.g.: \c 0x000801ff = 0.8.1 + * + * @remark This value should only be used during compile time, + * for runtime checks of version use rd_kafka_version() + */ +#define RD_KAFKA_VERSION 0x010802ff + +/** + * @brief Returns the librdkafka version as integer. + * + * @returns Version integer. + * + * @sa See RD_KAFKA_VERSION for how to parse the integer format. + * @sa Use rd_kafka_version_str() to retreive the version as a string. + */ +RD_EXPORT +int rd_kafka_version(void); + +/** + * @brief Returns the librdkafka version as string. + * + * @returns Version string + */ +RD_EXPORT +const char *rd_kafka_version_str (void); + +/**@}*/ + + +/** + * @name Constants, errors, types + * @{ + * + * + */ + + +/** + * @enum rd_kafka_type_t + * + * @brief rd_kafka_t handle type. + * + * @sa rd_kafka_new() + */ +typedef enum rd_kafka_type_t { + RD_KAFKA_PRODUCER, /**< Producer client */ + RD_KAFKA_CONSUMER /**< Consumer client */ +} rd_kafka_type_t; + + +/*! + * Timestamp types + * + * @sa rd_kafka_message_timestamp() + */ +typedef enum rd_kafka_timestamp_type_t { + RD_KAFKA_TIMESTAMP_NOT_AVAILABLE, /**< Timestamp not available */ + RD_KAFKA_TIMESTAMP_CREATE_TIME, /**< Message creation time */ + RD_KAFKA_TIMESTAMP_LOG_APPEND_TIME /**< Log append time */ +} rd_kafka_timestamp_type_t; + + + +/** + * @brief Retrieve supported debug contexts for use with the \c \"debug\" + * configuration property. (runtime) + * + * @returns Comma-separated list of available debugging contexts. + */ +RD_EXPORT +const char *rd_kafka_get_debug_contexts(void); + +/** + * @brief Supported debug contexts. (compile time) + * + * @deprecated This compile time value may be outdated at runtime due to + * linking another version of the library. + * Use rd_kafka_get_debug_contexts() instead. + */ +#define RD_KAFKA_DEBUG_CONTEXTS \ + "all,generic,broker,topic,metadata,feature,queue,msg,protocol,cgrp,security,fetch,interceptor,plugin,consumer,admin,eos,mock,assignor,conf" + + +/* @cond NO_DOC */ +/* Private types to provide ABI compatibility */ +typedef struct rd_kafka_s rd_kafka_t; +typedef struct rd_kafka_topic_s rd_kafka_topic_t; +typedef struct rd_kafka_conf_s rd_kafka_conf_t; +typedef struct rd_kafka_topic_conf_s rd_kafka_topic_conf_t; +typedef struct rd_kafka_queue_s rd_kafka_queue_t; +typedef struct rd_kafka_op_s rd_kafka_event_t; +typedef struct rd_kafka_topic_result_s rd_kafka_topic_result_t; +typedef struct rd_kafka_consumer_group_metadata_s +rd_kafka_consumer_group_metadata_t; +typedef struct rd_kafka_error_s rd_kafka_error_t; +typedef struct rd_kafka_headers_s rd_kafka_headers_t; +typedef struct rd_kafka_group_result_s rd_kafka_group_result_t; +/* @endcond */ + + +/** + * @enum rd_kafka_resp_err_t + * @brief Error codes. + * + * The negative error codes delimited by two underscores + * (\c RD_KAFKA_RESP_ERR__..) denotes errors internal to librdkafka and are + * displayed as \c \"Local: \\", while the error codes + * delimited by a single underscore (\c RD_KAFKA_RESP_ERR_..) denote broker + * errors and are displayed as \c \"Broker: \\". + * + * @sa Use rd_kafka_err2str() to translate an error code a human readable string + */ +typedef enum { + /* Internal errors to rdkafka: */ + /** Begin internal error codes */ + RD_KAFKA_RESP_ERR__BEGIN = -200, + /** Received message is incorrect */ + RD_KAFKA_RESP_ERR__BAD_MSG = -199, + /** Bad/unknown compression */ + RD_KAFKA_RESP_ERR__BAD_COMPRESSION = -198, + /** Broker is going away */ + RD_KAFKA_RESP_ERR__DESTROY = -197, + /** Generic failure */ + RD_KAFKA_RESP_ERR__FAIL = -196, + /** Broker transport failure */ + RD_KAFKA_RESP_ERR__TRANSPORT = -195, + /** Critical system resource */ + RD_KAFKA_RESP_ERR__CRIT_SYS_RESOURCE = -194, + /** Failed to resolve broker */ + RD_KAFKA_RESP_ERR__RESOLVE = -193, + /** Produced message timed out*/ + RD_KAFKA_RESP_ERR__MSG_TIMED_OUT = -192, + /** Reached the end of the topic+partition queue on + * the broker. Not really an error. + * This event is disabled by default, + * see the `enable.partition.eof` configuration property. */ + RD_KAFKA_RESP_ERR__PARTITION_EOF = -191, + /** Permanent: Partition does not exist in cluster. */ + RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION = -190, + /** File or filesystem error */ + RD_KAFKA_RESP_ERR__FS = -189, + /** Permanent: Topic does not exist in cluster. */ + RD_KAFKA_RESP_ERR__UNKNOWN_TOPIC = -188, + /** All broker connections are down. */ + RD_KAFKA_RESP_ERR__ALL_BROKERS_DOWN = -187, + /** Invalid argument, or invalid configuration */ + RD_KAFKA_RESP_ERR__INVALID_ARG = -186, + /** Operation timed out */ + RD_KAFKA_RESP_ERR__TIMED_OUT = -185, + /** Queue is full */ + RD_KAFKA_RESP_ERR__QUEUE_FULL = -184, + /** ISR count < required.acks */ + RD_KAFKA_RESP_ERR__ISR_INSUFF = -183, + /** Broker node update */ + RD_KAFKA_RESP_ERR__NODE_UPDATE = -182, + /** SSL error */ + RD_KAFKA_RESP_ERR__SSL = -181, + /** Waiting for coordinator to become available. */ + RD_KAFKA_RESP_ERR__WAIT_COORD = -180, + /** Unknown client group */ + RD_KAFKA_RESP_ERR__UNKNOWN_GROUP = -179, + /** Operation in progress */ + RD_KAFKA_RESP_ERR__IN_PROGRESS = -178, + /** Previous operation in progress, wait for it to finish. */ + RD_KAFKA_RESP_ERR__PREV_IN_PROGRESS = -177, + /** This operation would interfere with an existing subscription */ + RD_KAFKA_RESP_ERR__EXISTING_SUBSCRIPTION = -176, + /** Assigned partitions (rebalance_cb) */ + RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS = -175, + /** Revoked partitions (rebalance_cb) */ + RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS = -174, + /** Conflicting use */ + RD_KAFKA_RESP_ERR__CONFLICT = -173, + /** Wrong state */ + RD_KAFKA_RESP_ERR__STATE = -172, + /** Unknown protocol */ + RD_KAFKA_RESP_ERR__UNKNOWN_PROTOCOL = -171, + /** Not implemented */ + RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED = -170, + /** Authentication failure*/ + RD_KAFKA_RESP_ERR__AUTHENTICATION = -169, + /** No stored offset */ + RD_KAFKA_RESP_ERR__NO_OFFSET = -168, + /** Outdated */ + RD_KAFKA_RESP_ERR__OUTDATED = -167, + /** Timed out in queue */ + RD_KAFKA_RESP_ERR__TIMED_OUT_QUEUE = -166, + /** Feature not supported by broker */ + RD_KAFKA_RESP_ERR__UNSUPPORTED_FEATURE = -165, + /** Awaiting cache update */ + RD_KAFKA_RESP_ERR__WAIT_CACHE = -164, + /** Operation interrupted (e.g., due to yield)) */ + RD_KAFKA_RESP_ERR__INTR = -163, + /** Key serialization error */ + RD_KAFKA_RESP_ERR__KEY_SERIALIZATION = -162, + /** Value serialization error */ + RD_KAFKA_RESP_ERR__VALUE_SERIALIZATION = -161, + /** Key deserialization error */ + RD_KAFKA_RESP_ERR__KEY_DESERIALIZATION = -160, + /** Value deserialization error */ + RD_KAFKA_RESP_ERR__VALUE_DESERIALIZATION = -159, + /** Partial response */ + RD_KAFKA_RESP_ERR__PARTIAL = -158, + /** Modification attempted on read-only object */ + RD_KAFKA_RESP_ERR__READ_ONLY = -157, + /** No such entry / item not found */ + RD_KAFKA_RESP_ERR__NOENT = -156, + /** Read underflow */ + RD_KAFKA_RESP_ERR__UNDERFLOW = -155, + /** Invalid type */ + RD_KAFKA_RESP_ERR__INVALID_TYPE = -154, + /** Retry operation */ + RD_KAFKA_RESP_ERR__RETRY = -153, + /** Purged in queue */ + RD_KAFKA_RESP_ERR__PURGE_QUEUE = -152, + /** Purged in flight */ + RD_KAFKA_RESP_ERR__PURGE_INFLIGHT = -151, + /** Fatal error: see rd_kafka_fatal_error() */ + RD_KAFKA_RESP_ERR__FATAL = -150, + /** Inconsistent state */ + RD_KAFKA_RESP_ERR__INCONSISTENT = -149, + /** Gap-less ordering would not be guaranteed if proceeding */ + RD_KAFKA_RESP_ERR__GAPLESS_GUARANTEE = -148, + /** Maximum poll interval exceeded */ + RD_KAFKA_RESP_ERR__MAX_POLL_EXCEEDED = -147, + /** Unknown broker */ + RD_KAFKA_RESP_ERR__UNKNOWN_BROKER = -146, + /** Functionality not configured */ + RD_KAFKA_RESP_ERR__NOT_CONFIGURED = -145, + /** Instance has been fenced */ + RD_KAFKA_RESP_ERR__FENCED = -144, + /** Application generated error */ + RD_KAFKA_RESP_ERR__APPLICATION = -143, + /** Assignment lost */ + RD_KAFKA_RESP_ERR__ASSIGNMENT_LOST = -142, + /** No operation performed */ + RD_KAFKA_RESP_ERR__NOOP = -141, + /** No offset to automatically reset to */ + RD_KAFKA_RESP_ERR__AUTO_OFFSET_RESET = -140, + + /** End internal error codes */ + RD_KAFKA_RESP_ERR__END = -100, + + /* Kafka broker errors: */ + /** Unknown broker error */ + RD_KAFKA_RESP_ERR_UNKNOWN = -1, + /** Success */ + RD_KAFKA_RESP_ERR_NO_ERROR = 0, + /** Offset out of range */ + RD_KAFKA_RESP_ERR_OFFSET_OUT_OF_RANGE = 1, + /** Invalid message */ + RD_KAFKA_RESP_ERR_INVALID_MSG = 2, + /** Unknown topic or partition */ + RD_KAFKA_RESP_ERR_UNKNOWN_TOPIC_OR_PART = 3, + /** Invalid message size */ + RD_KAFKA_RESP_ERR_INVALID_MSG_SIZE = 4, + /** Leader not available */ + RD_KAFKA_RESP_ERR_LEADER_NOT_AVAILABLE = 5, + /** Not leader for partition */ + RD_KAFKA_RESP_ERR_NOT_LEADER_FOR_PARTITION = 6, + /** Request timed out */ + RD_KAFKA_RESP_ERR_REQUEST_TIMED_OUT = 7, + /** Broker not available */ + RD_KAFKA_RESP_ERR_BROKER_NOT_AVAILABLE = 8, + /** Replica not available */ + RD_KAFKA_RESP_ERR_REPLICA_NOT_AVAILABLE = 9, + /** Message size too large */ + RD_KAFKA_RESP_ERR_MSG_SIZE_TOO_LARGE = 10, + /** StaleControllerEpochCode */ + RD_KAFKA_RESP_ERR_STALE_CTRL_EPOCH = 11, + /** Offset metadata string too large */ + RD_KAFKA_RESP_ERR_OFFSET_METADATA_TOO_LARGE = 12, + /** Broker disconnected before response received */ + RD_KAFKA_RESP_ERR_NETWORK_EXCEPTION = 13, + /** Coordinator load in progress */ + RD_KAFKA_RESP_ERR_COORDINATOR_LOAD_IN_PROGRESS = 14, + /** Group coordinator load in progress */ +#define RD_KAFKA_RESP_ERR_GROUP_LOAD_IN_PROGRESS \ + RD_KAFKA_RESP_ERR_COORDINATOR_LOAD_IN_PROGRESS + /** Coordinator not available */ + RD_KAFKA_RESP_ERR_COORDINATOR_NOT_AVAILABLE = 15, + /** Group coordinator not available */ +#define RD_KAFKA_RESP_ERR_GROUP_COORDINATOR_NOT_AVAILABLE \ + RD_KAFKA_RESP_ERR_COORDINATOR_NOT_AVAILABLE + /** Not coordinator */ + RD_KAFKA_RESP_ERR_NOT_COORDINATOR = 16, + /** Not coordinator for group */ +#define RD_KAFKA_RESP_ERR_NOT_COORDINATOR_FOR_GROUP \ + RD_KAFKA_RESP_ERR_NOT_COORDINATOR + /** Invalid topic */ + RD_KAFKA_RESP_ERR_TOPIC_EXCEPTION = 17, + /** Message batch larger than configured server segment size */ + RD_KAFKA_RESP_ERR_RECORD_LIST_TOO_LARGE = 18, + /** Not enough in-sync replicas */ + RD_KAFKA_RESP_ERR_NOT_ENOUGH_REPLICAS = 19, + /** Message(s) written to insufficient number of in-sync replicas */ + RD_KAFKA_RESP_ERR_NOT_ENOUGH_REPLICAS_AFTER_APPEND = 20, + /** Invalid required acks value */ + RD_KAFKA_RESP_ERR_INVALID_REQUIRED_ACKS = 21, + /** Specified group generation id is not valid */ + RD_KAFKA_RESP_ERR_ILLEGAL_GENERATION = 22, + /** Inconsistent group protocol */ + RD_KAFKA_RESP_ERR_INCONSISTENT_GROUP_PROTOCOL = 23, + /** Invalid group.id */ + RD_KAFKA_RESP_ERR_INVALID_GROUP_ID = 24, + /** Unknown member */ + RD_KAFKA_RESP_ERR_UNKNOWN_MEMBER_ID = 25, + /** Invalid session timeout */ + RD_KAFKA_RESP_ERR_INVALID_SESSION_TIMEOUT = 26, + /** Group rebalance in progress */ + RD_KAFKA_RESP_ERR_REBALANCE_IN_PROGRESS = 27, + /** Commit offset data size is not valid */ + RD_KAFKA_RESP_ERR_INVALID_COMMIT_OFFSET_SIZE = 28, + /** Topic authorization failed */ + RD_KAFKA_RESP_ERR_TOPIC_AUTHORIZATION_FAILED = 29, + /** Group authorization failed */ + RD_KAFKA_RESP_ERR_GROUP_AUTHORIZATION_FAILED = 30, + /** Cluster authorization failed */ + RD_KAFKA_RESP_ERR_CLUSTER_AUTHORIZATION_FAILED = 31, + /** Invalid timestamp */ + RD_KAFKA_RESP_ERR_INVALID_TIMESTAMP = 32, + /** Unsupported SASL mechanism */ + RD_KAFKA_RESP_ERR_UNSUPPORTED_SASL_MECHANISM = 33, + /** Illegal SASL state */ + RD_KAFKA_RESP_ERR_ILLEGAL_SASL_STATE = 34, + /** Unuspported version */ + RD_KAFKA_RESP_ERR_UNSUPPORTED_VERSION = 35, + /** Topic already exists */ + RD_KAFKA_RESP_ERR_TOPIC_ALREADY_EXISTS = 36, + /** Invalid number of partitions */ + RD_KAFKA_RESP_ERR_INVALID_PARTITIONS = 37, + /** Invalid replication factor */ + RD_KAFKA_RESP_ERR_INVALID_REPLICATION_FACTOR = 38, + /** Invalid replica assignment */ + RD_KAFKA_RESP_ERR_INVALID_REPLICA_ASSIGNMENT = 39, + /** Invalid config */ + RD_KAFKA_RESP_ERR_INVALID_CONFIG = 40, + /** Not controller for cluster */ + RD_KAFKA_RESP_ERR_NOT_CONTROLLER = 41, + /** Invalid request */ + RD_KAFKA_RESP_ERR_INVALID_REQUEST = 42, + /** Message format on broker does not support request */ + RD_KAFKA_RESP_ERR_UNSUPPORTED_FOR_MESSAGE_FORMAT = 43, + /** Policy violation */ + RD_KAFKA_RESP_ERR_POLICY_VIOLATION = 44, + /** Broker received an out of order sequence number */ + RD_KAFKA_RESP_ERR_OUT_OF_ORDER_SEQUENCE_NUMBER = 45, + /** Broker received a duplicate sequence number */ + RD_KAFKA_RESP_ERR_DUPLICATE_SEQUENCE_NUMBER = 46, + /** Producer attempted an operation with an old epoch */ + RD_KAFKA_RESP_ERR_INVALID_PRODUCER_EPOCH = 47, + /** Producer attempted a transactional operation in an invalid state */ + RD_KAFKA_RESP_ERR_INVALID_TXN_STATE = 48, + /** Producer attempted to use a producer id which is not + * currently assigned to its transactional id */ + RD_KAFKA_RESP_ERR_INVALID_PRODUCER_ID_MAPPING = 49, + /** Transaction timeout is larger than the maximum + * value allowed by the broker's max.transaction.timeout.ms */ + RD_KAFKA_RESP_ERR_INVALID_TRANSACTION_TIMEOUT = 50, + /** Producer attempted to update a transaction while another + * concurrent operation on the same transaction was ongoing */ + RD_KAFKA_RESP_ERR_CONCURRENT_TRANSACTIONS = 51, + /** Indicates that the transaction coordinator sending a + * WriteTxnMarker is no longer the current coordinator for a + * given producer */ + RD_KAFKA_RESP_ERR_TRANSACTION_COORDINATOR_FENCED = 52, + /** Transactional Id authorization failed */ + RD_KAFKA_RESP_ERR_TRANSACTIONAL_ID_AUTHORIZATION_FAILED = 53, + /** Security features are disabled */ + RD_KAFKA_RESP_ERR_SECURITY_DISABLED = 54, + /** Operation not attempted */ + RD_KAFKA_RESP_ERR_OPERATION_NOT_ATTEMPTED = 55, + /** Disk error when trying to access log file on the disk */ + RD_KAFKA_RESP_ERR_KAFKA_STORAGE_ERROR = 56, + /** The user-specified log directory is not found in the broker config */ + RD_KAFKA_RESP_ERR_LOG_DIR_NOT_FOUND = 57, + /** SASL Authentication failed */ + RD_KAFKA_RESP_ERR_SASL_AUTHENTICATION_FAILED = 58, + /** Unknown Producer Id */ + RD_KAFKA_RESP_ERR_UNKNOWN_PRODUCER_ID = 59, + /** Partition reassignment is in progress */ + RD_KAFKA_RESP_ERR_REASSIGNMENT_IN_PROGRESS = 60, + /** Delegation Token feature is not enabled */ + RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_AUTH_DISABLED = 61, + /** Delegation Token is not found on server */ + RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_NOT_FOUND = 62, + /** Specified Principal is not valid Owner/Renewer */ + RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_OWNER_MISMATCH = 63, + /** Delegation Token requests are not allowed on this connection */ + RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_REQUEST_NOT_ALLOWED = 64, + /** Delegation Token authorization failed */ + RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_AUTHORIZATION_FAILED = 65, + /** Delegation Token is expired */ + RD_KAFKA_RESP_ERR_DELEGATION_TOKEN_EXPIRED = 66, + /** Supplied principalType is not supported */ + RD_KAFKA_RESP_ERR_INVALID_PRINCIPAL_TYPE = 67, + /** The group is not empty */ + RD_KAFKA_RESP_ERR_NON_EMPTY_GROUP = 68, + /** The group id does not exist */ + RD_KAFKA_RESP_ERR_GROUP_ID_NOT_FOUND = 69, + /** The fetch session ID was not found */ + RD_KAFKA_RESP_ERR_FETCH_SESSION_ID_NOT_FOUND = 70, + /** The fetch session epoch is invalid */ + RD_KAFKA_RESP_ERR_INVALID_FETCH_SESSION_EPOCH = 71, + /** No matching listener */ + RD_KAFKA_RESP_ERR_LISTENER_NOT_FOUND = 72, + /** Topic deletion is disabled */ + RD_KAFKA_RESP_ERR_TOPIC_DELETION_DISABLED = 73, + /** Leader epoch is older than broker epoch */ + RD_KAFKA_RESP_ERR_FENCED_LEADER_EPOCH = 74, + /** Leader epoch is newer than broker epoch */ + RD_KAFKA_RESP_ERR_UNKNOWN_LEADER_EPOCH = 75, + /** Unsupported compression type */ + RD_KAFKA_RESP_ERR_UNSUPPORTED_COMPRESSION_TYPE = 76, + /** Broker epoch has changed */ + RD_KAFKA_RESP_ERR_STALE_BROKER_EPOCH = 77, + /** Leader high watermark is not caught up */ + RD_KAFKA_RESP_ERR_OFFSET_NOT_AVAILABLE = 78, + /** Group member needs a valid member ID */ + RD_KAFKA_RESP_ERR_MEMBER_ID_REQUIRED = 79, + /** Preferred leader was not available */ + RD_KAFKA_RESP_ERR_PREFERRED_LEADER_NOT_AVAILABLE = 80, + /** Consumer group has reached maximum size */ + RD_KAFKA_RESP_ERR_GROUP_MAX_SIZE_REACHED = 81, + /** Static consumer fenced by other consumer with same + * group.instance.id. */ + RD_KAFKA_RESP_ERR_FENCED_INSTANCE_ID = 82, + /** Eligible partition leaders are not available */ + RD_KAFKA_RESP_ERR_ELIGIBLE_LEADERS_NOT_AVAILABLE = 83, + /** Leader election not needed for topic partition */ + RD_KAFKA_RESP_ERR_ELECTION_NOT_NEEDED = 84, + /** No partition reassignment is in progress */ + RD_KAFKA_RESP_ERR_NO_REASSIGNMENT_IN_PROGRESS = 85, + /** Deleting offsets of a topic while the consumer group is + * subscribed to it */ + RD_KAFKA_RESP_ERR_GROUP_SUBSCRIBED_TO_TOPIC = 86, + /** Broker failed to validate record */ + RD_KAFKA_RESP_ERR_INVALID_RECORD = 87, + /** There are unstable offsets that need to be cleared */ + RD_KAFKA_RESP_ERR_UNSTABLE_OFFSET_COMMIT = 88, + /** Throttling quota has been exceeded */ + RD_KAFKA_RESP_ERR_THROTTLING_QUOTA_EXCEEDED = 89, + /** There is a newer producer with the same transactionalId + * which fences the current one */ + RD_KAFKA_RESP_ERR_PRODUCER_FENCED = 90, + /** Request illegally referred to resource that does not exist */ + RD_KAFKA_RESP_ERR_RESOURCE_NOT_FOUND = 91, + /** Request illegally referred to the same resource twice */ + RD_KAFKA_RESP_ERR_DUPLICATE_RESOURCE = 92, + /** Requested credential would not meet criteria for acceptability */ + RD_KAFKA_RESP_ERR_UNACCEPTABLE_CREDENTIAL = 93, + /** Indicates that the either the sender or recipient of a + * voter-only request is not one of the expected voters */ + RD_KAFKA_RESP_ERR_INCONSISTENT_VOTER_SET = 94, + /** Invalid update version */ + RD_KAFKA_RESP_ERR_INVALID_UPDATE_VERSION = 95, + /** Unable to update finalized features due to server error */ + RD_KAFKA_RESP_ERR_FEATURE_UPDATE_FAILED = 96, + /** Request principal deserialization failed during forwarding */ + RD_KAFKA_RESP_ERR_PRINCIPAL_DESERIALIZATION_FAILURE = 97, + + RD_KAFKA_RESP_ERR_END_ALL, +} rd_kafka_resp_err_t; + + +/** + * @brief Error code value, name and description. + * Typically for use with language bindings to automatically expose + * the full set of librdkafka error codes. + */ +struct rd_kafka_err_desc { + rd_kafka_resp_err_t code;/**< Error code */ + const char *name; /**< Error name, same as code enum sans prefix */ + const char *desc; /**< Human readable error description. */ +}; + + +/** + * @brief Returns the full list of error codes. + */ +RD_EXPORT +void rd_kafka_get_err_descs (const struct rd_kafka_err_desc **errdescs, + size_t *cntp); + + + + +/** + * @brief Returns a human readable representation of a kafka error. + * + * @param err Error code to translate + */ +RD_EXPORT +const char *rd_kafka_err2str (rd_kafka_resp_err_t err); + + + +/** + * @brief Returns the error code name (enum name). + * + * @param err Error code to translate + */ +RD_EXPORT +const char *rd_kafka_err2name (rd_kafka_resp_err_t err); + + +/** + * @brief Returns the last error code generated by a legacy API call + * in the current thread. + * + * The legacy APIs are the ones using errno to propagate error value, namely: + * - rd_kafka_topic_new() + * - rd_kafka_consume_start() + * - rd_kafka_consume_stop() + * - rd_kafka_consume() + * - rd_kafka_consume_batch() + * - rd_kafka_consume_callback() + * - rd_kafka_consume_queue() + * - rd_kafka_produce() + * + * The main use for this function is to avoid converting system \p errno + * values to rd_kafka_resp_err_t codes for legacy APIs. + * + * @remark The last error is stored per-thread, if multiple rd_kafka_t handles + * are used in the same application thread the developer needs to + * make sure rd_kafka_last_error() is called immediately after + * a failed API call. + * + * @remark errno propagation from librdkafka is not safe on Windows + * and should not be used, use rd_kafka_last_error() instead. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_last_error (void); + + +/** + * @brief Converts the system errno value \p errnox to a rd_kafka_resp_err_t + * error code upon failure from the following functions: + * - rd_kafka_topic_new() + * - rd_kafka_consume_start() + * - rd_kafka_consume_stop() + * - rd_kafka_consume() + * - rd_kafka_consume_batch() + * - rd_kafka_consume_callback() + * - rd_kafka_consume_queue() + * - rd_kafka_produce() + * + * @param errnox System errno value to convert + * + * @returns Appropriate error code for \p errnox + * + * @remark A better alternative is to call rd_kafka_last_error() immediately + * after any of the above functions return -1 or NULL. + * + * @deprecated Use rd_kafka_last_error() to retrieve the last error code + * set by the legacy librdkafka APIs. + * + * @sa rd_kafka_last_error() + */ +RD_EXPORT RD_DEPRECATED +rd_kafka_resp_err_t rd_kafka_errno2err(int errnox); + + +/** + * @brief Returns the thread-local system errno + * + * On most platforms this is the same as \p errno but in case of different + * runtimes between library and application (e.g., Windows static DLLs) + * this provides a means for exposing the errno librdkafka uses. + * + * @remark The value is local to the current calling thread. + * + * @deprecated Use rd_kafka_last_error() to retrieve the last error code + * set by the legacy librdkafka APIs. + */ +RD_EXPORT RD_DEPRECATED +int rd_kafka_errno (void); + + + + +/** + * @brief Returns the first fatal error set on this client instance, + * or RD_KAFKA_RESP_ERR_NO_ERROR if no fatal error has occurred. + * + * This function is to be used with the Idempotent Producer and \c error_cb + * to detect fatal errors. + * + * Generally all errors raised by \c error_cb are to be considered + * informational and temporary, the client will try to recover from all + * errors in a graceful fashion (by retrying, etc). + * + * However, some errors should logically be considered fatal to retain + * consistency; in particular a set of errors that may occur when using the + * Idempotent Producer and the in-order or exactly-once producer guarantees + * can't be satisfied. + * + * @param rk Client instance. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written to if there is a fatal error. + * @param errstr_size Writable size in \p errstr. + * + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR if no fatal error has been raised, else + * any other error code. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_fatal_error (rd_kafka_t *rk, + char *errstr, size_t errstr_size); + + +/** + * @brief Trigger a fatal error for testing purposes. + * + * Since there is no practical way to trigger real fatal errors in the + * idempotent producer, this method allows an application to trigger + * fabricated fatal errors in tests to check its error handling code. + * + * @param rk Client instance. + * @param err The underlying error code. + * @param reason A human readable error reason. + * Will be prefixed with "test_fatal_error: " to differentiate + * from real fatal errors. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR if a fatal error was triggered, or + * RD_KAFKA_RESP_ERR__PREV_IN_PROGRESS if a previous fatal error + * has already been triggered. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_test_fatal_error (rd_kafka_t *rk, rd_kafka_resp_err_t err, + const char *reason); + + +/** + * @returns the error code for \p error or RD_KAFKA_RESP_ERR_NO_ERROR if + * \p error is NULL. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_error_code (const rd_kafka_error_t *error); + +/** + * @returns the error code name for \p error, e.g, "ERR_UNKNOWN_MEMBER_ID", + * or an empty string if \p error is NULL. + * + * @remark The lifetime of the returned pointer is the same as the error object. + * + * @sa rd_kafka_err2name() + */ +RD_EXPORT +const char *rd_kafka_error_name (const rd_kafka_error_t *error); + +/** + * @returns a human readable error string for \p error, + * or an empty string if \p error is NULL. + * + * @remark The lifetime of the returned pointer is the same as the error object. + */ +RD_EXPORT +const char *rd_kafka_error_string (const rd_kafka_error_t *error); + + +/** + * @returns 1 if the error is a fatal error, indicating that the client + * instance is no longer usable, else 0 (also if \p error is NULL). + */ +RD_EXPORT +int rd_kafka_error_is_fatal (const rd_kafka_error_t *error); + + +/** + * @returns 1 if the operation may be retried, + * else 0 (also if \p error is NULL). + */ +RD_EXPORT +int rd_kafka_error_is_retriable (const rd_kafka_error_t *error); + + +/** + * @returns 1 if the error is an abortable transaction error in which case + * the application must call rd_kafka_abort_transaction() and + * start a new transaction with rd_kafka_begin_transaction() if it + * wishes to proceed with transactions. + * Else returns 0 (also if \p error is NULL). + * + * @remark The return value of this method is only valid for errors returned + * by the transactional API. + */ +RD_EXPORT +int rd_kafka_error_txn_requires_abort (const rd_kafka_error_t *error); + +/** + * @brief Free and destroy an error object. + * + * @remark As a conveniance it is permitted to pass a NULL \p error. + */ +RD_EXPORT +void rd_kafka_error_destroy (rd_kafka_error_t *error); + + +/** + * @brief Create a new error object with error \p code and optional + * human readable error string in \p fmt. + * + * This method is mainly to be used for mocking errors in application test code. + * + * The returned object must be destroyed with rd_kafka_error_destroy(). + */ +RD_EXPORT +rd_kafka_error_t *rd_kafka_error_new (rd_kafka_resp_err_t code, + const char *fmt, ...) + RD_FORMAT(printf, 2, 3); + + +/** + * @brief Topic+Partition place holder + * + * Generic place holder for a Topic+Partition and its related information + * used for multiple purposes: + * - consumer offset (see rd_kafka_commit(), et.al.) + * - group rebalancing callback (rd_kafka_conf_set_rebalance_cb()) + * - offset commit result callback (rd_kafka_conf_set_offset_commit_cb()) + */ + +/** + * @brief Generic place holder for a specific Topic+Partition. + * + * @sa rd_kafka_topic_partition_list_new() + */ +typedef struct rd_kafka_topic_partition_s { + char *topic; /**< Topic name */ + int32_t partition; /**< Partition */ + int64_t offset; /**< Offset */ + void *metadata; /**< Metadata */ + size_t metadata_size; /**< Metadata size */ + void *opaque; /**< Opaque value for application use */ + rd_kafka_resp_err_t err; /**< Error code, depending on use. */ + void *_private; /**< INTERNAL USE ONLY, + * INITIALIZE TO ZERO, DO NOT TOUCH */ +} rd_kafka_topic_partition_t; + + +/** + * @brief Destroy a rd_kafka_topic_partition_t. + * @remark This must not be called for elements in a topic partition list. + */ +RD_EXPORT +void rd_kafka_topic_partition_destroy (rd_kafka_topic_partition_t *rktpar); + + +/** + * @brief A growable list of Topic+Partitions. + * + */ +typedef struct rd_kafka_topic_partition_list_s { + int cnt; /**< Current number of elements */ + int size; /**< Current allocated size */ + rd_kafka_topic_partition_t *elems; /**< Element array[] */ +} rd_kafka_topic_partition_list_t; + + +/** + * @brief Create a new list/vector Topic+Partition container. + * + * @param size Initial allocated size used when the expected number of + * elements is known or can be estimated. + * Avoids reallocation and possibly relocation of the + * elems array. + * + * @returns A newly allocated Topic+Partition list. + * + * @remark Use rd_kafka_topic_partition_list_destroy() to free all resources + * in use by a list and the list itself. + * @sa rd_kafka_topic_partition_list_add() + */ +RD_EXPORT +rd_kafka_topic_partition_list_t *rd_kafka_topic_partition_list_new (int size); + + +/** + * @brief Free all resources used by the list and the list itself. + */ +RD_EXPORT +void +rd_kafka_topic_partition_list_destroy (rd_kafka_topic_partition_list_t *rkparlist); + +/** + * @brief Add topic+partition to list + * + * @param rktparlist List to extend + * @param topic Topic name (copied) + * @param partition Partition id + * + * @returns The object which can be used to fill in additionals fields. + */ +RD_EXPORT +rd_kafka_topic_partition_t * +rd_kafka_topic_partition_list_add (rd_kafka_topic_partition_list_t *rktparlist, + const char *topic, int32_t partition); + + +/** + * @brief Add range of partitions from \p start to \p stop inclusive. + * + * @param rktparlist List to extend + * @param topic Topic name (copied) + * @param start Start partition of range + * @param stop Last partition of range (inclusive) + */ +RD_EXPORT +void +rd_kafka_topic_partition_list_add_range (rd_kafka_topic_partition_list_t + *rktparlist, + const char *topic, + int32_t start, int32_t stop); + + + +/** + * @brief Delete partition from list. + * + * @param rktparlist List to modify + * @param topic Topic name to match + * @param partition Partition to match + * + * @returns 1 if partition was found (and removed), else 0. + * + * @remark Any held indices to elems[] are unusable after this call returns 1. + */ +RD_EXPORT +int +rd_kafka_topic_partition_list_del (rd_kafka_topic_partition_list_t *rktparlist, + const char *topic, int32_t partition); + + +/** + * @brief Delete partition from list by elems[] index. + * + * @returns 1 if partition was found (and removed), else 0. + * + * @sa rd_kafka_topic_partition_list_del() + */ +RD_EXPORT +int +rd_kafka_topic_partition_list_del_by_idx ( + rd_kafka_topic_partition_list_t *rktparlist, + int idx); + + +/** + * @brief Make a copy of an existing list. + * + * @param src The existing list to copy. + * + * @returns A new list fully populated to be identical to \p src + */ +RD_EXPORT +rd_kafka_topic_partition_list_t * +rd_kafka_topic_partition_list_copy (const rd_kafka_topic_partition_list_t *src); + + + + +/** + * @brief Set offset to \p offset for \p topic and \p partition + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or + * RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION if \p partition was not found + * in the list. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_topic_partition_list_set_offset ( + rd_kafka_topic_partition_list_t *rktparlist, + const char *topic, int32_t partition, int64_t offset); + + + +/** + * @brief Find element by \p topic and \p partition. + * + * @returns a pointer to the first matching element, or NULL if not found. + */ +RD_EXPORT +rd_kafka_topic_partition_t * +rd_kafka_topic_partition_list_find ( + const rd_kafka_topic_partition_list_t *rktparlist, + const char *topic, int32_t partition); + + +/** + * @brief Sort list using comparator \p cmp. + * + * If \p cmp is NULL the default comparator will be used that + * sorts by ascending topic name and partition. + * + * \p cmp_opaque is provided as the \p cmp_opaque argument to \p cmp. + * + */ +RD_EXPORT void +rd_kafka_topic_partition_list_sort (rd_kafka_topic_partition_list_t *rktparlist, + int (*cmp) (const void *a, const void *b, + void *cmp_opaque), + void *cmp_opaque); + + +/**@}*/ + + + +/** + * @name Var-arg tag types + * @{ + * + */ + +/** + * @enum rd_kafka_vtype_t + * + * @brief Var-arg tag types + * + * @sa rd_kafka_producev() + */ +typedef enum rd_kafka_vtype_t { + RD_KAFKA_VTYPE_END, /**< va-arg sentinel */ + RD_KAFKA_VTYPE_TOPIC, /**< (const char *) Topic name */ + RD_KAFKA_VTYPE_RKT, /**< (rd_kafka_topic_t *) Topic handle */ + RD_KAFKA_VTYPE_PARTITION, /**< (int32_t) Partition */ + RD_KAFKA_VTYPE_VALUE, /**< (void *, size_t) Message value (payload)*/ + RD_KAFKA_VTYPE_KEY, /**< (void *, size_t) Message key */ + RD_KAFKA_VTYPE_OPAQUE, /**< (void *) Per-message application opaque + * value. This is the same as + * the _private field in + * rd_kafka_message_t, also known + * as the msg_opaque. */ + RD_KAFKA_VTYPE_MSGFLAGS, /**< (int) RD_KAFKA_MSG_F_.. flags */ + RD_KAFKA_VTYPE_TIMESTAMP, /**< (int64_t) Milliseconds since epoch UTC */ + RD_KAFKA_VTYPE_HEADER, /**< (const char *, const void *, ssize_t) + * Message Header */ + RD_KAFKA_VTYPE_HEADERS, /**< (rd_kafka_headers_t *) Headers list */ +} rd_kafka_vtype_t; + + +/** + * @brief VTYPE + argument container for use with rd_kafka_produce_va() + * + * See RD_KAFKA_V_..() macros below for which union field corresponds + * to which RD_KAFKA_VTYPE_... + */ +typedef struct rd_kafka_vu_s { + rd_kafka_vtype_t vtype; /**< RD_KAFKA_VTYPE_.. */ + /** Value union, see RD_KAFKA_V_.. macros for which field to use. */ + union { + const char *cstr; + rd_kafka_topic_t *rkt; + int i; + int32_t i32; + int64_t i64; + struct { + void *ptr; + size_t size; + } mem; + struct { + const char *name; + const void *val; + ssize_t size; + } header; + rd_kafka_headers_t *headers; + void *ptr; + char _pad[64]; /**< Padding size for future-proofness */ + } u; +} rd_kafka_vu_t; + +/** + * @brief Convenience macros for rd_kafka_vtype_t that takes the + * correct arguments for each vtype. + */ + +/*! + * va-arg end sentinel used to terminate the variable argument list + */ +#define RD_KAFKA_V_END RD_KAFKA_VTYPE_END + +/*! + * Topic name (const char *) + * + * rd_kafka_vu_t field: u.cstr + */ +#define RD_KAFKA_V_TOPIC(topic) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_TOPIC, const char *, topic), \ + (const char *)topic +/*! + * Topic object (rd_kafka_topic_t *) + * + * rd_kafka_vu_t field: u.rkt + */ +#define RD_KAFKA_V_RKT(rkt) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_RKT, rd_kafka_topic_t *, rkt), \ + (rd_kafka_topic_t *)rkt +/*! + * Partition (int32_t) + * + * rd_kafka_vu_t field: u.i32 + */ +#define RD_KAFKA_V_PARTITION(partition) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_PARTITION, int32_t, partition), \ + (int32_t)partition +/*! + * Message value/payload pointer and length (void *, size_t) + * + * rd_kafka_vu_t fields: u.mem.ptr, u.mem.size + */ +#define RD_KAFKA_V_VALUE(VALUE,LEN) \ + _LRK_TYPECHECK2(RD_KAFKA_VTYPE_VALUE, void *, VALUE, size_t, LEN), \ + (void *)VALUE, (size_t)LEN +/*! + * Message key pointer and length (const void *, size_t) + * + * rd_kafka_vu_t field: u.mem.ptr, rd_kafka_vu.t.u.mem.size + */ +#define RD_KAFKA_V_KEY(KEY,LEN) \ + _LRK_TYPECHECK2(RD_KAFKA_VTYPE_KEY, const void *, KEY, size_t, LEN), \ + (void *)KEY, (size_t)LEN +/*! + * Message opaque pointer (void *) + * Same as \c msg_opaque, \c produce(.., msg_opaque), + * and \c rkmessage->_private . + * + * rd_kafka_vu_t field: u.ptr + */ +#define RD_KAFKA_V_OPAQUE(msg_opaque) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_OPAQUE, void *, msg_opaque), \ + (void *)msg_opaque +/*! + * Message flags (int) + * @sa RD_KAFKA_MSG_F_COPY, et.al. + * + * rd_kafka_vu_t field: u.i + */ +#define RD_KAFKA_V_MSGFLAGS(msgflags) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_MSGFLAGS, int, msgflags), \ + (int)msgflags +/*! + * Timestamp in milliseconds since epoch UTC (int64_t). + * A value of 0 will use the current wall-clock time. + * + * rd_kafka_vu_t field: u.i64 + */ +#define RD_KAFKA_V_TIMESTAMP(timestamp) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_TIMESTAMP, int64_t, timestamp), \ + (int64_t)timestamp +/*! + * Add Message Header (const char *NAME, const void *VALUE, ssize_t LEN). + * @sa rd_kafka_header_add() + * @remark RD_KAFKA_V_HEADER() and RD_KAFKA_V_HEADERS() MUST NOT be mixed + * in the same call to producev(). + * + * rd_kafka_vu_t fields: u.header.name, u.header.val, u.header.size + */ +#define RD_KAFKA_V_HEADER(NAME,VALUE,LEN) \ + _LRK_TYPECHECK3(RD_KAFKA_VTYPE_HEADER, const char *, NAME, \ + const void *, VALUE, ssize_t, LEN), \ + (const char *)NAME, (const void *)VALUE, (ssize_t)LEN + +/*! + * Message Headers list (rd_kafka_headers_t *). + * The message object will assume ownership of the headers (unless producev() + * fails). + * Any existing headers will be replaced. + * @sa rd_kafka_message_set_headers() + * @remark RD_KAFKA_V_HEADER() and RD_KAFKA_V_HEADERS() MUST NOT be mixed + * in the same call to producev(). + * + * rd_kafka_vu_t fields: u.headers + */ +#define RD_KAFKA_V_HEADERS(HDRS) \ + _LRK_TYPECHECK(RD_KAFKA_VTYPE_HEADERS, rd_kafka_headers_t *, HDRS), \ + (rd_kafka_headers_t *)HDRS + + +/**@}*/ + + +/** + * @name Message headers + * @{ + * + * @brief Message headers consist of a list of (string key, binary value) pairs. + * Duplicate keys are supported and the order in which keys were + * added are retained. + * + * Header values are considered binary and may have three types of + * value: + * - proper value with size > 0 and a valid pointer + * - empty value with size = 0 and any non-NULL pointer + * - null value with size = 0 and a NULL pointer + * + * Headers require Apache Kafka broker version v0.11.0.0 or later. + * + * Header operations are O(n). + */ + + +/** + * @brief Create a new headers list. + * + * @param initial_count Preallocate space for this number of headers. + * Any number of headers may be added, updated and + * removed regardless of the initial count. + */ +RD_EXPORT rd_kafka_headers_t *rd_kafka_headers_new (size_t initial_count); + +/** + * @brief Destroy the headers list. The object and any returned value pointers + * are not usable after this call. + */ +RD_EXPORT void rd_kafka_headers_destroy (rd_kafka_headers_t *hdrs); + +/** + * @brief Make a copy of headers list \p src. + */ +RD_EXPORT rd_kafka_headers_t * +rd_kafka_headers_copy (const rd_kafka_headers_t *src); + +/** + * @brief Add header with name \p name and value \p val (copied) of size + * \p size (not including null-terminator). + * + * @param hdrs Headers list. + * @param name Header name. + * @param name_size Header name size (not including the null-terminator). + * If -1 the \p name length is automatically acquired using + * strlen(). + * @param value Pointer to header value, or NULL (set size to 0 or -1). + * @param value_size Size of header value. If -1 the \p value is assumed to be a + * null-terminated string and the length is automatically + * acquired using strlen(). + * + * @returns RD_KAFKA_RESP_ERR__READ_ONLY if the headers are read-only, + * else RD_KAFKA_RESP_ERR_NO_ERROR. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_header_add (rd_kafka_headers_t *hdrs, + const char *name, ssize_t name_size, + const void *value, ssize_t value_size); + +/** + * @brief Remove all headers for the given key (if any). + * + * @returns RD_KAFKA_RESP_ERR__READ_ONLY if the headers are read-only, + * RD_KAFKA_RESP_ERR__NOENT if no matching headers were found, + * else RD_KAFKA_RESP_ERR_NO_ERROR if headers were removed. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_header_remove (rd_kafka_headers_t *hdrs, const char *name); + + +/** + * @brief Find last header in list \p hdrs matching \p name. + * + * @param hdrs Headers list. + * @param name Header to find (last match). + * @param valuep (out) Set to a (null-terminated) const pointer to the value + * (may be NULL). + * @param sizep (out) Set to the value's size (not including null-terminator). + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR if an entry was found, else + * RD_KAFKA_RESP_ERR__NOENT. + * + * @remark The returned pointer in \p valuep includes a trailing null-terminator + * that is not accounted for in \p sizep. + * @remark The returned pointer is only valid as long as the headers list and + * the header item is valid. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_header_get_last (const rd_kafka_headers_t *hdrs, + const char *name, const void **valuep, size_t *sizep); + +/** + * @brief Iterator for headers matching \p name. + * + * Same semantics as rd_kafka_header_get_last() + * + * @param hdrs Headers to iterate. + * @param idx Iterator index, start at 0 and increment by one for each call + * as long as RD_KAFKA_RESP_ERR_NO_ERROR is returned. + * @param name Header name to match. + * @param valuep (out) Set to a (null-terminated) const pointer to the value + * (may be NULL). + * @param sizep (out) Set to the value's size (not including null-terminator). + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_header_get (const rd_kafka_headers_t *hdrs, size_t idx, + const char *name, const void **valuep, size_t *sizep); + + +/** + * @brief Iterator for all headers. + * + * Same semantics as rd_kafka_header_get() + * + * @sa rd_kafka_header_get() + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_header_get_all (const rd_kafka_headers_t *hdrs, size_t idx, + const char **namep, + const void **valuep, size_t *sizep); + + + +/**@}*/ + + + +/** + * @name Kafka messages + * @{ + * + */ + + + +// FIXME: This doesn't show up in docs for some reason +// "Compound rd_kafka_message_t is not documented." + +/** + * @brief A Kafka message as returned by the \c rd_kafka_consume*() family + * of functions as well as provided to the Producer \c dr_msg_cb(). + * + * For the consumer this object has two purposes: + * - provide the application with a consumed message. (\c err == 0) + * - report per-topic+partition consumer errors (\c err != 0) + * + * The application must check \c err to decide what action to take. + * + * When the application is finished with a message it must call + * rd_kafka_message_destroy() unless otherwise noted. + */ +typedef struct rd_kafka_message_s { + rd_kafka_resp_err_t err; /**< Non-zero for error signaling. */ + rd_kafka_topic_t *rkt; /**< Topic */ + int32_t partition; /**< Partition */ + void *payload; /**< Producer: original message payload. + * Consumer: Depends on the value of \c err : + * - \c err==0: Message payload. + * - \c err!=0: Error string */ + size_t len; /**< Depends on the value of \c err : + * - \c err==0: Message payload length + * - \c err!=0: Error string length */ + void *key; /**< Depends on the value of \c err : + * - \c err==0: Optional message key */ + size_t key_len; /**< Depends on the value of \c err : + * - \c err==0: Optional message key length*/ + int64_t offset; /**< Consumer: + * - Message offset (or offset for error + * if \c err!=0 if applicable). + * Producer, dr_msg_cb: + * Message offset assigned by broker. + * May be RD_KAFKA_OFFSET_INVALID + * for retried messages when + * idempotence is enabled. */ + void *_private; /**< Consumer: + * - rdkafka private pointer: DO NOT MODIFY + * Producer: + * - dr_msg_cb: + * msg_opaque from produce() call or + * RD_KAFKA_V_OPAQUE from producev(). */ +} rd_kafka_message_t; + + +/** + * @brief Frees resources for \p rkmessage and hands ownership back to rdkafka. + */ +RD_EXPORT +void rd_kafka_message_destroy(rd_kafka_message_t *rkmessage); + + + + +/** + * @brief Returns the error string for an errored rd_kafka_message_t or NULL if + * there was no error. + * + * @remark This function MUST NOT be used with the producer. + */ +RD_EXPORT +const char *rd_kafka_message_errstr (const rd_kafka_message_t *rkmessage); + + +/** + * @brief Returns the message timestamp for a consumed message. + * + * The timestamp is the number of milliseconds since the epoch (UTC). + * + * \p tstype (if not NULL) is updated to indicate the type of timestamp. + * + * @returns message timestamp, or -1 if not available. + * + * @remark Message timestamps require broker version 0.10.0 or later. + */ +RD_EXPORT +int64_t rd_kafka_message_timestamp (const rd_kafka_message_t *rkmessage, + rd_kafka_timestamp_type_t *tstype); + + + +/** + * @brief Returns the latency for a produced message measured from + * the produce() call. + * + * @returns the latency in microseconds, or -1 if not available. + */ +RD_EXPORT +int64_t rd_kafka_message_latency (const rd_kafka_message_t *rkmessage); + + +/** + * @brief Returns the broker id of the broker the message was produced to + * or fetched from. + * + * @returns a broker id if known, else -1. + */ +RD_EXPORT +int32_t rd_kafka_message_broker_id (const rd_kafka_message_t *rkmessage); + + +/** + * @brief Get the message header list. + * + * The returned pointer in \p *hdrsp is associated with the \p rkmessage and + * must not be used after destruction of the message object or the header + * list is replaced with rd_kafka_message_set_headers(). + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR if headers were returned, + * RD_KAFKA_RESP_ERR__NOENT if the message has no headers, + * or another error code if the headers could not be parsed. + * + * @remark Headers require broker version 0.11.0.0 or later. + * + * @remark As an optimization the raw protocol headers are parsed on + * the first call to this function. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_message_headers (const rd_kafka_message_t *rkmessage, + rd_kafka_headers_t **hdrsp); + +/** + * @brief Get the message header list and detach the list from the message + * making the application the owner of the headers. + * The application must eventually destroy the headers using + * rd_kafka_headers_destroy(). + * The message's headers will be set to NULL. + * + * Otherwise same semantics as rd_kafka_message_headers() + * + * @sa rd_kafka_message_headers + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_message_detach_headers (rd_kafka_message_t *rkmessage, + rd_kafka_headers_t **hdrsp); + + +/** + * @brief Replace the message's current headers with a new list. + * + * @param rkmessage The message to set headers. + * @param hdrs New header list. The message object assumes ownership of + * the list, the list will be destroyed automatically with + * the message object. + * The new headers list may be updated until the message object + * is passed or returned to librdkafka. + * + * @remark The existing headers object, if any, will be destroyed. + */ +RD_EXPORT +void rd_kafka_message_set_headers (rd_kafka_message_t *rkmessage, + rd_kafka_headers_t *hdrs); + + +/** + * @brief Returns the number of header key/value pairs + * + * @param hdrs Headers to count + */ +RD_EXPORT size_t rd_kafka_header_cnt (const rd_kafka_headers_t *hdrs); + + +/** + * @enum rd_kafka_msg_status_t + * @brief Message persistence status can be used by the application to + * find out if a produced message was persisted in the topic log. + */ +typedef enum { + /** Message was never transmitted to the broker, or failed with + * an error indicating it was not written to the log. + * Application retry risks ordering, but not duplication. */ + RD_KAFKA_MSG_STATUS_NOT_PERSISTED = 0, + + /** Message was transmitted to broker, but no acknowledgement was + * received. + * Application retry risks ordering and duplication. */ + RD_KAFKA_MSG_STATUS_POSSIBLY_PERSISTED = 1, + + /** Message was written to the log and acknowledged by the broker. + * No reason for application to retry. + * Note: this value should only be trusted with \c acks=all. */ + RD_KAFKA_MSG_STATUS_PERSISTED = 2 +} rd_kafka_msg_status_t; + + +/** + * @brief Returns the message's persistence status in the topic log. + * + * @remark The message status is not available in on_acknowledgement + * interceptors. + */ +RD_EXPORT rd_kafka_msg_status_t +rd_kafka_message_status (const rd_kafka_message_t *rkmessage); + +/**@}*/ + + +/** + * @name Configuration interface + * @{ + * + * @brief Main/global configuration property interface + * + */ + +/** + * @enum rd_kafka_conf_res_t + * @brief Configuration result type + */ +typedef enum { + RD_KAFKA_CONF_UNKNOWN = -2, /**< Unknown configuration name. */ + RD_KAFKA_CONF_INVALID = -1, /**< Invalid configuration value or + * property or value not supported in + * this build. */ + RD_KAFKA_CONF_OK = 0 /**< Configuration okay */ +} rd_kafka_conf_res_t; + + +/** + * @brief Create configuration object. + * + * When providing your own configuration to the \c rd_kafka_*_new_*() calls + * the rd_kafka_conf_t objects needs to be created with this function + * which will set up the defaults. + * I.e.: + * @code + * rd_kafka_conf_t *myconf; + * rd_kafka_conf_res_t res; + * + * myconf = rd_kafka_conf_new(); + * res = rd_kafka_conf_set(myconf, "socket.timeout.ms", "600", + * errstr, sizeof(errstr)); + * if (res != RD_KAFKA_CONF_OK) + * die("%s\n", errstr); + * + * rk = rd_kafka_new(..., myconf); + * @endcode + * + * Please see CONFIGURATION.md for the default settings or use + * rd_kafka_conf_properties_show() to provide the information at runtime. + * + * The properties are identical to the Apache Kafka configuration properties + * whenever possible. + * + * @remark A successful call to rd_kafka_new() will assume ownership of + * the conf object and rd_kafka_conf_destroy() must not be called. + * + * @returns A new rd_kafka_conf_t object with defaults set. + * + * @sa rd_kafka_new(), rd_kafka_conf_set(), rd_kafka_conf_destroy() + */ +RD_EXPORT +rd_kafka_conf_t *rd_kafka_conf_new(void); + + +/** + * @brief Destroys a conf object. + */ +RD_EXPORT +void rd_kafka_conf_destroy(rd_kafka_conf_t *conf); + + +/** + * @brief Creates a copy/duplicate of configuration object \p conf + * + * @remark Interceptors are NOT copied to the new configuration object. + * @sa rd_kafka_interceptor_f_on_conf_dup + */ +RD_EXPORT +rd_kafka_conf_t *rd_kafka_conf_dup(const rd_kafka_conf_t *conf); + + +/** + * @brief Same as rd_kafka_conf_dup() but with an array of property name + * prefixes to filter out (ignore) when copying. + */ +RD_EXPORT +rd_kafka_conf_t *rd_kafka_conf_dup_filter (const rd_kafka_conf_t *conf, + size_t filter_cnt, + const char **filter); + + + +/** + * @returns the configuration object used by an rd_kafka_t instance. + * For use with rd_kafka_conf_get(), et.al., to extract configuration + * properties from a running client. + * + * @remark the returned object is read-only and its lifetime is the same + * as the rd_kafka_t object. + */ +RD_EXPORT +const rd_kafka_conf_t *rd_kafka_conf (rd_kafka_t *rk); + + +/** + * @brief Sets a configuration property. + * + * \p conf must have been previously created with rd_kafka_conf_new(). + * + * Fallthrough: + * Topic-level configuration properties may be set using this interface + * in which case they are applied on the \c default_topic_conf. + * If no \c default_topic_conf has been set one will be created. + * Any sub-sequent rd_kafka_conf_set_default_topic_conf() calls will + * replace the current default topic configuration. + * + * @returns \c rd_kafka_conf_res_t to indicate success or failure. + * In case of failure \p errstr is updated to contain a human readable + * error string. + * + * @remark Setting properties or values that were disabled at build time due to + * missing dependencies will return RD_KAFKA_CONF_INVALID. + */ +RD_EXPORT +rd_kafka_conf_res_t rd_kafka_conf_set(rd_kafka_conf_t *conf, + const char *name, + const char *value, + char *errstr, size_t errstr_size); + + +/** + * @brief Enable event sourcing. + * \p events is a bitmask of \c RD_KAFKA_EVENT_* of events to enable + * for consumption by `rd_kafka_queue_poll()`. + */ +RD_EXPORT +void rd_kafka_conf_set_events(rd_kafka_conf_t *conf, int events); + + +/** + * @brief Generic event callback to be used with the event API to trigger + * callbacks for \c rd_kafka_event_t objects from a background + * thread serving the background queue. + * + * How to use: + * 1. First set the event callback on the configuration object with this + * function, followed by creating an rd_kafka_t instance + * with rd_kafka_new(). + * 2. Get the instance's background queue with rd_kafka_queue_get_background() + * and pass it as the reply/response queue to an API that takes an + * event queue, such as rd_kafka_CreateTopics(). + * 3. As the response event is ready and enqueued on the background queue the + * event callback will be triggered from the background thread. + * 4. Prior to destroying the client instance, loose your reference to the + * background queue by calling rd_kafka_queue_destroy(). + * + * The application must destroy the \c rkev passed to \p event cb using + * rd_kafka_event_destroy(). + * + * The \p event_cb \c opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @remark This callback is a specialized alternative to the poll-based + * event API described in the Event interface section. + * + * @remark The \p event_cb will be called spontaneously from a background + * thread completely managed by librdkafka. + * Take care to perform proper locking of application objects. + * + * @warning The application MUST NOT call rd_kafka_destroy() from the + * event callback. + * + * @sa rd_kafka_queue_get_background + */ +RD_EXPORT void +rd_kafka_conf_set_background_event_cb (rd_kafka_conf_t *conf, + void (*event_cb) (rd_kafka_t *rk, + rd_kafka_event_t *rkev, + void *opaque)); + + +/** + * @deprecated See rd_kafka_conf_set_dr_msg_cb() + */ +RD_EXPORT +void rd_kafka_conf_set_dr_cb(rd_kafka_conf_t *conf, + void (*dr_cb) (rd_kafka_t *rk, + void *payload, size_t len, + rd_kafka_resp_err_t err, + void *opaque, void *msg_opaque)); + +/** + * @brief \b Producer: Set delivery report callback in provided \p conf object. + * + * The delivery report callback will be called once for each message + * accepted by rd_kafka_produce() (et.al) with \p err set to indicate + * the result of the produce request. + * + * The callback is called when a message is succesfully produced or + * if librdkafka encountered a permanent failure. + * Delivery errors occur when the retry count is exceeded, when the + * message.timeout.ms timeout is exceeded or there is a permanent error + * like RD_KAFKA_RESP_ERR_UNKNOWN_TOPIC_OR_PART. + * + * An application must call rd_kafka_poll() at regular intervals to + * serve queued delivery report callbacks. + * + * The broker-assigned offset can be retrieved with \c rkmessage->offset + * and the timestamp can be retrieved using rd_kafka_message_timestamp(). + * + * The \p dr_msg_cb \c opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * The per-message msg_opaque value is available in + * \c rd_kafka_message_t._private. + * + * @remark The Idempotent Producer may return invalid timestamp + * (RD_KAFKA_TIMESTAMP_NOT_AVAILABLE), and + * and offset (RD_KAFKA_OFFSET_INVALID) for retried messages + * that were previously successfully delivered but not properly + * acknowledged. + */ +RD_EXPORT +void rd_kafka_conf_set_dr_msg_cb(rd_kafka_conf_t *conf, + void (*dr_msg_cb) (rd_kafka_t *rk, + const rd_kafka_message_t * + rkmessage, + void *opaque)); + + +/** + * @brief \b Consumer: Set consume callback for use with + * rd_kafka_consumer_poll() + * + * The \p consume_cb \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + */ +RD_EXPORT +void rd_kafka_conf_set_consume_cb (rd_kafka_conf_t *conf, + void (*consume_cb) (rd_kafka_message_t * + rkmessage, + void *opaque)); + +/** + * @brief \b Consumer: Set rebalance callback for use with + * coordinated consumer group balancing. + * + * The \p err field is set to either RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS + * or RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS and 'partitions' + * contains the full partition set that was either assigned or revoked. + * + * Registering a \p rebalance_cb turns off librdkafka's automatic + * partition assignment/revocation and instead delegates that responsibility + * to the application's \p rebalance_cb. + * + * The rebalance callback is responsible for updating librdkafka's + * assignment set based on the two events: RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS + * and RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS but should also be able to handle + * arbitrary rebalancing failures where \p err is neither of those. + * @remark In this latter case (arbitrary error), the application must + * call rd_kafka_assign(rk, NULL) to synchronize state. + * + * For eager/non-cooperative `partition.assignment.strategy` assignors, + * such as `range` and `roundrobin`, the application must use + * rd_kafka_assign() to set or clear the entire assignment. + * For the cooperative assignors, such as `cooperative-sticky`, the application + * must use rd_kafka_incremental_assign() for + * RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS and rd_kafka_incremental_unassign() + * for RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS. + * + * Without a rebalance callback this is done automatically by librdkafka + * but registering a rebalance callback gives the application flexibility + * in performing other operations along with the assigning/revocation, + * such as fetching offsets from an alternate location (on assign) + * or manually committing offsets (on revoke). + * + * rebalance_cb is always triggered exactly once when a rebalance completes + * with a new assignment, even if that assignment is empty. If an + * eager/non-cooperative assignor is configured, there will eventually be + * exactly one corresponding call to rebalance_cb to revoke these partitions + * (even if empty), whether this is due to a group rebalance or lost + * partitions. In the cooperative case, rebalance_cb will never be called if + * the set of partitions being revoked is empty (whether or not lost). + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @remark The \p partitions list is destroyed by librdkafka on return + * return from the rebalance_cb and must not be freed or + * saved by the application. + * + * @remark Be careful when modifying the \p partitions list. + * Changing this list should only be done to change the initial + * offsets for each partition. + * But a function like `rd_kafka_position()` might have unexpected + * effects for instance when a consumer gets assigned a partition + * it used to consume at an earlier rebalance. In this case, the + * list of partitions will be updated with the old offset for that + * partition. In this case, it is generally better to pass a copy + * of the list (see `rd_kafka_topic_partition_list_copy()`). + * The result of `rd_kafka_position()` is typically outdated in + * RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS. + * + * @sa rd_kafka_assign() + * @sa rd_kafka_incremental_assign() + * @sa rd_kafka_incremental_unassign() + * @sa rd_kafka_assignment_lost() + * @sa rd_kafka_rebalance_protocol() + * + * The following example shows the application's responsibilities: + * @code + * static void rebalance_cb (rd_kafka_t *rk, rd_kafka_resp_err_t err, + * rd_kafka_topic_partition_list_t *partitions, + * void *opaque) { + * + * switch (err) + * { + * case RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS: + * // application may load offets from arbitrary external + * // storage here and update \p partitions + * if (!strcmp(rd_kafka_rebalance_protocol(rk), "COOPERATIVE")) + * rd_kafka_incremental_assign(rk, partitions); + * else // EAGER + * rd_kafka_assign(rk, partitions); + * break; + * + * case RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS: + * if (manual_commits) // Optional explicit manual commit + * rd_kafka_commit(rk, partitions, 0); // sync commit + * + * if (!strcmp(rd_kafka_rebalance_protocol(rk), "COOPERATIVE")) + * rd_kafka_incremental_unassign(rk, partitions); + * else // EAGER + * rd_kafka_assign(rk, NULL); + * break; + * + * default: + * handle_unlikely_error(err); + * rd_kafka_assign(rk, NULL); // sync state + * break; + * } + * } + * @endcode + * + * @remark The above example lacks error handling for assign calls, see + * the examples/ directory. + */ +RD_EXPORT +void rd_kafka_conf_set_rebalance_cb ( + rd_kafka_conf_t *conf, + void (*rebalance_cb) (rd_kafka_t *rk, + rd_kafka_resp_err_t err, + rd_kafka_topic_partition_list_t *partitions, + void *opaque)); + + + +/** + * @brief \b Consumer: Set offset commit callback for use with consumer groups. + * + * The results of automatic or manual offset commits will be scheduled + * for this callback and is served by rd_kafka_consumer_poll(). + * + * If no partitions had valid offsets to commit this callback will be called + * with \p err == RD_KAFKA_RESP_ERR__NO_OFFSET which is not to be considered + * an error. + * + * The \p offsets list contains per-partition information: + * - \c offset: committed offset (attempted) + * - \c err: commit error + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + */ +RD_EXPORT +void rd_kafka_conf_set_offset_commit_cb ( + rd_kafka_conf_t *conf, + void (*offset_commit_cb) (rd_kafka_t *rk, + rd_kafka_resp_err_t err, + rd_kafka_topic_partition_list_t *offsets, + void *opaque)); + + +/** + * @brief Set error callback in provided conf object. + * + * The error callback is used by librdkafka to signal warnings and errors + * back to the application. + * + * These errors should generally be considered informational and non-permanent, + * the client will try to recover automatically from all type of errors. + * Given that the client and cluster configuration is correct the + * application should treat these as temporary errors. + * + * \p error_cb will be triggered with \c err set to RD_KAFKA_RESP_ERR__FATAL + * if a fatal error has been raised; in this case use rd_kafka_fatal_error() to + * retrieve the fatal error code and error string, and then begin terminating + * the client instance. + * + * If no \p error_cb is registered, or RD_KAFKA_EVENT_ERROR has not been set + * with rd_kafka_conf_set_events, then the errors will be logged instead. + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + */ +RD_EXPORT +void rd_kafka_conf_set_error_cb(rd_kafka_conf_t *conf, + void (*error_cb) (rd_kafka_t *rk, int err, + const char *reason, + void *opaque)); + +/** + * @brief Set throttle callback. + * + * The throttle callback is used to forward broker throttle times to the + * application for Produce and Fetch (consume) requests. + * + * Callbacks are triggered whenever a non-zero throttle time is returned by + * the broker, or when the throttle time drops back to zero. + * + * An application must call rd_kafka_poll() or rd_kafka_consumer_poll() at + * regular intervals to serve queued callbacks. + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @remark Requires broker version 0.9.0 or later. + */ +RD_EXPORT +void rd_kafka_conf_set_throttle_cb (rd_kafka_conf_t *conf, + void (*throttle_cb) ( + rd_kafka_t *rk, + const char *broker_name, + int32_t broker_id, + int throttle_time_ms, + void *opaque)); + + +/** + * @brief Set logger callback. + * + * The default is to print to stderr, but a syslog logger is also available, + * see rd_kafka_log_print and rd_kafka_log_syslog for the builtin alternatives. + * Alternatively the application may provide its own logger callback. + * Or pass \p func as NULL to disable logging. + * + * This is the configuration alternative to the deprecated rd_kafka_set_logger() + * + * @remark The log_cb will be called spontaneously from librdkafka's internal + * threads unless logs have been forwarded to a poll queue through + * \c rd_kafka_set_log_queue(). + * An application MUST NOT call any librdkafka APIs or do any prolonged + * work in a non-forwarded \c log_cb. + */ +RD_EXPORT +void rd_kafka_conf_set_log_cb(rd_kafka_conf_t *conf, + void (*log_cb) (const rd_kafka_t *rk, int level, + const char *fac, const char *buf)); + + +/** + * @brief Set statistics callback in provided conf object. + * + * The statistics callback is triggered from rd_kafka_poll() every + * \c statistics.interval.ms (needs to be configured separately). + * Function arguments: + * - \p rk - Kafka handle + * - \p json - String containing the statistics data in JSON format + * - \p json_len - Length of \p json string. + * - \p opaque - Application-provided opaque as set by + * rd_kafka_conf_set_opaque(). + * + * For more information on the format of \p json, see + * https://github.com/edenhill/librdkafka/wiki/Statistics + * + * If the application wishes to hold on to the \p json pointer and free + * it at a later time it must return 1 from the \p stats_cb. + * If the application returns 0 from the \p stats_cb then librdkafka + * will immediately free the \p json pointer. + * + * See STATISTICS.md for a full definition of the JSON object. + */ +RD_EXPORT +void rd_kafka_conf_set_stats_cb(rd_kafka_conf_t *conf, + int (*stats_cb) (rd_kafka_t *rk, + char *json, + size_t json_len, + void *opaque)); + +/** + * @brief Set SASL/OAUTHBEARER token refresh callback in provided conf object. + * + * @param conf the configuration to mutate. + * @param oauthbearer_token_refresh_cb the callback to set; callback function + * arguments:
+ * \p rk - Kafka handle
+ * \p oauthbearer_config - Value of configuration property + * sasl.oauthbearer.config. + * \p opaque - Application-provided opaque set via + * rd_kafka_conf_set_opaque() + * + * The SASL/OAUTHBEARER token refresh callback is triggered via rd_kafka_poll() + * whenever OAUTHBEARER is the SASL mechanism and a token needs to be retrieved, + * typically based on the configuration defined in \c sasl.oauthbearer.config. + * + * The callback should invoke rd_kafka_oauthbearer_set_token() + * or rd_kafka_oauthbearer_set_token_failure() to indicate success + * or failure, respectively. + * + * The refresh operation is eventable and may be received via + * rd_kafka_queue_poll() with an event type of + * \c RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH. + * + * Note that before any SASL/OAUTHBEARER broker connection can succeed the + * application must call rd_kafka_oauthbearer_set_token() once -- either + * directly or, more typically, by invoking either rd_kafka_poll() or + * rd_kafka_queue_poll() -- in order to cause retrieval of an initial token to + * occur. + * + * An unsecured JWT refresh handler is provided by librdkafka for development + * and testing purposes, it is enabled by setting + * the \c enable.sasl.oauthbearer.unsecure.jwt property to true and is + * mutually exclusive to using a refresh callback. + */ +RD_EXPORT +void rd_kafka_conf_set_oauthbearer_token_refresh_cb ( + rd_kafka_conf_t *conf, + void (*oauthbearer_token_refresh_cb) (rd_kafka_t *rk, + const char *oauthbearer_config, + void *opaque)); + +/** + * @brief Set socket callback. + * + * The socket callback is responsible for opening a socket + * according to the supplied \p domain, \p type and \p protocol. + * The socket shall be created with \c CLOEXEC set in a racefree fashion, if + * possible. + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * Default: + * - on linux: racefree CLOEXEC + * - others : non-racefree CLOEXEC + * + * @remark The callback will be called from an internal librdkafka thread. + */ +RD_EXPORT +void rd_kafka_conf_set_socket_cb(rd_kafka_conf_t *conf, + int (*socket_cb) (int domain, int type, + int protocol, + void *opaque)); + + + +/** + * @brief Set connect callback. + * + * The connect callback is responsible for connecting socket \p sockfd + * to peer address \p addr. + * The \p id field contains the broker identifier. + * + * \p connect_cb shall return 0 on success (socket connected) or an error + * number (errno) on error. + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @remark The callback will be called from an internal librdkafka thread. + */ +RD_EXPORT void +rd_kafka_conf_set_connect_cb (rd_kafka_conf_t *conf, + int (*connect_cb) (int sockfd, + const struct sockaddr *addr, + int addrlen, + const char *id, + void *opaque)); + +/** + * @brief Set close socket callback. + * + * Close a socket (optionally opened with socket_cb()). + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @remark The callback will be called from an internal librdkafka thread. + */ +RD_EXPORT void +rd_kafka_conf_set_closesocket_cb (rd_kafka_conf_t *conf, + int (*closesocket_cb) (int sockfd, + void *opaque)); + + + +#ifndef _WIN32 +/** + * @brief Set open callback. + * + * The open callback is responsible for opening the file specified by + * pathname, flags and mode. + * The file shall be opened with \c CLOEXEC set in a racefree fashion, if + * possible. + * + * Default: + * - on linux: racefree CLOEXEC + * - others : non-racefree CLOEXEC + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @remark The callback will be called from an internal librdkafka thread. + */ +RD_EXPORT +void rd_kafka_conf_set_open_cb (rd_kafka_conf_t *conf, + int (*open_cb) (const char *pathname, + int flags, mode_t mode, + void *opaque)); +#endif + + +/** + * @brief Sets the verification callback of the broker certificate + * + * The verification callback is triggered from internal librdkafka threads + * upon connecting to a broker. On each connection attempt the callback + * will be called for each certificate in the broker's certificate chain, + * starting at the root certification, as long as the application callback + * returns 1 (valid certificate). + * \c broker_name and \c broker_id correspond to the broker the connection + * is being made to. + * The \c x509_error argument indicates if OpenSSL's verification of + * the certificate succeed (0) or failed (an OpenSSL error code). + * The application may set the SSL context error code by returning 0 + * from the verify callback and providing a non-zero SSL context error code + * in \c x509_error. + * If the verify callback sets \c x509_error to 0, returns 1, and the + * original \c x509_error was non-zero, the error on the SSL context will + * be cleared. + * \c x509_error is always a valid pointer to an int. + * + * \c depth is the depth of the current certificate in the chain, starting + * at the root certificate. + * + * The certificate itself is passed in binary DER format in \c buf of + * size \c size. + * + * The callback must return 1 if verification succeeds, or + * 0 if verification fails and then write a human-readable error message + * to \c errstr (limited to \c errstr_size bytes, including nul-term). + * + * The callback's \p opaque argument is the opaque set with + * rd_kafka_conf_set_opaque(). + * + * @returns RD_KAFKA_CONF_OK if SSL is supported in this build, else + * RD_KAFKA_CONF_INVALID. + * + * @warning This callback will be called from internal librdkafka threads. + * + * @remark See in the OpenSSL source distribution + * for a list of \p x509_error codes. + */ +RD_EXPORT +rd_kafka_conf_res_t rd_kafka_conf_set_ssl_cert_verify_cb ( + rd_kafka_conf_t *conf, + int (*ssl_cert_verify_cb) (rd_kafka_t *rk, + const char *broker_name, + int32_t broker_id, + int *x509_error, + int depth, + const char *buf, size_t size, + char *errstr, size_t errstr_size, + void *opaque)); + + +/** + * @enum rd_kafka_cert_type_t + * + * @brief SSL certificate type + * + * @sa rd_kafka_conf_set_ssl_cert + */ +typedef enum rd_kafka_cert_type_t { + RD_KAFKA_CERT_PUBLIC_KEY, /**< Client's public key */ + RD_KAFKA_CERT_PRIVATE_KEY, /**< Client's private key */ + RD_KAFKA_CERT_CA, /**< CA certificate */ + RD_KAFKA_CERT__CNT, +} rd_kafka_cert_type_t; + +/** + * @enum rd_kafka_cert_enc_t + * + * @brief SSL certificate encoding + * + * @sa rd_kafka_conf_set_ssl_cert + */ +typedef enum rd_kafka_cert_enc_t { + RD_KAFKA_CERT_ENC_PKCS12, /**< PKCS#12 */ + RD_KAFKA_CERT_ENC_DER, /**< DER / binary X.509 ASN1 */ + RD_KAFKA_CERT_ENC_PEM, /**< PEM */ + RD_KAFKA_CERT_ENC__CNT, +} rd_kafka_cert_enc_t; + + +/** + * @brief Set certificate/key \p cert_type from the \p cert_enc encoded + * memory at \p buffer of \p size bytes. + * + * @param conf Configuration object. + * @param cert_type Certificate or key type to configure. + * @param cert_enc Buffer \p encoding type. + * @param buffer Memory pointer to encoded certificate or key. + * The memory is not referenced after this function returns. + * @param size Size of memory at \p buffer. + * @param errstr Memory were a human-readable error string will be written + * on failure. + * @param errstr_size Size of \p errstr, including space for nul-terminator. + * + * @returns RD_KAFKA_CONF_OK on success or RD_KAFKA_CONF_INVALID if the + * memory in \p buffer is of incorrect encoding, or if librdkafka + * was not built with SSL support. + * + * @remark Calling this method multiple times with the same \p cert_type + * will replace the previous value. + * + * @remark Calling this method with \p buffer set to NULL will clear the + * configuration for \p cert_type. + * + * @remark The private key may require a password, which must be specified + * with the `ssl.key.password` configuration property prior to + * calling this function. + * + * @remark Private and public keys in PEM format may also be set with the + * `ssl.key.pem` and `ssl.certificate.pem` configuration properties. + * + * @remark CA certificate in PEM format may also be set with the + * `ssl.ca.pem` configuration property. + */ +RD_EXPORT rd_kafka_conf_res_t +rd_kafka_conf_set_ssl_cert (rd_kafka_conf_t *conf, + rd_kafka_cert_type_t cert_type, + rd_kafka_cert_enc_t cert_enc, + const void *buffer, size_t size, + char *errstr, size_t errstr_size); + + +/** + * @brief Set callback_data for OpenSSL engine. + * + * @param conf Configuration object. + * @param callback_data passed to engine callbacks, + * e.g. \c ENGINE_load_ssl_client_cert. + * + * @remark The \c ssl.engine.location configuration must be set for this + * to have affect. + * + * @remark The memory pointed to by \p value must remain valid for the + * lifetime of the configuration object and any Kafka clients that + * use it. + */ +RD_EXPORT +void rd_kafka_conf_set_engine_callback_data (rd_kafka_conf_t *conf, + void *callback_data); + + +/** + * @brief Sets the application's opaque pointer that will be passed to callbacks + * + * @sa rd_kafka_opaque() + */ +RD_EXPORT +void rd_kafka_conf_set_opaque(rd_kafka_conf_t *conf, void *opaque); + +/** + * @brief Retrieves the opaque pointer previously set + * with rd_kafka_conf_set_opaque() + */ +RD_EXPORT +void *rd_kafka_opaque(const rd_kafka_t *rk); + + + +/** + * @brief Sets the default topic configuration to use for automatically + * subscribed topics (e.g., through pattern-matched topics). + * The topic config object is not usable after this call. + * + * @warning Any topic configuration settings that have been set on the + * global rd_kafka_conf_t object will be overwritten by this call + * since the implicitly created default topic config object is + * replaced by the user-supplied one. + * + * @deprecated Set default topic level configuration on the + * global rd_kafka_conf_t object instead. + */ +RD_EXPORT +void rd_kafka_conf_set_default_topic_conf (rd_kafka_conf_t *conf, + rd_kafka_topic_conf_t *tconf); + +/** + * @brief Gets the default topic configuration as previously set with + * rd_kafka_conf_set_default_topic_conf() or that was implicitly created + * by configuring a topic-level property on the global \p conf object. + * + * @returns the \p conf's default topic configuration (if any), or NULL. + * + * @warning The returned topic configuration object is owned by the \p conf + * object. It may be modified but not destroyed and its lifetime is + * the same as the \p conf object or the next call to + * rd_kafka_conf_set_default_topic_conf(). + */ +RD_EXPORT rd_kafka_topic_conf_t * +rd_kafka_conf_get_default_topic_conf (rd_kafka_conf_t *conf); + + +/** + * @brief Retrieve configuration value for property \p name. + * + * If \p dest is non-NULL the value will be written to \p dest with at + * most \p dest_size. + * + * \p *dest_size is updated to the full length of the value, thus if + * \p *dest_size initially is smaller than the full length the application + * may reallocate \p dest to fit the returned \p *dest_size and try again. + * + * If \p dest is NULL only the full length of the value is returned. + * + * Fallthrough: + * Topic-level configuration properties from the \c default_topic_conf + * may be retrieved using this interface. + * + * @returns \p RD_KAFKA_CONF_OK if the property name matched, else + * \p RD_KAFKA_CONF_UNKNOWN. + */ +RD_EXPORT +rd_kafka_conf_res_t rd_kafka_conf_get (const rd_kafka_conf_t *conf, + const char *name, + char *dest, size_t *dest_size); + + +/** + * @brief Retrieve topic configuration value for property \p name. + * + * @sa rd_kafka_conf_get() + */ +RD_EXPORT +rd_kafka_conf_res_t rd_kafka_topic_conf_get (const rd_kafka_topic_conf_t *conf, + const char *name, + char *dest, size_t *dest_size); + + +/** + * @brief Dump the configuration properties and values of \p conf to an array + * with \"key\", \"value\" pairs. + * + * The number of entries in the array is returned in \p *cntp. + * + * The dump must be freed with `rd_kafka_conf_dump_free()`. + */ +RD_EXPORT +const char **rd_kafka_conf_dump(rd_kafka_conf_t *conf, size_t *cntp); + + +/** + * @brief Dump the topic configuration properties and values of \p conf + * to an array with \"key\", \"value\" pairs. + * + * The number of entries in the array is returned in \p *cntp. + * + * The dump must be freed with `rd_kafka_conf_dump_free()`. + */ +RD_EXPORT +const char **rd_kafka_topic_conf_dump(rd_kafka_topic_conf_t *conf, + size_t *cntp); + +/** + * @brief Frees a configuration dump returned from `rd_kafka_conf_dump()` or + * `rd_kafka_topic_conf_dump(). + */ +RD_EXPORT +void rd_kafka_conf_dump_free(const char **arr, size_t cnt); + +/** + * @brief Prints a table to \p fp of all supported configuration properties, + * their default values as well as a description. + * + * @remark All properties and properties and values are shown, even those + * that have been disabled at build time due to missing dependencies. + */ +RD_EXPORT +void rd_kafka_conf_properties_show(FILE *fp); + +/**@}*/ + + +/** + * @name Topic configuration + * @{ + * + * @brief Topic configuration property interface + * + */ + + +/** + * @brief Create topic configuration object + * + * @sa Same semantics as for rd_kafka_conf_new(). + */ +RD_EXPORT +rd_kafka_topic_conf_t *rd_kafka_topic_conf_new(void); + + +/** + * @brief Creates a copy/duplicate of topic configuration object \p conf. + */ +RD_EXPORT +rd_kafka_topic_conf_t *rd_kafka_topic_conf_dup(const rd_kafka_topic_conf_t + *conf); + +/** + * @brief Creates a copy/duplicate of \p rk 's default topic configuration + * object. + */ +RD_EXPORT +rd_kafka_topic_conf_t *rd_kafka_default_topic_conf_dup (rd_kafka_t *rk); + + +/** + * @brief Destroys a topic conf object. + */ +RD_EXPORT +void rd_kafka_topic_conf_destroy(rd_kafka_topic_conf_t *topic_conf); + + +/** + * @brief Sets a single rd_kafka_topic_conf_t value by property name. + * + * \p topic_conf should have been previously set up + * with `rd_kafka_topic_conf_new()`. + * + * @returns rd_kafka_conf_res_t to indicate success or failure. + */ +RD_EXPORT +rd_kafka_conf_res_t rd_kafka_topic_conf_set(rd_kafka_topic_conf_t *conf, + const char *name, + const char *value, + char *errstr, size_t errstr_size); + +/** + * @brief Sets the application's opaque pointer that will be passed to all topic + * callbacks as the \c rkt_opaque argument. + * + * @sa rd_kafka_topic_opaque() + */ +RD_EXPORT +void rd_kafka_topic_conf_set_opaque(rd_kafka_topic_conf_t *conf, + void *rkt_opaque); + + +/** + * @brief \b Producer: Set partitioner callback in provided topic conf object. + * + * The partitioner may be called in any thread at any time, + * it may be called multiple times for the same message/key. + * + * The callback's \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The callback's \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * Partitioner function constraints: + * - MUST NOT call any rd_kafka_*() functions except: + * rd_kafka_topic_partition_available() + * - MUST NOT block or execute for prolonged periods of time. + * - MUST return a value between 0 and partition_cnt-1, or the + * special \c RD_KAFKA_PARTITION_UA value if partitioning + * could not be performed. + */ +RD_EXPORT +void +rd_kafka_topic_conf_set_partitioner_cb (rd_kafka_topic_conf_t *topic_conf, + int32_t (*partitioner) ( + const rd_kafka_topic_t *rkt, + const void *keydata, + size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, + void *msg_opaque)); + + +/** + * @brief \b Producer: Set message queueing order comparator callback. + * + * The callback may be called in any thread at any time, + * it may be called multiple times for the same message. + * + * Ordering comparator function constraints: + * - MUST be stable sort (same input gives same output). + * - MUST NOT call any rd_kafka_*() functions. + * - MUST NOT block or execute for prolonged periods of time. + * + * The comparator shall compare the two messages and return: + * - < 0 if message \p a should be inserted before message \p b. + * - >=0 if message \p a should be inserted after message \p b. + * + * @remark Insert sorting will be used to enqueue the message in the + * correct queue position, this comes at a cost of O(n). + * + * @remark If `queuing.strategy=fifo` new messages are enqueued to the + * tail of the queue regardless of msg_order_cmp, but retried messages + * are still affected by msg_order_cmp. + * + * @warning THIS IS AN EXPERIMENTAL API, SUBJECT TO CHANGE OR REMOVAL, + * DO NOT USE IN PRODUCTION. + */ +RD_EXPORT void +rd_kafka_topic_conf_set_msg_order_cmp (rd_kafka_topic_conf_t *topic_conf, + int (*msg_order_cmp) ( + const rd_kafka_message_t *a, + const rd_kafka_message_t *b)); + + +/** + * @brief Check if partition is available (has a leader broker). + * + * @returns 1 if the partition is available, else 0. + * + * @warning This function must only be called from inside a partitioner function + */ +RD_EXPORT +int rd_kafka_topic_partition_available(const rd_kafka_topic_t *rkt, + int32_t partition); + + +/******************************************************************* + * * + * Partitioners provided by rdkafka * + * * + *******************************************************************/ + +/** + * @brief Random partitioner. + * + * Will try not to return unavailable partitions. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a random partition between 0 and \p partition_cnt - 1. + * + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_random(const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, void *msg_opaque); + +/** + * @brief Consistent partitioner. + * + * Uses consistent hashing to map identical keys onto identical partitions. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a \"random\" partition between 0 and \p partition_cnt - 1 based on + * the CRC value of the key + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_consistent (const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, void *msg_opaque); + +/** + * @brief Consistent-Random partitioner. + * + * This is the default partitioner. + * Uses consistent hashing to map identical keys onto identical partitions, and + * messages without keys will be assigned via the random partitioner. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a \"random\" partition between 0 and \p partition_cnt - 1 based on + * the CRC value of the key (if provided) + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_consistent_random (const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, void *msg_opaque); + + +/** + * @brief Murmur2 partitioner (Java compatible). + * + * Uses consistent hashing to map identical keys onto identical partitions + * using Java-compatible Murmur2 hashing. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a partition between 0 and \p partition_cnt - 1. + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_murmur2 (const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, + void *msg_opaque); + +/** + * @brief Consistent-Random Murmur2 partitioner (Java compatible). + * + * Uses consistent hashing to map identical keys onto identical partitions + * using Java-compatible Murmur2 hashing. + * Messages without keys will be assigned via the random partitioner. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a partition between 0 and \p partition_cnt - 1. + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_murmur2_random (const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, + void *msg_opaque); + + +/** + * @brief FNV-1a partitioner. + * + * Uses consistent hashing to map identical keys onto identical partitions + * using FNV-1a hashing. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a partition between 0 and \p partition_cnt - 1. + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_fnv1a (const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, + void *msg_opaque); + + +/** + * @brief Consistent-Random FNV-1a partitioner. + * + * Uses consistent hashing to map identical keys onto identical partitions + * using FNV-1a hashing. + * Messages without keys will be assigned via the random partitioner. + * + * The \p rkt_opaque argument is the opaque set by + * rd_kafka_topic_conf_set_opaque(). + * The \p msg_opaque argument is the per-message opaque + * passed to produce(). + * + * @returns a partition between 0 and \p partition_cnt - 1. + */ +RD_EXPORT +int32_t rd_kafka_msg_partitioner_fnv1a_random (const rd_kafka_topic_t *rkt, + const void *key, size_t keylen, + int32_t partition_cnt, + void *rkt_opaque, + void *msg_opaque); + + +/**@}*/ + + + +/** + * @name Main Kafka and Topic object handles + * @{ + * + * + */ + + + + +/** + * @brief Creates a new Kafka handle and starts its operation according to the + * specified \p type (\p RD_KAFKA_CONSUMER or \p RD_KAFKA_PRODUCER). + * + * \p conf is an optional struct created with `rd_kafka_conf_new()` that will + * be used instead of the default configuration. + * The \p conf object is freed by this function on success and must not be used + * or destroyed by the application sub-sequently. + * See `rd_kafka_conf_set()` et.al for more information. + * + * \p errstr must be a pointer to memory of at least size \p errstr_size where + * `rd_kafka_new()` may write a human readable error message in case the + * creation of a new handle fails. In which case the function returns NULL. + * + * @remark \b RD_KAFKA_CONSUMER: When a new \p RD_KAFKA_CONSUMER + * rd_kafka_t handle is created it may either operate in the + * legacy simple consumer mode using the rd_kafka_consume_start() + * interface, or the High-level KafkaConsumer API. + * @remark An application must only use one of these groups of APIs on a given + * rd_kafka_t RD_KAFKA_CONSUMER handle. + + * + * @returns The Kafka handle on success or NULL on error (see \p errstr) + * + * @sa To destroy the Kafka handle, use rd_kafka_destroy(). + */ +RD_EXPORT +rd_kafka_t *rd_kafka_new(rd_kafka_type_t type, rd_kafka_conf_t *conf, + char *errstr, size_t errstr_size); + + +/** + * @brief Destroy Kafka handle. + * + * @remark This is a blocking operation. + * @remark rd_kafka_consumer_close() will be called from this function + * if the instance type is RD_KAFKA_CONSUMER, a \c group.id was + * configured, and the rd_kafka_consumer_close() was not + * explicitly called by the application. This in turn may + * trigger consumer callbacks, such as rebalance_cb. + * Use rd_kafka_destroy_flags() with + * RD_KAFKA_DESTROY_F_NO_CONSUMER_CLOSE to avoid this behaviour. + * + * @sa rd_kafka_destroy_flags() + */ +RD_EXPORT +void rd_kafka_destroy(rd_kafka_t *rk); + + +/** + * @brief Destroy Kafka handle according to specified destroy flags + * + */ +RD_EXPORT +void rd_kafka_destroy_flags (rd_kafka_t *rk, int flags); + +/** + * @brief Flags for rd_kafka_destroy_flags() + */ + +/*! + * Don't call consumer_close() to leave group and commit final offsets. + * + * This also disables consumer callbacks to be called from rd_kafka_destroy*(), + * such as rebalance_cb. + * + * The consumer group handler is still closed internally, but from an + * application perspective none of the functionality from consumer_close() + * is performed. + */ +#define RD_KAFKA_DESTROY_F_NO_CONSUMER_CLOSE 0x8 + + + +/** + * @brief Returns Kafka handle name. + */ +RD_EXPORT +const char *rd_kafka_name(const rd_kafka_t *rk); + + +/** + * @brief Returns Kafka handle type. + */ +RD_EXPORT +rd_kafka_type_t rd_kafka_type(const rd_kafka_t *rk); + + +/** + * @brief Returns this client's broker-assigned group member id. + * + * @remark This currently requires the high-level KafkaConsumer + * + * @returns An allocated string containing the current broker-assigned group + * member id, or NULL if not available. + * The application must free the string with \p free() or + * rd_kafka_mem_free() + */ +RD_EXPORT +char *rd_kafka_memberid (const rd_kafka_t *rk); + + + +/** + * @brief Returns the ClusterId as reported in broker metadata. + * + * @param rk Client instance. + * @param timeout_ms If there is no cached value from metadata retrieval + * then this specifies the maximum amount of time + * (in milliseconds) the call will block waiting + * for metadata to be retrieved. + * Use 0 for non-blocking calls. + + * @remark Requires broker version >=0.10.0 and api.version.request=true. + * + * @remark The application must free the returned pointer + * using rd_kafka_mem_free(). + * + * @returns a newly allocated string containing the ClusterId, or NULL + * if no ClusterId could be retrieved in the allotted timespan. + */ +RD_EXPORT +char *rd_kafka_clusterid (rd_kafka_t *rk, int timeout_ms); + + +/** + * @brief Returns the current ControllerId as reported in broker metadata. + * + * @param rk Client instance. + * @param timeout_ms If there is no cached value from metadata retrieval + * then this specifies the maximum amount of time + * (in milliseconds) the call will block waiting + * for metadata to be retrieved. + * Use 0 for non-blocking calls. + + * @remark Requires broker version >=0.10.0 and api.version.request=true. + * + * @returns the controller broker id (>= 0), or -1 if no ControllerId could be + * retrieved in the allotted timespan. + */ +RD_EXPORT +int32_t rd_kafka_controllerid (rd_kafka_t *rk, int timeout_ms); + + +/** + * @brief Creates a new topic handle for topic named \p topic. + * + * \p conf is an optional configuration for the topic created with + * `rd_kafka_topic_conf_new()` that will be used instead of the default + * topic configuration. + * The \p conf object is freed by this function and must not be used or + * destroyed by the application sub-sequently. + * See `rd_kafka_topic_conf_set()` et.al for more information. + * + * Topic handles are refcounted internally and calling rd_kafka_topic_new() + * again with the same topic name will return the previous topic handle + * without updating the original handle's configuration. + * Applications must eventually call rd_kafka_topic_destroy() for each + * succesfull call to rd_kafka_topic_new() to clear up resources. + * + * @returns the new topic handle or NULL on error (use rd_kafka_errno2err() + * to convert system \p errno to an rd_kafka_resp_err_t error code. + * + * @sa rd_kafka_topic_destroy() + */ +RD_EXPORT +rd_kafka_topic_t *rd_kafka_topic_new(rd_kafka_t *rk, const char *topic, + rd_kafka_topic_conf_t *conf); + + + +/** + * @brief Loose application's topic handle refcount as previously created + * with `rd_kafka_topic_new()`. + * + * @remark Since topic objects are refcounted (both internally and for the app) + * the topic object might not actually be destroyed by this call, + * but the application must consider the object destroyed. + */ +RD_EXPORT +void rd_kafka_topic_destroy(rd_kafka_topic_t *rkt); + + +/** + * @brief Returns the topic name. + */ +RD_EXPORT +const char *rd_kafka_topic_name(const rd_kafka_topic_t *rkt); + + +/** + * @brief Get the \p rkt_opaque pointer that was set in the topic configuration + * with rd_kafka_topic_conf_set_opaque(). + */ +RD_EXPORT +void *rd_kafka_topic_opaque (const rd_kafka_topic_t *rkt); + + +/** + * @brief Unassigned partition. + * + * The unassigned partition is used by the producer API for messages + * that should be partitioned using the configured or default partitioner. + */ +#define RD_KAFKA_PARTITION_UA ((int32_t)-1) + + +/** + * @brief Polls the provided kafka handle for events. + * + * Events will cause application provided callbacks to be called. + * + * The \p timeout_ms argument specifies the maximum amount of time + * (in milliseconds) that the call will block waiting for events. + * For non-blocking calls, provide 0 as \p timeout_ms. + * To wait indefinately for an event, provide -1. + * + * @remark An application should make sure to call poll() at regular + * intervals to serve any queued callbacks waiting to be called. + * @remark If your producer doesn't have any callback set (in particular + * via rd_kafka_conf_set_dr_msg_cb or rd_kafka_conf_set_error_cb) + * you might chose not to call poll(), though this is not + * recommended. + * + * Events: + * - delivery report callbacks (if dr_cb/dr_msg_cb is configured) [producer] + * - error callbacks (rd_kafka_conf_set_error_cb()) [all] + * - stats callbacks (rd_kafka_conf_set_stats_cb()) [all] + * - throttle callbacks (rd_kafka_conf_set_throttle_cb()) [all] + * - OAUTHBEARER token refresh callbacks (rd_kafka_conf_set_oauthbearer_token_refresh_cb()) [all] + * + * @returns the number of events served. + */ +RD_EXPORT +int rd_kafka_poll(rd_kafka_t *rk, int timeout_ms); + + +/** + * @brief Cancels the current callback dispatcher (rd_kafka_poll(), + * rd_kafka_consume_callback(), etc). + * + * A callback may use this to force an immediate return to the calling + * code (caller of e.g. rd_kafka_poll()) without processing any further + * events. + * + * @remark This function MUST ONLY be called from within a librdkafka callback. + */ +RD_EXPORT +void rd_kafka_yield (rd_kafka_t *rk); + + + + +/** + * @brief Pause producing or consumption for the provided list of partitions. + * + * Success or error is returned per-partition \p err in the \p partitions list. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_pause_partitions (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *partitions); + + + +/** + * @brief Resume producing consumption for the provided list of partitions. + * + * Success or error is returned per-partition \p err in the \p partitions list. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_resume_partitions (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *partitions); + + + + +/** + * @brief Query broker for low (oldest/beginning) and high (newest/end) offsets + * for partition. + * + * Offsets are returned in \p *low and \p *high respectively. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an error code on failure. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_query_watermark_offsets (rd_kafka_t *rk, + const char *topic, int32_t partition, + int64_t *low, int64_t *high, int timeout_ms); + + +/** + * @brief Get last known low (oldest/beginning) and high (newest/end) offsets + * for partition. + * + * The low offset is updated periodically (if statistics.interval.ms is set) + * while the high offset is updated on each fetched message set from the broker. + * + * If there is no cached offset (either low or high, or both) then + * RD_KAFKA_OFFSET_INVALID will be returned for the respective offset. + * + * Offsets are returned in \p *low and \p *high respectively. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an error code on failure. + * + * @remark Shall only be used with an active consumer instance. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_get_watermark_offsets (rd_kafka_t *rk, + const char *topic, int32_t partition, + int64_t *low, int64_t *high); + + + +/** + * @brief Look up the offsets for the given partitions by timestamp. + * + * The returned offset for each partition is the earliest offset whose + * timestamp is greater than or equal to the given timestamp in the + * corresponding partition. + * + * The timestamps to query are represented as \c offset in \p offsets + * on input, and \c offset will contain the offset on output. + * + * The function will block for at most \p timeout_ms milliseconds. + * + * @remark Duplicate Topic+Partitions are not supported. + * @remark Per-partition errors may be returned in \c rd_kafka_topic_partition_t.err + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR if offsets were be queried (do note + * that per-partition errors might be set), + * RD_KAFKA_RESP_ERR__TIMED_OUT if not all offsets could be fetched + * within \p timeout_ms, + * RD_KAFKA_RESP_ERR__INVALID_ARG if the \p offsets list is empty, + * RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION if all partitions are unknown, + * RD_KAFKA_RESP_ERR_LEADER_NOT_AVAILABLE if unable to query leaders + * for the given partitions. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_offsets_for_times (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *offsets, + int timeout_ms); + + + +/** + * @brief Allocate and zero memory using the same allocator librdkafka uses. + * + * This is typically an abstraction for the calloc(3) call and makes sure + * the application can use the same memory allocator as librdkafka for + * allocating pointers that are used by librdkafka. + * + * \p rk can be set to return memory allocated by a specific \c rk instance + * otherwise pass NULL for \p rk. + * + * @remark Memory allocated by rd_kafka_mem_calloc() must be freed using + * rd_kafka_mem_free() + */ +RD_EXPORT +void *rd_kafka_mem_calloc (rd_kafka_t *rk, size_t num, size_t size); + + + +/** + * @brief Allocate memory using the same allocator librdkafka uses. + * + * This is typically an abstraction for the malloc(3) call and makes sure + * the application can use the same memory allocator as librdkafka for + * allocating pointers that are used by librdkafka. + * + * \p rk can be set to return memory allocated by a specific \c rk instance + * otherwise pass NULL for \p rk. + * + * @remark Memory allocated by rd_kafka_mem_malloc() must be freed using + * rd_kafka_mem_free() + */ +RD_EXPORT +void *rd_kafka_mem_malloc (rd_kafka_t *rk, size_t size); + + + +/** + * @brief Free pointer returned by librdkafka + * + * This is typically an abstraction for the free(3) call and makes sure + * the application can use the same memory allocator as librdkafka for + * freeing pointers returned by librdkafka. + * + * In standard setups it is usually not necessary to use this interface + * rather than the free(3) functione. + * + * \p rk must be set for memory returned by APIs that take an \c rk argument, + * for other APIs pass NULL for \p rk. + * + * @remark rd_kafka_mem_free() must only be used for pointers returned by APIs + * that explicitly mention using this function for freeing. + */ +RD_EXPORT +void rd_kafka_mem_free (rd_kafka_t *rk, void *ptr); + + +/**@}*/ + + + + + +/** + * @name Queue API + * @{ + * + * Message queues allows the application to re-route consumed messages + * from multiple topic+partitions into one single queue point. + * This queue point containing messages from a number of topic+partitions + * may then be served by a single rd_kafka_consume*_queue() call, + * rather than one call per topic+partition combination. + */ + + +/** + * @brief Create a new message queue. + * + * See rd_kafka_consume_start_queue(), rd_kafka_consume_queue(), et.al. + */ +RD_EXPORT +rd_kafka_queue_t *rd_kafka_queue_new(rd_kafka_t *rk); + +/** + * Destroy a queue, purging all of its enqueued messages. + */ +RD_EXPORT +void rd_kafka_queue_destroy(rd_kafka_queue_t *rkqu); + + +/** + * @returns a reference to the main librdkafka event queue. + * This is the queue served by rd_kafka_poll(). + * + * Use rd_kafka_queue_destroy() to loose the reference. + */ +RD_EXPORT +rd_kafka_queue_t *rd_kafka_queue_get_main (rd_kafka_t *rk); + + +/** + * @returns a reference to the librdkafka consumer queue. + * This is the queue served by rd_kafka_consumer_poll(). + * + * Use rd_kafka_queue_destroy() to loose the reference. + * + * @remark rd_kafka_queue_destroy() MUST be called on this queue + * prior to calling rd_kafka_consumer_close(). + */ +RD_EXPORT +rd_kafka_queue_t *rd_kafka_queue_get_consumer (rd_kafka_t *rk); + +/** + * @returns a reference to the partition's queue, or NULL if + * partition is invalid. + * + * Use rd_kafka_queue_destroy() to loose the reference. + * + * @remark rd_kafka_queue_destroy() MUST be called on this queue + * + * @remark This function only works on consumers. + */ +RD_EXPORT +rd_kafka_queue_t *rd_kafka_queue_get_partition (rd_kafka_t *rk, + const char *topic, + int32_t partition); + +/** + * @returns a reference to the background thread queue, or NULL if the + * background queue is not enabled. + * + * To enable the background thread queue set a generic event handler callback + * with rd_kafka_conf_set_background_event_cb() on the client instance + * configuration object (rd_kafka_conf_t). + * + * The background queue is polled and served by librdkafka and MUST NOT be + * polled, forwarded, or otherwise managed by the application, it may only + * be used as the destination queue passed to queue-enabled APIs, such as + * the Admin API. + * + * The background thread queue provides the application with an automatically + * polled queue that triggers the event callback in a background thread, + * this background thread is completely managed by librdkafka. + * + * Use rd_kafka_queue_destroy() to loose the reference. + * + * @warning The background queue MUST NOT be read from (polled, consumed, etc), + * or forwarded from. + */ +RD_EXPORT +rd_kafka_queue_t *rd_kafka_queue_get_background (rd_kafka_t *rk); + + +/** + * @brief Forward/re-route queue \p src to \p dst. + * If \p dst is \c NULL the forwarding is removed. + * + * The internal refcounts for both queues are increased. + * + * @remark Regardless of whether \p dst is NULL or not, after calling this + * function, \p src will not forward it's fetch queue to the consumer + * queue. + */ +RD_EXPORT +void rd_kafka_queue_forward (rd_kafka_queue_t *src, rd_kafka_queue_t *dst); + +/** + * @brief Forward librdkafka logs (and debug) to the specified queue + * for serving with one of the ..poll() calls. + * + * This allows an application to serve log callbacks (\c log_cb) + * in its thread of choice. + * + * @param rk Client instance. + * @param rkqu Queue to forward logs to. If the value is NULL the logs + * are forwarded to the main queue. + * + * @remark The configuration property \c log.queue MUST also be set to true. + * + * @remark librdkafka maintains its own reference to the provided queue. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an error code on error. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_set_log_queue (rd_kafka_t *rk, + rd_kafka_queue_t *rkqu); + + +/** + * @returns the current number of elements in queue. + */ +RD_EXPORT +size_t rd_kafka_queue_length (rd_kafka_queue_t *rkqu); + + +/** + * @brief Enable IO event triggering for queue. + * + * To ease integration with IO based polling loops this API + * allows an application to create a separate file-descriptor + * that librdkafka will write \p payload (of size \p size) to + * whenever a new element is enqueued on a previously empty queue. + * + * To remove event triggering call with \p fd = -1. + * + * librdkafka will maintain a copy of the \p payload. + * + * @remark IO and callback event triggering are mutually exclusive. + * @remark When using forwarded queues the IO event must only be enabled + * on the final forwarded-to (destination) queue. + * @remark The file-descriptor/socket must be set to non-blocking. + */ +RD_EXPORT +void rd_kafka_queue_io_event_enable (rd_kafka_queue_t *rkqu, int fd, + const void *payload, size_t size); + +/** + * @brief Enable callback event triggering for queue. + * + * The callback will be called from an internal librdkafka thread + * when a new element is enqueued on a previously empty queue. + * + * To remove event triggering call with \p event_cb = NULL. + * + * The \p qev_opaque is passed to the callback's \p qev_opaque argument. + * + * @remark IO and callback event triggering are mutually exclusive. + * @remark Since the callback may be triggered from internal librdkafka + * threads, the application must not perform any pro-longed work in + * the callback, or call any librdkafka APIs (for the same rd_kafka_t + * handle). + */ +RD_EXPORT +void rd_kafka_queue_cb_event_enable (rd_kafka_queue_t *rkqu, + void (*event_cb) (rd_kafka_t *rk, + void *qev_opaque), + void *qev_opaque); + + +/** + * @brief Cancels the current rd_kafka_queue_poll() on \p rkqu. + * + * An application may use this from another thread to force + * an immediate return to the calling code (caller of rd_kafka_queue_poll()). + * Must not be used from signal handlers since that may cause deadlocks. + */ +RD_EXPORT +void rd_kafka_queue_yield (rd_kafka_queue_t *rkqu); + + +/**@}*/ + +/** + * + * @name Simple Consumer API (legacy) + * @{ + * + */ + + +#define RD_KAFKA_OFFSET_BEGINNING -2 /**< Start consuming from beginning of + * kafka partition queue: oldest msg */ +#define RD_KAFKA_OFFSET_END -1 /**< Start consuming from end of kafka + * partition queue: next msg */ +#define RD_KAFKA_OFFSET_STORED -1000 /**< Start consuming from offset retrieved + * from offset store */ +#define RD_KAFKA_OFFSET_INVALID -1001 /**< Invalid offset */ + + +/** @cond NO_DOC */ +#define RD_KAFKA_OFFSET_TAIL_BASE -2000 /* internal: do not use */ +/** @endcond */ + +/** + * @brief Start consuming \p CNT messages from topic's current end offset. + * + * That is, if current end offset is 12345 and \p CNT is 200, it will start + * consuming from offset \c 12345-200 = \c 12145. */ +#define RD_KAFKA_OFFSET_TAIL(CNT) (RD_KAFKA_OFFSET_TAIL_BASE - (CNT)) + +/** + * @brief Start consuming messages for topic \p rkt and \p partition + * at offset \p offset which may either be an absolute \c (0..N) + * or one of the logical offsets: + * - RD_KAFKA_OFFSET_BEGINNING + * - RD_KAFKA_OFFSET_END + * - RD_KAFKA_OFFSET_STORED + * - RD_KAFKA_OFFSET_TAIL + * + * rdkafka will attempt to keep \c queued.min.messages (config property) + * messages in the local queue by repeatedly fetching batches of messages + * from the broker until the threshold is reached. + * + * The application shall use one of the `rd_kafka_consume*()` functions + * to consume messages from the local queue, each kafka message being + * represented as a `rd_kafka_message_t *` object. + * + * `rd_kafka_consume_start()` must not be called multiple times for the same + * topic and partition without stopping consumption first with + * `rd_kafka_consume_stop()`. + * + * @returns 0 on success or -1 on error in which case errno is set accordingly: + * - EBUSY - Conflicts with an existing or previous subscription + * (RD_KAFKA_RESP_ERR__CONFLICT) + * - EINVAL - Invalid offset, or incomplete configuration (lacking group.id) + * (RD_KAFKA_RESP_ERR__INVALID_ARG) + * - ESRCH - requested \p partition is invalid. + * (RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION) + * - ENOENT - topic is unknown in the Kafka cluster. + * (RD_KAFKA_RESP_ERR__UNKNOWN_TOPIC) + * + * Use `rd_kafka_errno2err()` to convert sytem \c errno to `rd_kafka_resp_err_t` + */ +RD_EXPORT +int rd_kafka_consume_start(rd_kafka_topic_t *rkt, int32_t partition, + int64_t offset); + +/** + * @brief Same as rd_kafka_consume_start() but re-routes incoming messages to + * the provided queue \p rkqu (which must have been previously allocated + * with `rd_kafka_queue_new()`. + * + * The application must use one of the `rd_kafka_consume_*_queue()` functions + * to receive fetched messages. + * + * `rd_kafka_consume_start_queue()` must not be called multiple times for the + * same topic and partition without stopping consumption first with + * `rd_kafka_consume_stop()`. + * `rd_kafka_consume_start()` and `rd_kafka_consume_start_queue()` must not + * be combined for the same topic and partition. + */ +RD_EXPORT +int rd_kafka_consume_start_queue(rd_kafka_topic_t *rkt, int32_t partition, + int64_t offset, rd_kafka_queue_t *rkqu); + +/** + * @brief Stop consuming messages for topic \p rkt and \p partition, purging + * all messages currently in the local queue. + * + * NOTE: To enforce synchronisation this call will block until the internal + * fetcher has terminated and offsets are committed to configured + * storage method. + * + * The application needs to be stop all consumers before calling + * `rd_kafka_destroy()` on the main object handle. + * + * @returns 0 on success or -1 on error (see `errno`). + */ +RD_EXPORT +int rd_kafka_consume_stop(rd_kafka_topic_t *rkt, int32_t partition); + + + +/** + * @brief Seek consumer for topic+partition to \p offset which is either an + * absolute or logical offset. + * + * If \p timeout_ms is not 0 the call will wait this long for the + * seek to be performed. If the timeout is reached the internal state + * will be unknown and this function returns `RD_KAFKA_RESP_ERR__TIMED_OUT`. + * If \p timeout_ms is 0 it will initiate the seek but return + * immediately without any error reporting (e.g., async). + * + * This call will purge all pre-fetched messages for the given partition, which + * may be up to \c queued.max.message.kbytes in size. Repeated use of seek + * may thus lead to increased network usage as messages are re-fetched from + * the broker. + * + * @remark Seek must only be performed for already assigned/consumed partitions, + * use rd_kafka_assign() (et.al) to set the initial starting offset + * for a new assignmenmt. + * + * @returns `RD_KAFKA_RESP_ERR__NO_ERROR` on success else an error code. + * + * @deprecated Use rd_kafka_seek_partitions(). + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_seek (rd_kafka_topic_t *rkt, + int32_t partition, + int64_t offset, + int timeout_ms); + + + +/** + * @brief Seek consumer for partitions in \p partitions to the per-partition + * offset in the \c .offset field of \p partitions. + * + * The offset may be either absolute (>= 0) or a logical offset. + * + * If \p timeout_ms is not 0 the call will wait this long for the + * seeks to be performed. If the timeout is reached the internal state + * will be unknown for the remaining partitions to seek and this function + * will return an error with the error code set to + * `RD_KAFKA_RESP_ERR__TIMED_OUT`. + * + * If \p timeout_ms is 0 it will initiate the seek but return + * immediately without any error reporting (e.g., async). + * + * This call will purge all pre-fetched messages for the given partition, which + * may be up to \c queued.max.message.kbytes in size. Repeated use of seek + * may thus lead to increased network usage as messages are re-fetched from + * the broker. + * + * Individual partition errors are reported in the per-partition \c .err field + * of \p partitions. + * + * @remark Seek must only be performed for already assigned/consumed partitions, + * use rd_kafka_assign() (et.al) to set the initial starting offset + * for a new assignmenmt. + * + * @returns NULL on success or an error object on failure. + */ +RD_EXPORT rd_kafka_error_t * +rd_kafka_seek_partitions (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *partitions, + int timeout_ms); + + +/** + * @brief Consume a single message from topic \p rkt and \p partition + * + * \p timeout_ms is maximum amount of time to wait for a message to be received. + * Consumer must have been previously started with `rd_kafka_consume_start()`. + * + * @returns a message object on success or \c NULL on error. + * The message object must be destroyed with `rd_kafka_message_destroy()` + * when the application is done with it. + * + * Errors (when returning NULL): + * - ETIMEDOUT - \p timeout_ms was reached with no new messages fetched. + * - ENOENT - \p rkt + \p partition is unknown. + * (no prior `rd_kafka_consume_start()` call) + * + * NOTE: The returned message's \c ..->err must be checked for errors. + * NOTE: \c ..->err \c == \c RD_KAFKA_RESP_ERR__PARTITION_EOF signals that the + * end of the partition has been reached, which should typically not be + * considered an error. The application should handle this case + * (e.g., ignore). + * + * @remark on_consume() interceptors may be called from this function prior to + * passing message to application. + */ +RD_EXPORT +rd_kafka_message_t *rd_kafka_consume(rd_kafka_topic_t *rkt, int32_t partition, + int timeout_ms); + + + +/** + * @brief Consume up to \p rkmessages_size from topic \p rkt and \p partition + * putting a pointer to each message in the application provided + * array \p rkmessages (of size \p rkmessages_size entries). + * + * `rd_kafka_consume_batch()` provides higher throughput performance + * than `rd_kafka_consume()`. + * + * \p timeout_ms is the maximum amount of time to wait for all of + * \p rkmessages_size messages to be put into \p rkmessages. + * If no messages were available within the timeout period this function + * returns 0 and \p rkmessages remains untouched. + * This differs somewhat from `rd_kafka_consume()`. + * + * The message objects must be destroyed with `rd_kafka_message_destroy()` + * when the application is done with it. + * + * @returns the number of rkmessages added in \p rkmessages, + * or -1 on error (same error codes as for `rd_kafka_consume()`. + * + * @sa rd_kafka_consume() + * + * @remark on_consume() interceptors may be called from this function prior to + * passing message to application. + */ +RD_EXPORT +ssize_t rd_kafka_consume_batch(rd_kafka_topic_t *rkt, int32_t partition, + int timeout_ms, + rd_kafka_message_t **rkmessages, + size_t rkmessages_size); + + + +/** + * @brief Consumes messages from topic \p rkt and \p partition, calling + * the provided callback for each consumed messsage. + * + * `rd_kafka_consume_callback()` provides higher throughput performance + * than both `rd_kafka_consume()` and `rd_kafka_consume_batch()`. + * + * \p timeout_ms is the maximum amount of time to wait for one or more messages + * to arrive. + * + * The provided \p consume_cb function is called for each message, + * the application \b MUST \b NOT call `rd_kafka_message_destroy()` on the + * provided \p rkmessage. + * + * The \p commit_opaque argument is passed to the \p consume_cb + * as \p commit_opaque. + * + * @returns the number of messages processed or -1 on error. + * + * @sa rd_kafka_consume() + * + * @remark on_consume() interceptors may be called from this function prior to + * passing message to application. + * + * @remark This function will return early if a transaction control message is + * received, these messages are not exposed to the application but + * still enqueued on the consumer queue to make sure their + * offsets are stored. + * + * @deprecated This API is deprecated and subject for future removal. + * There is no new callback-based consume interface, use the + * poll/queue based alternatives. + */ +RD_EXPORT +int rd_kafka_consume_callback(rd_kafka_topic_t *rkt, int32_t partition, + int timeout_ms, + void (*consume_cb) (rd_kafka_message_t + *rkmessage, + void *commit_opaque), + void *commit_opaque); + + +/** + * @name Simple Consumer API (legacy): Queue consumers + * @{ + * + * The following `..._queue()` functions are analogue to the functions above + * but reads messages from the provided queue \p rkqu instead. + * \p rkqu must have been previously created with `rd_kafka_queue_new()` + * and the topic consumer must have been started with + * `rd_kafka_consume_start_queue()` utilising the the same queue. + */ + +/** + * @brief Consume from queue + * + * @sa rd_kafka_consume() + */ +RD_EXPORT +rd_kafka_message_t *rd_kafka_consume_queue(rd_kafka_queue_t *rkqu, + int timeout_ms); + +/** + * @brief Consume batch of messages from queue + * + * @sa rd_kafka_consume_batch() + */ +RD_EXPORT +ssize_t rd_kafka_consume_batch_queue(rd_kafka_queue_t *rkqu, + int timeout_ms, + rd_kafka_message_t **rkmessages, + size_t rkmessages_size); + +/** + * @brief Consume multiple messages from queue with callback + * + * @sa rd_kafka_consume_callback() + * + * @deprecated This API is deprecated and subject for future removal. + * There is no new callback-based consume interface, use the + * poll/queue based alternatives. + */ +RD_EXPORT +int rd_kafka_consume_callback_queue (rd_kafka_queue_t *rkqu, + int timeout_ms, + void (*consume_cb) (rd_kafka_message_t + *rkmessage, + void *commit_opaque), + void *commit_opaque); + + +/**@}*/ + + + + +/** + * @name Simple Consumer API (legacy): Topic+partition offset store. + * @{ + * + * If \c auto.commit.enable is true the offset is stored automatically prior to + * returning of the message(s) in each of the rd_kafka_consume*() functions + * above. + */ + + +/** + * @brief Store offset \p offset + 1 for topic \p rkt partition \p partition. + * + * The \c offset + 1 will be committed (written) to broker (or file) according + * to \c `auto.commit.interval.ms` or manual offset-less commit() + * + * @remark \c `enable.auto.offset.store` must be set to "false" when using + * this API. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an error code on error. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_offset_store (rd_kafka_topic_t *rkt, + int32_t partition, int64_t offset); + + +/** + * @brief Store offsets for next auto-commit for one or more partitions. + * + * The offset will be committed (written) to the offset store according + * to \c `auto.commit.interval.ms` or manual offset-less commit(). + * + * Per-partition success/error status propagated through each partition's + * \c .err field. + * + * @remark The \c .offset field is stored as is, it will NOT be + 1. + * + * @remark \c `enable.auto.offset.store` must be set to "false" when using + * this API. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, or + * RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION if none of the + * offsets could be stored, or + * RD_KAFKA_RESP_ERR__INVALID_ARG if \c enable.auto.offset.store + * is true. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_offsets_store (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *offsets); +/**@}*/ + + + + +/** + * @name KafkaConsumer (C) + * @{ + * @brief High-level KafkaConsumer C API + * + * + * + */ + +/** + * @brief Subscribe to topic set using balanced consumer groups. + * + * Wildcard (regex) topics are supported: + * any topic name in the \p topics list that is prefixed with \c \"^\" will + * be regex-matched to the full list of topics in the cluster and matching + * topics will be added to the subscription list. + * + * The full topic list is retrieved every \c topic.metadata.refresh.interval.ms + * to pick up new or delete topics that match the subscription. + * If there is any change to the matched topics the consumer will + * immediately rejoin the group with the updated set of subscribed topics. + * + * Regex and full topic names can be mixed in \p topics. + * + * @remark Only the \c .topic field is used in the supplied \p topics list, + * all other fields are ignored. + * + * @remark subscribe() is an asynchronous method which returns immediately: + * background threads will (re)join the group, wait for group rebalance, + * issue any registered rebalance_cb, assign() the assigned partitions, + * and then start fetching messages. This cycle may take up to + * \c session.timeout.ms * 2 or more to complete. + * + * @remark A consumer error will be raised for each unavailable topic in the + * \p topics. The error will be RD_KAFKA_RESP_ERR_UNKNOWN_TOPIC_OR_PART + * for non-existent topics, and + * RD_KAFKA_RESP_ERR_TOPIC_AUTHORIZATION_FAILED for unauthorized topics. + * The consumer error will be raised through rd_kafka_consumer_poll() + * (et.al.) with the \c rd_kafka_message_t.err field set to one of the + * error codes mentioned above. + * The subscribe function itself is asynchronous and will not return + * an error on unavailable topics. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or + * RD_KAFKA_RESP_ERR__INVALID_ARG if list is empty, contains invalid + * topics or regexes or duplicate entries, + * RD_KAFKA_RESP_ERR__FATAL if the consumer has raised a fatal error. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_subscribe (rd_kafka_t *rk, + const rd_kafka_topic_partition_list_t *topics); + + +/** + * @brief Unsubscribe from the current subscription set. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_unsubscribe (rd_kafka_t *rk); + + +/** + * @brief Returns the current topic subscription + * + * @returns An error code on failure, otherwise \p topic is updated + * to point to a newly allocated topic list (possibly empty). + * + * @remark The application is responsible for calling + * rd_kafka_topic_partition_list_destroy on the returned list. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_subscription (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t **topics); + + + +/** + * @brief Poll the consumer for messages or events. + * + * Will block for at most \p timeout_ms milliseconds. + * + * @remark An application should make sure to call consumer_poll() at regular + * intervals, even if no messages are expected, to serve any + * queued callbacks waiting to be called. This is especially + * important when a rebalance_cb has been registered as it needs + * to be called and handled properly to synchronize internal + * consumer state. + * + * @returns A message object which is a proper message if \p ->err is + * RD_KAFKA_RESP_ERR_NO_ERROR, or an event or error for any other + * value. + * + * @remark on_consume() interceptors may be called from this function prior to + * passing message to application. + * + * @remark When subscribing to topics the application must call poll at + * least every \c max.poll.interval.ms to remain a member of the + * consumer group. + * + * Noteworthy errors returned in \c ->err: + * - RD_KAFKA_RESP_ERR__MAX_POLL_EXCEEDED - application failed to call + * poll within `max.poll.interval.ms`. + * + * @sa rd_kafka_message_t + */ +RD_EXPORT +rd_kafka_message_t *rd_kafka_consumer_poll (rd_kafka_t *rk, int timeout_ms); + +/** + * @brief Close down the KafkaConsumer. + * + * @remark This call will block until the consumer has revoked its assignment, + * calling the \c rebalance_cb if it is configured, committed offsets + * to broker, and left the consumer group. + * The maximum blocking time is roughly limited to session.timeout.ms. + * + * @returns An error code indicating if the consumer close was succesful + * or not. + * RD_KAFKA_RESP_ERR__FATAL is returned if the consumer has raised + * a fatal error. + * + * @remark The application still needs to call rd_kafka_destroy() after + * this call finishes to clean up the underlying handle resources. + * + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_consumer_close (rd_kafka_t *rk); + + +/** + * @brief Incrementally add \p partitions to the current assignment. + * + * If a COOPERATIVE assignor (i.e. incremental rebalancing) is being used, + * this method should be used in a rebalance callback to adjust the current + * assignment appropriately in the case where the rebalance type is + * RD_KAFKA_RESP_ERR__ASSIGN_PARTITIONS. The application must pass the + * partition list passed to the callback (or a copy of it), even if the + * list is empty. \p partitions must not be NULL. This method may also be + * used outside the context of a rebalance callback. + * + * @returns NULL on success, or an error object if the operation was + * unsuccessful. + * + * @remark The returned error object (if not NULL) must be destroyed with + * rd_kafka_error_destroy(). + */ +RD_EXPORT rd_kafka_error_t * +rd_kafka_incremental_assign (rd_kafka_t *rk, + const rd_kafka_topic_partition_list_t + *partitions); + + +/** + * @brief Incrementally remove \p partitions from the current assignment. + * + * If a COOPERATIVE assignor (i.e. incremental rebalancing) is being used, + * this method should be used in a rebalance callback to adjust the current + * assignment appropriately in the case where the rebalance type is + * RD_KAFKA_RESP_ERR__REVOKE_PARTITIONS. The application must pass the + * partition list passed to the callback (or a copy of it), even if the + * list is empty. \p partitions must not be NULL. This method may also be + * used outside the context of a rebalance callback. + * + * @returns NULL on success, or an error object if the operation was + * unsuccessful. + * + * @remark The returned error object (if not NULL) must be destroyed with + * rd_kafka_error_destroy(). + */ +RD_EXPORT rd_kafka_error_t * +rd_kafka_incremental_unassign (rd_kafka_t *rk, + const rd_kafka_topic_partition_list_t + *partitions); + + +/** + * @brief The rebalance protocol currently in use. This will be + * "NONE" if the consumer has not (yet) joined a group, else it will + * match the rebalance protocol ("EAGER", "COOPERATIVE") of the + * configured and selected assignor(s). All configured + * assignors must have the same protocol type, meaning + * online migration of a consumer group from using one + * protocol to another (in particular upgading from EAGER + * to COOPERATIVE) without a restart is not currently + * supported. + * + * @returns NULL on error, or one of "NONE", "EAGER", "COOPERATIVE" on success. + */ +RD_EXPORT +const char *rd_kafka_rebalance_protocol (rd_kafka_t *rk); + + +/** + * @brief Atomic assignment of partitions to consume. + * + * The new \p partitions will replace the existing assignment. + * + * A zero-length \p partitions will treat the partitions as a valid, + * albeit empty assignment, and maintain internal state, while a \c NULL + * value for \p partitions will reset and clear the internal state. + * + * When used from a rebalance callback, the application should pass the + * partition list passed to the callback (or a copy of it) even if the list + * is empty (i.e. should not pass NULL in this case) so as to maintain + * internal join state. This is not strictly required - the application + * may adjust the assignment provided by the group. However, this is rarely + * useful in practice. + * + * @returns An error code indicating if the new assignment was applied or not. + * RD_KAFKA_RESP_ERR__FATAL is returned if the consumer has raised + * a fatal error. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_assign (rd_kafka_t *rk, + const rd_kafka_topic_partition_list_t *partitions); + +/** + * @brief Returns the current partition assignment as set by rd_kafka_assign() + * or rd_kafka_incremental_assign(). + * + * @returns An error code on failure, otherwise \p partitions is updated + * to point to a newly allocated partition list (possibly empty). + * + * @remark The application is responsible for calling + * rd_kafka_topic_partition_list_destroy on the returned list. + * + * @remark This assignment represents the partitions assigned through the + * assign functions and not the partitions assigned to this consumer + * instance by the consumer group leader. + * They are usually the same following a rebalance but not necessarily + * since an application is free to assign any partitions. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_assignment (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t **partitions); + + +/** + * @brief Check whether the consumer considers the current assignment to + * have been lost involuntarily. This method is only applicable for + * use with a high level subscribing consumer. Assignments are revoked + * immediately when determined to have been lost, so this method + * is only useful when reacting to a RD_KAFKA_EVENT_REBALANCE event + * or from within a rebalance_cb. Partitions that have been lost may + * already be owned by other members in the group and therefore + * commiting offsets, for example, may fail. + * + * @remark Calling rd_kafka_assign(), rd_kafka_incremental_assign() or + * rd_kafka_incremental_unassign() resets this flag. + * + * @returns Returns 1 if the current partition assignment is considered + * lost, 0 otherwise. + */ +RD_EXPORT int +rd_kafka_assignment_lost (rd_kafka_t *rk); + + +/** + * @brief Commit offsets on broker for the provided list of partitions. + * + * \p offsets should contain \c topic, \c partition, \c offset and possibly + * \c metadata. The \c offset should be the offset where consumption will + * resume, i.e., the last processed offset + 1. + * If \p offsets is NULL the current partition assignment will be used instead. + * + * If \p async is false this operation will block until the broker offset commit + * is done, returning the resulting success or error code. + * + * If a rd_kafka_conf_set_offset_commit_cb() offset commit callback has been + * configured the callback will be enqueued for a future call to + * rd_kafka_poll(), rd_kafka_consumer_poll() or similar. + * + * @returns An error code indiciating if the commit was successful, + * or successfully scheduled if asynchronous, or failed. + * RD_KAFKA_RESP_ERR__FATAL is returned if the consumer has raised + * a fatal error. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_commit (rd_kafka_t *rk, const rd_kafka_topic_partition_list_t *offsets, + int async); + + +/** + * @brief Commit message's offset on broker for the message's partition. + * The committed offset is the message's offset + 1. + * + * @sa rd_kafka_commit + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_commit_message (rd_kafka_t *rk, const rd_kafka_message_t *rkmessage, + int async); + + +/** + * @brief Commit offsets on broker for the provided list of partitions. + * + * See rd_kafka_commit for \p offsets semantics. + * + * The result of the offset commit will be posted on the provided \p rkqu queue. + * + * If the application uses one of the poll APIs (rd_kafka_poll(), + * rd_kafka_consumer_poll(), rd_kafka_queue_poll(), ..) to serve the queue + * the \p cb callback is required. + * + * The \p commit_opaque argument is passed to the callback as \p commit_opaque, + * or if using the event API the callback is ignored and the offset commit + * result will be returned as an RD_KAFKA_EVENT_COMMIT event and the + * \p commit_opaque value will be available with rd_kafka_event_opaque(). + * + * If \p rkqu is NULL a temporary queue will be created and the callback will + * be served by this call. + * + * @sa rd_kafka_commit() + * @sa rd_kafka_conf_set_offset_commit_cb() + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_commit_queue (rd_kafka_t *rk, + const rd_kafka_topic_partition_list_t *offsets, + rd_kafka_queue_t *rkqu, + void (*cb) (rd_kafka_t *rk, + rd_kafka_resp_err_t err, + rd_kafka_topic_partition_list_t *offsets, + void *commit_opaque), + void *commit_opaque); + + +/** + * @brief Retrieve committed offsets for topics+partitions. + * + * The \p offset field of each requested partition will either be set to + * stored offset or to RD_KAFKA_OFFSET_INVALID in case there was no stored + * offset for that partition. + * + * Committed offsets will be returned according to the `isolation.level` + * configuration property, if set to `read_committed` (default) then only + * stable offsets for fully committed transactions will be returned, while + * `read_uncommitted` may return offsets for not yet committed transactions. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success in which case the + * \p offset or \p err field of each \p partitions' element is filled + * in with the stored offset, or a partition specific error. + * Else returns an error code. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_committed (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *partitions, + int timeout_ms); + + + +/** + * @brief Retrieve current positions (offsets) for topics+partitions. + * + * The \p offset field of each requested partition will be set to the offset + * of the last consumed message + 1, or RD_KAFKA_OFFSET_INVALID in case there was + * no previous message. + * + * @remark In this context the last consumed message is the offset consumed + * by the current librdkafka instance and, in case of rebalancing, not + * necessarily the last message fetched from the partition. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success in which case the + * \p offset or \p err field of each \p partitions' element is filled + * in with the stored offset, or a partition specific error. + * Else returns an error code. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_position (rd_kafka_t *rk, + rd_kafka_topic_partition_list_t *partitions); + + + + +/** + * @returns the current consumer group metadata associated with this consumer, + * or NULL if \p rk is not a consumer configured with a \c group.id. + * This metadata object should be passed to the transactional + * producer's rd_kafka_send_offsets_to_transaction() API. + * + * @remark The returned pointer must be freed by the application using + * rd_kafka_consumer_group_metadata_destroy(). + * + * @sa rd_kafka_send_offsets_to_transaction() + */ +RD_EXPORT rd_kafka_consumer_group_metadata_t * +rd_kafka_consumer_group_metadata (rd_kafka_t *rk); + + +/** + * @brief Create a new consumer group metadata object. + * This is typically only used for writing tests. + * + * @param group_id The group id. + * + * @remark The returned pointer must be freed by the application using + * rd_kafka_consumer_group_metadata_destroy(). + */ +RD_EXPORT rd_kafka_consumer_group_metadata_t * +rd_kafka_consumer_group_metadata_new (const char *group_id); + + +/** + * @brief Create a new consumer group metadata object. + * This is typically only used for writing tests. + * + * @param group_id The group id. + * @param generation_id The group generation id. + * @param member_id The group member id. + * @param group_instance_id The group instance id (may be NULL). + * + * @remark The returned pointer must be freed by the application using + * rd_kafka_consumer_group_metadata_destroy(). + */ +RD_EXPORT rd_kafka_consumer_group_metadata_t * +rd_kafka_consumer_group_metadata_new_with_genid (const char *group_id, + int32_t generation_id, + const char *member_id, + const char + *group_instance_id); + + +/** + * @brief Frees the consumer group metadata object as returned by + * rd_kafka_consumer_group_metadata(). + */ +RD_EXPORT void +rd_kafka_consumer_group_metadata_destroy (rd_kafka_consumer_group_metadata_t *); + + +/** + * @brief Serialize the consumer group metadata to a binary format. + * This is mainly for client binding use and not for application use. + * + * @remark The serialized metadata format is private and is not compatible + * across different versions or even builds of librdkafka. + * It should only be used in the same process runtime and must only + * be passed to rd_kafka_consumer_group_metadata_read(). + * + * @param cgmd Metadata to be serialized. + * @param bufferp On success this pointer will be updated to point to na + * allocated buffer containing the serialized metadata. + * The buffer must be freed with rd_kafka_mem_free(). + * @param sizep The pointed to size will be updated with the size of + * the serialized buffer. + * + * @returns NULL on success or an error object on failure. + * + * @sa rd_kafka_consumer_group_metadata_read() + */ +RD_EXPORT rd_kafka_error_t * +rd_kafka_consumer_group_metadata_write ( + const rd_kafka_consumer_group_metadata_t *cgmd, + void **bufferp, size_t *sizep); + +/** + * @brief Reads serialized consumer group metadata and returns a + * consumer group metadata object. + * This is mainly for client binding use and not for application use. + * + * @remark The serialized metadata format is private and is not compatible + * across different versions or even builds of librdkafka. + * It should only be used in the same process runtime and must only + * be passed to rd_kafka_consumer_group_metadata_read(). + * + * @param cgmdp On success this pointer will be updated to point to a new + * consumer group metadata object which must be freed with + * rd_kafka_consumer_group_metadata_destroy(). + * @param buffer Pointer to the serialized data. + * @param size Size of the serialized data. + * + * @returns NULL on success or an error object on failure. + * + * @sa rd_kafka_consumer_group_metadata_write() + */ +RD_EXPORT rd_kafka_error_t * +rd_kafka_consumer_group_metadata_read ( + rd_kafka_consumer_group_metadata_t **cgmdp, + const void *buffer, size_t size); + +/**@}*/ + + + +/** + * @name Producer API + * @{ + * + * + */ + + +/** + * @brief Producer message flags + */ +#define RD_KAFKA_MSG_F_FREE 0x1 /**< Delegate freeing of payload to rdkafka. */ +#define RD_KAFKA_MSG_F_COPY 0x2 /**< rdkafka will make a copy of the payload. */ +#define RD_KAFKA_MSG_F_BLOCK 0x4 /**< Block produce*() on message queue full. + * WARNING: If a delivery report callback + * is used the application MUST + * call rd_kafka_poll() (or equiv.) + * to make sure delivered messages + * are drained from the internal + * delivery report queue. + * Failure to do so will result + * in indefinately blocking on + * the produce() call when the + * message queue is full. */ +#define RD_KAFKA_MSG_F_PARTITION 0x8 /**< produce_batch() will honor + * per-message partition. */ + + + +/** + * @brief Produce and send a single message to broker. + * + * \p rkt is the target topic which must have been previously created with + * `rd_kafka_topic_new()`. + * + * `rd_kafka_produce()` is an asynch non-blocking API. + * See `rd_kafka_conf_set_dr_msg_cb` on how to setup a callback to be called + * once the delivery status (success or failure) is known. The delivery report + * is trigged by the application calling `rd_kafka_poll()` (at regular + * intervals) or `rd_kafka_flush()` (at termination). + * + * Since producing is asynchronous, you should call `rd_kafka_flush()` before + * you destroy the producer. Otherwise, any outstanding messages will be + * silently discarded. + * + * When temporary errors occur, librdkafka automatically retries to produce the + * messages. Retries are triggered after retry.backoff.ms and when the + * leader broker for the given partition is available. Otherwise, librdkafka + * falls back to polling the topic metadata to monitor when a new leader is + * elected (see the topic.metadata.refresh.fast.interval.ms and + * topic.metadata.refresh.interval.ms configurations) and then performs a + * retry. A delivery error will occur if the message could not be produced + * within message.timeout.ms. + * + * See the "Message reliability" chapter in INTRODUCTION.md for more + * information. + * + * \p partition is the target partition, either: + * - RD_KAFKA_PARTITION_UA (unassigned) for + * automatic partitioning using the topic's partitioner function, or + * - a fixed partition (0..N) + * + * \p msgflags is zero or more of the following flags OR:ed together: + * RD_KAFKA_MSG_F_BLOCK - block \p produce*() call if + * \p queue.buffering.max.messages or + * \p queue.buffering.max.kbytes are exceeded. + * Messages are considered in-queue from the point they + * are accepted by produce() until their corresponding + * delivery report callback/event returns. + * It is thus a requirement to call + * rd_kafka_poll() (or equiv.) from a separate + * thread when F_BLOCK is used. + * See WARNING on \c RD_KAFKA_MSG_F_BLOCK above. + * + * RD_KAFKA_MSG_F_FREE - rdkafka will free(3) \p payload when it is done + * with it. + * RD_KAFKA_MSG_F_COPY - the \p payload data will be copied and the + * \p payload pointer will not be used by rdkafka + * after the call returns. + * RD_KAFKA_MSG_F_PARTITION - produce_batch() will honour per-message + * partition, either set manually or by the + * configured partitioner. + * + * .._F_FREE and .._F_COPY are mutually exclusive. If neither of these are + * set, the caller must ensure that the memory backing \p payload remains + * valid and is not modified or reused until the delivery callback is + * invoked. Other buffers passed to `rd_kafka_produce()` don't have this + * restriction on reuse, i.e. the memory backing the key or the topic name + * may be reused as soon as `rd_kafka_produce()` returns. + * + * If the function returns -1 and RD_KAFKA_MSG_F_FREE was specified, then + * the memory associated with the payload is still the caller's + * responsibility. + * + * \p payload is the message payload of size \p len bytes. + * + * \p key is an optional message key of size \p keylen bytes, if non-NULL it + * will be passed to the topic partitioner as well as be sent with the + * message to the broker and passed on to the consumer. + * + * \p msg_opaque is an optional application-provided per-message opaque + * pointer that will provided in the message's delivery report callback + * (\c dr_msg_cb or \c dr_cb) and the \c rd_kafka_message_t \c _private field. + * + * @remark on_send() and on_acknowledgement() interceptors may be called + * from this function. on_acknowledgement() will only be called if the + * message fails partitioning. + * + * @remark If the producer is transactional (\c transactional.id is configured) + * producing is only allowed during an on-going transaction, namely + * after rd_kafka_begin_transaction() has been called. + * + * @returns 0 on success or -1 on error in which case errno is set accordingly: + * - ENOBUFS - maximum number of outstanding messages has been reached: + * "queue.buffering.max.messages" + * (RD_KAFKA_RESP_ERR__QUEUE_FULL) + * - EMSGSIZE - message is larger than configured max size: + * "messages.max.bytes". + * (RD_KAFKA_RESP_ERR_MSG_SIZE_TOO_LARGE) + * - ESRCH - requested \p partition is unknown in the Kafka cluster. + * (RD_KAFKA_RESP_ERR__UNKNOWN_PARTITION) + * - ENOENT - topic is unknown in the Kafka cluster. + * (RD_KAFKA_RESP_ERR__UNKNOWN_TOPIC) + * - ECANCELED - fatal error has been raised on producer, see + * rd_kafka_fatal_error(), + * (RD_KAFKA_RESP_ERR__FATAL). + * - ENOEXEC - transactional state forbids producing + * (RD_KAFKA_RESP_ERR__STATE) + * + * @sa Use rd_kafka_errno2err() to convert `errno` to rdkafka error code. + */ +RD_EXPORT +int rd_kafka_produce(rd_kafka_topic_t *rkt, int32_t partition, + int msgflags, + void *payload, size_t len, + const void *key, size_t keylen, + void *msg_opaque); + + +/** + * @brief Produce and send a single message to broker. + * + * The message is defined by a va-arg list using \c rd_kafka_vtype_t + * tag tuples which must be terminated with a single \c RD_KAFKA_V_END. + * + * @returns \c RD_KAFKA_RESP_ERR_NO_ERROR on success, else an error code as + * described in rd_kafka_produce(). + * \c RD_KAFKA_RESP_ERR__CONFLICT is returned if _V_HEADER and + * _V_HEADERS are mixed. + * + * @sa rd_kafka_produce, rd_kafka_produceva, RD_KAFKA_V_END + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_producev (rd_kafka_t *rk, ...); + + +/** + * @brief Produce and send a single message to broker. + * + * The message is defined by an array of \c rd_kafka_vu_t of + * count \p cnt. + * + * @returns an error object on failure or NULL on success. + * See rd_kafka_producev() for specific error codes. + * + * @sa rd_kafka_produce, rd_kafka_producev, RD_KAFKA_V_END + */ +RD_EXPORT +rd_kafka_error_t *rd_kafka_produceva (rd_kafka_t *rk, + const rd_kafka_vu_t *vus, + size_t cnt); + + +/** + * @brief Produce multiple messages. + * + * If partition is RD_KAFKA_PARTITION_UA the configured partitioner will + * be run for each message (slower), otherwise the messages will be enqueued + * to the specified partition directly (faster). + * + * The messages are provided in the array \p rkmessages of count \p message_cnt + * elements. + * The \p partition and \p msgflags are used for all provided messages. + * + * Honoured \p rkmessages[] fields are: + * - payload,len Message payload and length + * - key,key_len Optional message key + * - _private Message opaque pointer (msg_opaque) + * - err Will be set according to success or failure, see + * rd_kafka_produce() for possible error codes. + * Application only needs to check for errors if + * return value != \p message_cnt. + * + * @remark If \c RD_KAFKA_MSG_F_PARTITION is set in \p msgflags, the + * \c .partition field of the \p rkmessages is used instead of + * \p partition. + * + * @returns the number of messages succesfully enqueued for producing. + * + * @remark This interface does NOT support setting message headers on + * the provided \p rkmessages. + */ +RD_EXPORT +int rd_kafka_produce_batch(rd_kafka_topic_t *rkt, int32_t partition, + int msgflags, + rd_kafka_message_t *rkmessages, int message_cnt); + + + + +/** + * @brief Wait until all outstanding produce requests, et.al, are completed. + * This should typically be done prior to destroying a producer instance + * to make sure all queued and in-flight produce requests are completed + * before terminating. + * + * @remark This function will call rd_kafka_poll() and thus trigger callbacks. + * + * @remark The \c linger.ms time will be ignored for the duration of the call, + * queued messages will be sent to the broker as soon as possible. + * + * @remark If RD_KAFKA_EVENT_DR has been enabled + * (through rd_kafka_conf_set_events()) this function will not call + * rd_kafka_poll() but instead wait for the librdkafka-handled + * message count to reach zero. This requires the application to + * serve the event queue in a separate thread. + * In this mode only messages are counted, not other types of + * queued events. + * + * @returns RD_KAFKA_RESP_ERR__TIMED_OUT if \p timeout_ms was reached before all + * outstanding requests were completed, else RD_KAFKA_RESP_ERR_NO_ERROR + * + * @sa rd_kafka_outq_len() + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_flush (rd_kafka_t *rk, int timeout_ms); + + + +/** + * @brief Purge messages currently handled by the producer instance. + * + * @param rk Client instance. + * @param purge_flags Tells which messages to purge and how. + * + * The application will need to call rd_kafka_poll() or rd_kafka_flush() + * afterwards to serve the delivery report callbacks of the purged messages. + * + * Messages purged from internal queues fail with the delivery report + * error code set to RD_KAFKA_RESP_ERR__PURGE_QUEUE, while purged messages that + * are in-flight to or from the broker will fail with the error code set to + * RD_KAFKA_RESP_ERR__PURGE_INFLIGHT. + * + * @warning Purging messages that are in-flight to or from the broker + * will ignore any sub-sequent acknowledgement for these messages + * received from the broker, effectively making it impossible + * for the application to know if the messages were successfully + * produced or not. This may result in duplicate messages if the + * application retries these messages at a later time. + * + * @remark This call may block for a short time while background thread + * queues are purged. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, + * RD_KAFKA_RESP_ERR__INVALID_ARG if the \p purge flags are invalid + * or unknown, + * RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED if called on a non-producer + * client instance. + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_purge (rd_kafka_t *rk, int purge_flags); + + +/** + * @brief Flags for rd_kafka_purge() + */ + +/*! + * Purge messages in internal queues. + */ +#define RD_KAFKA_PURGE_F_QUEUE 0x1 + +/*! + * Purge messages in-flight to or from the broker. + * Purging these messages will void any future acknowledgements from the + * broker, making it impossible for the application to know if these + * messages were successfully delivered or not. + * Retrying these messages may lead to duplicates. + */ +#define RD_KAFKA_PURGE_F_INFLIGHT 0x2 + + +/*! + * Don't wait for background thread queue purging to finish. + */ +#define RD_KAFKA_PURGE_F_NON_BLOCKING 0x4 + + +/**@}*/ + + +/** +* @name Metadata API +* @{ +* +* +*/ + + +/** + * @brief Broker information + */ +typedef struct rd_kafka_metadata_broker { + int32_t id; /**< Broker Id */ + char *host; /**< Broker hostname */ + int port; /**< Broker listening port */ +} rd_kafka_metadata_broker_t; + +/** + * @brief Partition information + */ +typedef struct rd_kafka_metadata_partition { + int32_t id; /**< Partition Id */ + rd_kafka_resp_err_t err; /**< Partition error reported by broker */ + int32_t leader; /**< Leader broker */ + int replica_cnt; /**< Number of brokers in \p replicas */ + int32_t *replicas; /**< Replica brokers */ + int isr_cnt; /**< Number of ISR brokers in \p isrs */ + int32_t *isrs; /**< In-Sync-Replica brokers */ +} rd_kafka_metadata_partition_t; + +/** + * @brief Topic information + */ +typedef struct rd_kafka_metadata_topic { + char *topic; /**< Topic name */ + int partition_cnt; /**< Number of partitions in \p partitions*/ + struct rd_kafka_metadata_partition *partitions; /**< Partitions */ + rd_kafka_resp_err_t err; /**< Topic error reported by broker */ +} rd_kafka_metadata_topic_t; + + +/** + * @brief Metadata container + */ +typedef struct rd_kafka_metadata { + int broker_cnt; /**< Number of brokers in \p brokers */ + struct rd_kafka_metadata_broker *brokers; /**< Brokers */ + + int topic_cnt; /**< Number of topics in \p topics */ + struct rd_kafka_metadata_topic *topics; /**< Topics */ + + int32_t orig_broker_id; /**< Broker originating this metadata */ + char *orig_broker_name; /**< Name of originating broker */ +} rd_kafka_metadata_t; + + +/** + * @brief Request Metadata from broker. + * + * Parameters: + * - \p all_topics if non-zero: request info about all topics in cluster, + * if zero: only request info about locally known topics. + * - \p only_rkt only request info about this topic + * - \p metadatap pointer to hold metadata result. + * The \p *metadatap pointer must be released + * with rd_kafka_metadata_destroy(). + * - \p timeout_ms maximum response time before failing. + * + * @remark Consumer: If \p all_topics is non-zero the Metadata response + * information may trigger a re-join if any subscribed topics + * have changed partition count or existence state. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success (in which case *metadatap) + * will be set, else RD_KAFKA_RESP_ERR__TIMED_OUT on timeout or + * other error code on error. + */ +RD_EXPORT +rd_kafka_resp_err_t +rd_kafka_metadata (rd_kafka_t *rk, int all_topics, + rd_kafka_topic_t *only_rkt, + const struct rd_kafka_metadata **metadatap, + int timeout_ms); + +/** + * @brief Release metadata memory. + */ +RD_EXPORT +void rd_kafka_metadata_destroy(const struct rd_kafka_metadata *metadata); + + +/**@}*/ + + + +/** +* @name Client group information +* @{ +* +* +*/ + + +/** + * @brief Group member information + * + * For more information on \p member_metadata format, see + * https://cwiki.apache.org/confluence/display/KAFKA/A+Guide+To+The+Kafka+Protocol#AGuideToTheKafkaProtocol-GroupMembershipAPI + * + */ +struct rd_kafka_group_member_info { + char *member_id; /**< Member id (generated by broker) */ + char *client_id; /**< Client's \p client.id */ + char *client_host; /**< Client's hostname */ + void *member_metadata; /**< Member metadata (binary), + * format depends on \p protocol_type. */ + int member_metadata_size; /**< Member metadata size in bytes */ + void *member_assignment; /**< Member assignment (binary), + * format depends on \p protocol_type. */ + int member_assignment_size; /**< Member assignment size in bytes */ +}; + +/** + * @brief Group information + */ +struct rd_kafka_group_info { + struct rd_kafka_metadata_broker broker; /**< Originating broker info */ + char *group; /**< Group name */ + rd_kafka_resp_err_t err; /**< Broker-originated error */ + char *state; /**< Group state */ + char *protocol_type; /**< Group protocol type */ + char *protocol; /**< Group protocol */ + struct rd_kafka_group_member_info *members; /**< Group members */ + int member_cnt; /**< Group member count */ +}; + +/** + * @brief List of groups + * + * @sa rd_kafka_group_list_destroy() to release list memory. + */ +struct rd_kafka_group_list { + struct rd_kafka_group_info *groups; /**< Groups */ + int group_cnt; /**< Group count */ +}; + + +/** + * @brief List and describe client groups in cluster. + * + * \p group is an optional group name to describe, otherwise (\p NULL) all + * groups are returned. + * + * \p timeout_ms is the (approximate) maximum time to wait for response + * from brokers and must be a positive value. + * + * @returns \c RD_KAFKA_RESP_ERR__NO_ERROR on success and \p grplistp is + * updated to point to a newly allocated list of groups. + * \c RD_KAFKA_RESP_ERR__PARTIAL if not all brokers responded + * in time but at least one group is returned in \p grplistlp. + * \c RD_KAFKA_RESP_ERR__TIMED_OUT if no groups were returned in the + * given timeframe but not all brokers have yet responded, or + * if the list of brokers in the cluster could not be obtained within + * the given timeframe. + * \c RD_KAFKA_RESP_ERR__TRANSPORT if no brokers were found. + * Other error codes may also be returned from the request layer. + * + * The \p grplistp remains untouched if any error code is returned, + * with the exception of RD_KAFKA_RESP_ERR__PARTIAL which behaves + * as RD_KAFKA_RESP_ERR__NO_ERROR (success) but with an incomplete + * group list. + * + * @sa Use rd_kafka_group_list_destroy() to release list memory. + */ +RD_EXPORT +rd_kafka_resp_err_t +rd_kafka_list_groups (rd_kafka_t *rk, const char *group, + const struct rd_kafka_group_list **grplistp, + int timeout_ms); + +/** + * @brief Release list memory + */ +RD_EXPORT +void rd_kafka_group_list_destroy (const struct rd_kafka_group_list *grplist); + + +/**@}*/ + + + +/** + * @name Miscellaneous APIs + * @{ + * + */ + + +/** + * @brief Adds one or more brokers to the kafka handle's list of initial + * bootstrap brokers. + * + * Additional brokers will be discovered automatically as soon as rdkafka + * connects to a broker by querying the broker metadata. + * + * If a broker name resolves to multiple addresses (and possibly + * address families) all will be used for connection attempts in + * round-robin fashion. + * + * \p brokerlist is a ,-separated list of brokers in the format: + * \c \,\,.. + * Where each broker is in either the host or URL based format: + * \c \[:\] + * \c \://\[:port] + * \c \ is either \c PLAINTEXT, \c SSL, \c SASL, \c SASL_PLAINTEXT + * The two formats can be mixed but ultimately the value of the + * `security.protocol` config property decides what brokers are allowed. + * + * Example: + * brokerlist = "broker1:10000,broker2" + * brokerlist = "SSL://broker3:9000,ssl://broker2" + * + * @returns the number of brokers successfully added. + * + * @remark Brokers may also be defined with the \c metadata.broker.list or + * \c bootstrap.servers configuration property (preferred method). + * + * @deprecated Set bootstrap servers with the \c bootstrap.servers + * configuration property. + */ +RD_EXPORT +int rd_kafka_brokers_add(rd_kafka_t *rk, const char *brokerlist); + + + + +/** + * @brief Set logger function. + * + * The default is to print to stderr, but a syslog logger is also available, + * see rd_kafka_log_(print|syslog) for the builtin alternatives. + * Alternatively the application may provide its own logger callback. + * Or pass 'func' as NULL to disable logging. + * + * @deprecated Use rd_kafka_conf_set_log_cb() + * + * @remark \p rk may be passed as NULL in the callback. + */ +RD_EXPORT RD_DEPRECATED +void rd_kafka_set_logger(rd_kafka_t *rk, + void (*func) (const rd_kafka_t *rk, int level, + const char *fac, const char *buf)); + + +/** + * @brief Specifies the maximum logging level emitted by + * internal kafka logging and debugging. + * + * @deprecated Set the \c "log_level" configuration property instead. + * + * @remark If the \p \"debug\" configuration property is set the log level is + * automatically adjusted to \c LOG_DEBUG (7). + */ +RD_EXPORT +void rd_kafka_set_log_level(rd_kafka_t *rk, int level); + + +/** + * @brief Builtin (default) log sink: print to stderr + */ +RD_EXPORT +void rd_kafka_log_print(const rd_kafka_t *rk, int level, + const char *fac, const char *buf); + + +/** + * @brief Builtin log sink: print to syslog. + * @remark This logger is only available if librdkafka was built + * with syslog support. + */ +RD_EXPORT +void rd_kafka_log_syslog(const rd_kafka_t *rk, int level, + const char *fac, const char *buf); + + +/** + * @brief Returns the current out queue length. + * + * The out queue length is the sum of: + * - number of messages waiting to be sent to, or acknowledged by, + * the broker. + * - number of delivery reports (e.g., dr_msg_cb) waiting to be served + * by rd_kafka_poll() or rd_kafka_flush(). + * - number of callbacks (e.g., error_cb, stats_cb, etc) waiting to be + * served by rd_kafka_poll(), rd_kafka_consumer_poll() or rd_kafka_flush(). + * - number of events waiting to be served by background_event_cb() in + * the background queue (see rd_kafka_conf_set_background_event_cb). + * + * An application should wait for the return value of this function to reach + * zero before terminating to make sure outstanding messages, + * requests (such as offset commits), callbacks and events are fully processed. + * See rd_kafka_flush(). + * + * @returns number of messages and events waiting in queues. + * + * @sa rd_kafka_flush() + */ +RD_EXPORT +int rd_kafka_outq_len(rd_kafka_t *rk); + + + +/** + * @brief Dumps rdkafka's internal state for handle \p rk to stream \p fp + * + * This is only useful for debugging rdkafka, showing state and statistics + * for brokers, topics, partitions, etc. + */ +RD_EXPORT +void rd_kafka_dump(FILE *fp, rd_kafka_t *rk); + + + +/** + * @brief Retrieve the current number of threads in use by librdkafka. + * + * Used by regression tests. + */ +RD_EXPORT +int rd_kafka_thread_cnt(void); + + +/** + * @enum rd_kafka_thread_type_t + * + * @brief librdkafka internal thread type. + * + * @sa rd_kafka_interceptor_add_on_thread_start() + */ +typedef enum rd_kafka_thread_type_t { + RD_KAFKA_THREAD_MAIN, /**< librdkafka's internal main thread */ + RD_KAFKA_THREAD_BACKGROUND, /**< Background thread (if enabled) */ + RD_KAFKA_THREAD_BROKER /**< Per-broker thread */ +} rd_kafka_thread_type_t; + + +/** + * @brief Wait for all rd_kafka_t objects to be destroyed. + * + * Returns 0 if all kafka objects are now destroyed, or -1 if the + * timeout was reached. + * + * @remark This function is deprecated. + */ +RD_EXPORT +int rd_kafka_wait_destroyed(int timeout_ms); + + +/** + * @brief Run librdkafka's built-in unit-tests. + * + * @returns the number of failures, or 0 if all tests passed. + */ +RD_EXPORT +int rd_kafka_unittest (void); + + +/**@}*/ + + + + +/** + * @name Experimental APIs + * @{ + */ + +/** + * @brief Redirect the main (rd_kafka_poll()) queue to the KafkaConsumer's + * queue (rd_kafka_consumer_poll()). + * + * @warning It is not permitted to call rd_kafka_poll() after directing the + * main queue with rd_kafka_poll_set_consumer(). + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_poll_set_consumer (rd_kafka_t *rk); + + +/**@}*/ + +/** + * @name Event interface + * + * @brief The event API provides an alternative pollable non-callback interface + * to librdkafka's message and event queues. + * + * @{ + */ + + +/** + * @brief Event types + */ +typedef int rd_kafka_event_type_t; +#define RD_KAFKA_EVENT_NONE 0x0 /**< Unset value */ +#define RD_KAFKA_EVENT_DR 0x1 /**< Producer Delivery report batch */ +#define RD_KAFKA_EVENT_FETCH 0x2 /**< Fetched message (consumer) */ +#define RD_KAFKA_EVENT_LOG 0x4 /**< Log message */ +#define RD_KAFKA_EVENT_ERROR 0x8 /**< Error */ +#define RD_KAFKA_EVENT_REBALANCE 0x10 /**< Group rebalance (consumer) */ +#define RD_KAFKA_EVENT_OFFSET_COMMIT 0x20 /**< Offset commit result */ +#define RD_KAFKA_EVENT_STATS 0x40 /**< Stats */ +#define RD_KAFKA_EVENT_CREATETOPICS_RESULT 100 /**< CreateTopics_result_t */ +#define RD_KAFKA_EVENT_DELETETOPICS_RESULT 101 /**< DeleteTopics_result_t */ +#define RD_KAFKA_EVENT_CREATEPARTITIONS_RESULT 102 /**< CreatePartitions_result_t */ +#define RD_KAFKA_EVENT_ALTERCONFIGS_RESULT 103 /**< AlterConfigs_result_t */ +#define RD_KAFKA_EVENT_DESCRIBECONFIGS_RESULT 104 /**< DescribeConfigs_result_t */ +#define RD_KAFKA_EVENT_DELETERECORDS_RESULT 105 /**< DeleteRecords_result_t */ +#define RD_KAFKA_EVENT_DELETEGROUPS_RESULT 106 /**< DeleteGroups_result_t */ +/** DeleteConsumerGroupOffsets_result_t */ +#define RD_KAFKA_EVENT_DELETECONSUMERGROUPOFFSETS_RESULT 107 +#define RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH 0x100 /**< SASL/OAUTHBEARER + token needs to be + refreshed */ + + +/** + * @returns the event type for the given event. + * + * @remark As a convenience it is okay to pass \p rkev as NULL in which case + * RD_KAFKA_EVENT_NONE is returned. + */ +RD_EXPORT +rd_kafka_event_type_t rd_kafka_event_type (const rd_kafka_event_t *rkev); + +/** + * @returns the event type's name for the given event. + * + * @remark As a convenience it is okay to pass \p rkev as NULL in which case + * the name for RD_KAFKA_EVENT_NONE is returned. + */ +RD_EXPORT +const char *rd_kafka_event_name (const rd_kafka_event_t *rkev); + + +/** + * @brief Destroy an event. + * + * @remark Any references to this event, such as extracted messages, + * will not be usable after this call. + * + * @remark As a convenience it is okay to pass \p rkev as NULL in which case + * no action is performed. + */ +RD_EXPORT +void rd_kafka_event_destroy (rd_kafka_event_t *rkev); + + +/** + * @returns the next message from an event. + * + * Call repeatedly until it returns NULL. + * + * Event types: + * - RD_KAFKA_EVENT_FETCH (1 message) + * - RD_KAFKA_EVENT_DR (>=1 message(s)) + * + * @remark The returned message(s) MUST NOT be + * freed with rd_kafka_message_destroy(). + * + * @remark on_consume() interceptor may be called + * from this function prior to passing message to application. + */ +RD_EXPORT +const rd_kafka_message_t *rd_kafka_event_message_next (rd_kafka_event_t *rkev); + + +/** + * @brief Extacts \p size message(s) from the event into the + * pre-allocated array \p rkmessages. + * + * Event types: + * - RD_KAFKA_EVENT_FETCH (1 message) + * - RD_KAFKA_EVENT_DR (>=1 message(s)) + * + * @returns the number of messages extracted. + * + * @remark on_consume() interceptor may be called + * from this function prior to passing message to application. + */ +RD_EXPORT +size_t rd_kafka_event_message_array (rd_kafka_event_t *rkev, + const rd_kafka_message_t **rkmessages, + size_t size); + + +/** + * @returns the number of remaining messages in the event. + * + * Event types: + * - RD_KAFKA_EVENT_FETCH (1 message) + * - RD_KAFKA_EVENT_DR (>=1 message(s)) + */ +RD_EXPORT +size_t rd_kafka_event_message_count (rd_kafka_event_t *rkev); + + +/** + * @returns the associated configuration string for the event, or NULL + * if the configuration property is not set or if + * not applicable for the given event type. + * + * The returned memory is read-only and its lifetime is the same as the + * event object. + * + * Event types: + * - RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH: value of sasl.oauthbearer.config + */ +RD_EXPORT +const char *rd_kafka_event_config_string (rd_kafka_event_t *rkev); + + +/** + * @returns the error code for the event. + * + * Use rd_kafka_event_error_is_fatal() to detect if this is a fatal error. + * + * Event types: + * - all + */ +RD_EXPORT +rd_kafka_resp_err_t rd_kafka_event_error (rd_kafka_event_t *rkev); + + +/** + * @returns the error string (if any). + * An application should check that rd_kafka_event_error() returns + * non-zero before calling this function. + * + * Event types: + * - all + */ +RD_EXPORT +const char *rd_kafka_event_error_string (rd_kafka_event_t *rkev); + + +/** + * @returns 1 if the error is a fatal error, else 0. + * + * Event types: + * - RD_KAFKA_EVENT_ERROR + * + * @sa rd_kafka_fatal_error() + */ +RD_EXPORT +int rd_kafka_event_error_is_fatal (rd_kafka_event_t *rkev); + + +/** + * @returns the event opaque (if any) as passed to rd_kafka_commit() (et.al) or + * rd_kafka_AdminOptions_set_opaque(), depending on event type. + * + * Event types: + * - RD_KAFKA_EVENT_OFFSET_COMMIT + * - RD_KAFKA_EVENT_CREATETOPICS_RESULT + * - RD_KAFKA_EVENT_DELETETOPICS_RESULT + * - RD_KAFKA_EVENT_CREATEPARTITIONS_RESULT + * - RD_KAFKA_EVENT_ALTERCONFIGS_RESULT + * - RD_KAFKA_EVENT_DESCRIBECONFIGS_RESULT + * - RD_KAFKA_EVENT_DELETEGROUPS_RESULT + * - RD_KAFKA_EVENT_DELETECONSUMERGROUPOFFSETS_RESULT + * - RD_KAFKA_EVENT_DELETERECORDS_RESULT + */ +RD_EXPORT +void *rd_kafka_event_opaque (rd_kafka_event_t *rkev); + + +/** + * @brief Extract log message from the event. + * + * Event types: + * - RD_KAFKA_EVENT_LOG + * + * @returns 0 on success or -1 if unsupported event type. + */ +RD_EXPORT +int rd_kafka_event_log (rd_kafka_event_t *rkev, + const char **fac, const char **str, int *level); + + +/** + * @brief Extract log debug context from event. + * + * Event types: + * - RD_KAFKA_EVENT_LOG + * + * @param rkev the event to extract data from. + * @param dst destination string for comma separated list. + * @param dstsize size of provided dst buffer. + * @returns 0 on success or -1 if unsupported event type. + */ +RD_EXPORT +int rd_kafka_event_debug_contexts (rd_kafka_event_t *rkev, + char *dst, size_t dstsize); + + +/** + * @brief Extract stats from the event. + * + * Event types: + * - RD_KAFKA_EVENT_STATS + * + * @returns stats json string. + * + * @remark the returned string will be freed automatically along with the event object + * + */ +RD_EXPORT +const char *rd_kafka_event_stats (rd_kafka_event_t *rkev); + + +/** + * @returns the topic partition list from the event. + * + * @remark The list MUST NOT be freed with rd_kafka_topic_partition_list_destroy() + * + * Event types: + * - RD_KAFKA_EVENT_REBALANCE + * - RD_KAFKA_EVENT_OFFSET_COMMIT + */ +RD_EXPORT rd_kafka_topic_partition_list_t * +rd_kafka_event_topic_partition_list (rd_kafka_event_t *rkev); + + +/** + * @returns a newly allocated topic_partition container, if applicable for the event type, + * else NULL. + * + * @remark The returned pointer MUST be freed with rd_kafka_topic_partition_destroy(). + * + * Event types: + * RD_KAFKA_EVENT_ERROR (for partition level errors) + */ +RD_EXPORT rd_kafka_topic_partition_t * +rd_kafka_event_topic_partition (rd_kafka_event_t *rkev); + + +/*! CreateTopics result type */ +typedef rd_kafka_event_t rd_kafka_CreateTopics_result_t; +/*! DeleteTopics result type */ +typedef rd_kafka_event_t rd_kafka_DeleteTopics_result_t; +/*! CreatePartitions result type */ +typedef rd_kafka_event_t rd_kafka_CreatePartitions_result_t; +/*! AlterConfigs result type */ +typedef rd_kafka_event_t rd_kafka_AlterConfigs_result_t; +/*! CreateTopics result type */ +typedef rd_kafka_event_t rd_kafka_DescribeConfigs_result_t; +/*! DeleteRecords result type */ +typedef rd_kafka_event_t rd_kafka_DeleteRecords_result_t; +/*! DeleteGroups result type */ +typedef rd_kafka_event_t rd_kafka_DeleteGroups_result_t; +/*! DeleteConsumerGroupOffsets result type */ +typedef rd_kafka_event_t rd_kafka_DeleteConsumerGroupOffsets_result_t; + +/** + * @brief Get CreateTopics result. + * + * @returns the result of a CreateTopics request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_CREATETOPICS_RESULT + */ +RD_EXPORT const rd_kafka_CreateTopics_result_t * +rd_kafka_event_CreateTopics_result (rd_kafka_event_t *rkev); + +/** + * @brief Get DeleteTopics result. + * + * @returns the result of a DeleteTopics request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_DELETETOPICS_RESULT + */ +RD_EXPORT const rd_kafka_DeleteTopics_result_t * +rd_kafka_event_DeleteTopics_result (rd_kafka_event_t *rkev); + +/** + * @brief Get CreatePartitions result. + * + * @returns the result of a CreatePartitions request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_CREATEPARTITIONS_RESULT + */ +RD_EXPORT const rd_kafka_CreatePartitions_result_t * +rd_kafka_event_CreatePartitions_result (rd_kafka_event_t *rkev); + +/** + * @brief Get AlterConfigs result. + * + * @returns the result of a AlterConfigs request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_ALTERCONFIGS_RESULT + */ +RD_EXPORT const rd_kafka_AlterConfigs_result_t * +rd_kafka_event_AlterConfigs_result (rd_kafka_event_t *rkev); + +/** + * @brief Get DescribeConfigs result. + * + * @returns the result of a DescribeConfigs request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_DESCRIBECONFIGS_RESULT + */ +RD_EXPORT const rd_kafka_DescribeConfigs_result_t * +rd_kafka_event_DescribeConfigs_result (rd_kafka_event_t *rkev); + +/** + * @returns the result of a DeleteRecords request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_DELETERECORDS_RESULT + */ +RD_EXPORT const rd_kafka_DeleteRecords_result_t * +rd_kafka_event_DeleteRecords_result (rd_kafka_event_t *rkev); + +/** + * @brief Get DeleteGroups result. + * + * @returns the result of a DeleteGroups request, or NULL if event is of + * different type. + * + * Event types: + * RD_KAFKA_EVENT_DELETEGROUPS_RESULT + */ +RD_EXPORT const rd_kafka_DeleteGroups_result_t * +rd_kafka_event_DeleteGroups_result (rd_kafka_event_t *rkev); + +/** + * @brief Get DeleteConsumerGroupOffsets result. + * + * @returns the result of a DeleteConsumerGroupOffsets request, or NULL if + * event is of different type. + * + * Event types: + * RD_KAFKA_EVENT_DELETECONSUMERGROUPOFFSETS_RESULT + */ +RD_EXPORT const rd_kafka_DeleteConsumerGroupOffsets_result_t * +rd_kafka_event_DeleteConsumerGroupOffsets_result (rd_kafka_event_t *rkev); + +/** + * @brief Poll a queue for an event for max \p timeout_ms. + * + * @returns an event, or NULL. + * + * @remark Use rd_kafka_event_destroy() to free the event. + * + * @sa rd_kafka_conf_set_background_event_cb() + */ +RD_EXPORT +rd_kafka_event_t *rd_kafka_queue_poll (rd_kafka_queue_t *rkqu, int timeout_ms); + +/** +* @brief Poll a queue for events served through callbacks for max \p timeout_ms. +* +* @returns the number of events served. +* +* @remark This API must only be used for queues with callbacks registered +* for all expected event types. E.g., not a message queue. +* +* @remark Also see rd_kafka_conf_set_background_event_cb() for triggering +* event callbacks from a librdkafka-managed background thread. +* +* @sa rd_kafka_conf_set_background_event_cb() +*/ +RD_EXPORT +int rd_kafka_queue_poll_callback (rd_kafka_queue_t *rkqu, int timeout_ms); + + +/**@}*/ + + +/** + * @name Plugin interface + * + * @brief A plugin interface that allows external runtime-loaded libraries + * to integrate with a client instance without modifications to + * the application code. + * + * Plugins are loaded when referenced through the `plugin.library.paths` + * configuration property and operates on the \c rd_kafka_conf_t + * object prior \c rd_kafka_t instance creation. + * + * @warning Plugins require the application to link librdkafka dynamically + * and not statically. Failure to do so will lead to missing symbols + * or finding symbols in another librdkafka library than the + * application was linked with. + */ + + +/** + * @brief Plugin's configuration initializer method called each time the + * library is referenced from configuration (even if previously loaded by + * another client instance). + * + * @remark This method MUST be implemented by plugins and have the symbol name + * \c conf_init + * + * @param conf Configuration set up to this point. + * @param plug_opaquep Plugin can set this pointer to a per-configuration + * opaque pointer. + * @param errstr String buffer of size \p errstr_size where plugin must write + * a human readable error string in the case the initializer + * fails (returns non-zero). + * @param errstr_size Maximum space (including \0) in \p errstr. + * + * @remark A plugin may add an on_conf_destroy() interceptor to clean up + * plugin-specific resources created in the plugin's conf_init() method. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an error code on error. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_plugin_f_conf_init_t) (rd_kafka_conf_t *conf, + void **plug_opaquep, + char *errstr, size_t errstr_size); + +/**@}*/ + + + +/** + * @name Interceptors + * + * @{ + * + * @brief A callback interface that allows message interception for both + * producer and consumer data pipelines. + * + * Except for the on_new(), on_conf_set(), on_conf_dup() and on_conf_destroy() + * interceptors, interceptors are added to the + * newly created rd_kafka_t client instance. These interceptors MUST only + * be added from on_new() and MUST NOT be added after rd_kafka_new() returns. + * + * The on_new(), on_conf_set(), on_conf_dup() and on_conf_destroy() interceptors + * are added to the configuration object which is later passed to + * rd_kafka_new() where on_new() is called to allow addition of + * other interceptors. + * + * Each interceptor reference consists of a display name (ic_name), + * a callback function, and an application-specified opaque value that is + * passed as-is to the callback. + * The ic_name must be unique for the interceptor implementation and is used + * to reject duplicate interceptor methods. + * + * Any number of interceptors can be added and they are called in the order + * they were added, unless otherwise noted. + * The list of registered interceptor methods are referred to as + * interceptor chains. + * + * @remark Contrary to the Java client the librdkafka interceptor interface + * does not support message key and value modification. + * Message mutability is discouraged in the Java client and the + * combination of serializers and headers cover most use-cases. + * + * @remark Interceptors are NOT copied to the new configuration on + * rd_kafka_conf_dup() since it would be hard for interceptors to + * track usage of the interceptor's opaque value. + * An interceptor should rely on the plugin, which will be copied + * in rd_kafka_conf_conf_dup(), to set up the initial interceptors. + * An interceptor should implement the on_conf_dup() method + * to manually set up its internal configuration on the newly created + * configuration object that is being copied-to based on the + * interceptor-specific configuration properties. + * conf_dup() should thus be treated the same as conf_init(). + * + * @remark Interceptors are keyed by the interceptor type (on_..()), the + * interceptor name (ic_name) and the interceptor method function. + * Duplicates are not allowed and the .._add_on_..() method will + * return RD_KAFKA_RESP_ERR__CONFLICT if attempting to add a duplicate + * method. + * The only exception is on_conf_destroy() which may be added multiple + * times by the same interceptor to allow proper cleanup of + * interceptor configuration state. + */ + + +/** + * @brief on_conf_set() is called from rd_kafka_*_conf_set() in the order + * the interceptors were added. + * + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * @param name The configuration property to set. + * @param val The configuration value to set, or NULL for reverting to default + * in which case the previous value should be freed. + * @param errstr A human readable error string in case the interceptor fails. + * @param errstr_size Maximum space (including \0) in \p errstr. + * + * @returns RD_KAFKA_CONF_OK if the property was known and successfully + * handled by the interceptor, RD_KAFKA_CONF_INVALID if the + * property was handled by the interceptor but the value was invalid, + * or RD_KAFKA_CONF_UNKNOWN if the interceptor did not handle + * this property, in which case the property is passed on on the + * interceptor in the chain, finally ending up at the built-in + * configuration handler. + */ +typedef rd_kafka_conf_res_t +(rd_kafka_interceptor_f_on_conf_set_t) (rd_kafka_conf_t *conf, + const char *name, const char *val, + char *errstr, size_t errstr_size, + void *ic_opaque); + + +/** + * @brief on_conf_dup() is called from rd_kafka_conf_dup() in the + * order the interceptors were added and is used to let + * an interceptor re-register its conf interecptors with a new + * opaque value. + * The on_conf_dup() method is called prior to the configuration from + * \p old_conf being copied to \p new_conf. + * + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an error code + * on failure (which is logged but otherwise ignored). + * + * @remark No on_conf_* interceptors are copied to the new configuration + * object on rd_kafka_conf_dup(). + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_conf_dup_t) (rd_kafka_conf_t *new_conf, + const rd_kafka_conf_t *old_conf, + size_t filter_cnt, + const char **filter, + void *ic_opaque); + + +/** + * @brief on_conf_destroy() is called from rd_kafka_*_conf_destroy() in the + * order the interceptors were added. + * + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_conf_destroy_t) (void *ic_opaque); + + +/** + * @brief on_new() is called from rd_kafka_new() prior toreturning + * the newly created client instance to the application. + * + * @param rk The client instance. + * @param conf The client instance's final configuration. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * @param errstr A human readable error string in case the interceptor fails. + * @param errstr_size Maximum space (including \0) in \p errstr. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + * + * @warning The \p rk client instance will not be fully set up when this + * interceptor is called and the interceptor MUST NOT call any + * other rk-specific APIs than rd_kafka_interceptor_add..(). + * + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_new_t) (rd_kafka_t *rk, const rd_kafka_conf_t *conf, + void *ic_opaque, + char *errstr, size_t errstr_size); + + +/** + * @brief on_destroy() is called from rd_kafka_destroy() or (rd_kafka_new() + * if rd_kafka_new() fails during initialization). + * + * @param rk The client instance. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_destroy_t) (rd_kafka_t *rk, void *ic_opaque); + + + + +/** + * @brief on_send() is called from rd_kafka_produce*() (et.al) prior to + * the partitioner being called. + * + * @param rk The client instance. + * @param rkmessage The message being produced. Immutable. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @remark This interceptor is only used by producer instances. + * + * @remark The \p rkmessage object is NOT mutable and MUST NOT be modified + * by the interceptor. + * + * @remark If the partitioner fails or an unknown partition was specified, + * the on_acknowledgement() interceptor chain will be called from + * within the rd_kafka_produce*() call to maintain send-acknowledgement + * symmetry. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_send_t) (rd_kafka_t *rk, + rd_kafka_message_t *rkmessage, + void *ic_opaque); + +/** + * @brief on_acknowledgement() is called to inform interceptors that a message + * was succesfully delivered or permanently failed delivery. + * The interceptor chain is called from internal librdkafka background + * threads, or rd_kafka_produce*() if the partitioner failed. + * + * @param rk The client instance. + * @param rkmessage The message being produced. Immutable. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @remark This interceptor is only used by producer instances. + * + * @remark The \p rkmessage object is NOT mutable and MUST NOT be modified + * by the interceptor. + * + * @warning The on_acknowledgement() method may be called from internal + * librdkafka threads. An on_acknowledgement() interceptor MUST NOT + * call any librdkafka API's associated with the \p rk, or perform + * any blocking or prolonged work. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_acknowledgement_t) (rd_kafka_t *rk, + rd_kafka_message_t *rkmessage, + void *ic_opaque); + + +/** + * @brief on_consume() is called just prior to passing the message to the + * application in rd_kafka_consumer_poll(), rd_kafka_consume*(), + * the event interface, etc. + * + * @param rk The client instance. + * @param rkmessage The message being consumed. Immutable. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @remark This interceptor is only used by consumer instances. + * + * @remark The \p rkmessage object is NOT mutable and MUST NOT be modified + * by the interceptor. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_consume_t) (rd_kafka_t *rk, + rd_kafka_message_t *rkmessage, + void *ic_opaque); + +/** + * @brief on_commit() is called on completed or failed offset commit. + * It is called from internal librdkafka threads. + * + * @param rk The client instance. + * @param offsets List of topic+partition+offset+error that were committed. + * The error message of each partition should be checked for + * error. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @remark This interceptor is only used by consumer instances. + * + * @warning The on_commit() interceptor is called from internal + * librdkafka threads. An on_commit() interceptor MUST NOT + * call any librdkafka API's associated with the \p rk, or perform + * any blocking or prolonged work. + * + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_commit_t) ( + rd_kafka_t *rk, + const rd_kafka_topic_partition_list_t *offsets, + rd_kafka_resp_err_t err, void *ic_opaque); + + +/** + * @brief on_request_sent() is called when a request has been fully written + * to a broker TCP connections socket. + * + * @param rk The client instance. + * @param sockfd Socket file descriptor. + * @param brokername Broker request is being sent to. + * @param brokerid Broker request is being sent to. + * @param ApiKey Kafka protocol request type. + * @param ApiVersion Kafka protocol request type version. + * @param Corrid Kafka protocol request correlation id. + * @param size Size of request. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @warning The on_request_sent() interceptor is called from internal + * librdkafka broker threads. An on_request_sent() interceptor MUST NOT + * call any librdkafka API's associated with the \p rk, or perform + * any blocking or prolonged work. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_request_sent_t) ( + rd_kafka_t *rk, + int sockfd, + const char *brokername, + int32_t brokerid, + int16_t ApiKey, + int16_t ApiVersion, + int32_t CorrId, + size_t size, + void *ic_opaque); + + +/** + * @brief on_response_received() is called when a protocol response has been + * fully received from a broker TCP connection socket but before the + * response payload is parsed. + * + * @param rk The client instance. + * @param sockfd Socket file descriptor (always -1). + * @param brokername Broker response was received from, possibly empty string + * on error. + * @param brokerid Broker response was received from. + * @param ApiKey Kafka protocol request type or -1 on error. + * @param ApiVersion Kafka protocol request type version or -1 on error. + * @param Corrid Kafka protocol request correlation id, possibly -1 on error. + * @param size Size of response, possibly 0 on error. + * @param rtt Request round-trip-time in microseconds, possibly -1 on error. + * @param err Receive error. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @warning The on_response_received() interceptor is called from internal + * librdkafka broker threads. An on_response_received() interceptor + * MUST NOT call any librdkafka API's associated with the \p rk, or + * perform any blocking or prolonged work. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_response_received_t) ( + rd_kafka_t *rk, + int sockfd, + const char *brokername, + int32_t brokerid, + int16_t ApiKey, + int16_t ApiVersion, + int32_t CorrId, + size_t size, + int64_t rtt, + rd_kafka_resp_err_t err, + void *ic_opaque); + + +/** + * @brief on_thread_start() is called from a newly created librdkafka-managed + * thread. + + * @param rk The client instance. + * @param thread_type Thread type. + * @param thread_name Human-readable thread name, may not be unique. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @warning The on_thread_start() interceptor is called from internal + * librdkafka threads. An on_thread_start() interceptor MUST NOT + * call any librdkafka API's associated with the \p rk, or perform + * any blocking or prolonged work. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_thread_start_t) ( + rd_kafka_t *rk, + rd_kafka_thread_type_t thread_type, + const char *thread_name, + void *ic_opaque); + + +/** + * @brief on_thread_exit() is called just prior to a librdkafka-managed + * thread exiting from the exiting thread itself. + * + * @param rk The client instance. + * @param thread_type Thread type.n + * @param thread_name Human-readable thread name, may not be unique. + * @param ic_opaque The interceptor's opaque pointer specified in ..add..(). + * + * @remark Depending on the thread type, librdkafka may execute additional + * code on the thread after on_thread_exit() returns. + * + * @warning The on_thread_exit() interceptor is called from internal + * librdkafka threads. An on_thread_exit() interceptor MUST NOT + * call any librdkafka API's associated with the \p rk, or perform + * any blocking or prolonged work. + * + * @returns an error code on failure, the error is logged but otherwise ignored. + */ +typedef rd_kafka_resp_err_t +(rd_kafka_interceptor_f_on_thread_exit_t) ( + rd_kafka_t *rk, + rd_kafka_thread_type_t thread_type, + const char *thread_name, + void *ic_opaque); + + + +/** + * @brief Append an on_conf_set() interceptor. + * + * @param conf Configuration object. + * @param ic_name Interceptor name, used in logging. + * @param on_conf_set Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_conf_interceptor_add_on_conf_set ( + rd_kafka_conf_t *conf, const char *ic_name, + rd_kafka_interceptor_f_on_conf_set_t *on_conf_set, + void *ic_opaque); + + +/** + * @brief Append an on_conf_dup() interceptor. + * + * @param conf Configuration object. + * @param ic_name Interceptor name, used in logging. + * @param on_conf_dup Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_conf_interceptor_add_on_conf_dup ( + rd_kafka_conf_t *conf, const char *ic_name, + rd_kafka_interceptor_f_on_conf_dup_t *on_conf_dup, + void *ic_opaque); + +/** + * @brief Append an on_conf_destroy() interceptor. + * + * @param conf Configuration object. + * @param ic_name Interceptor name, used in logging. + * @param on_conf_destroy Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR + * + * @remark Multiple on_conf_destroy() interceptors are allowed to be added + * to the same configuration object. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_conf_interceptor_add_on_conf_destroy ( + rd_kafka_conf_t *conf, const char *ic_name, + rd_kafka_interceptor_f_on_conf_destroy_t *on_conf_destroy, + void *ic_opaque); + + +/** + * @brief Append an on_new() interceptor. + * + * @param conf Configuration object. + * @param ic_name Interceptor name, used in logging. + * @param on_new Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @remark Since the on_new() interceptor is added to the configuration object + * it may be copied by rd_kafka_conf_dup(). + * An interceptor implementation must thus be able to handle + * the same interceptor,ic_opaque tuple to be used by multiple + * client instances. + * + * @remark An interceptor plugin should check the return value to make sure it + * has not already been added. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_conf_interceptor_add_on_new ( + rd_kafka_conf_t *conf, const char *ic_name, + rd_kafka_interceptor_f_on_new_t *on_new, + void *ic_opaque); + + + +/** + * @brief Append an on_destroy() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_destroy Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_destroy ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_destroy_t *on_destroy, + void *ic_opaque); + + +/** + * @brief Append an on_send() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_send Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_send ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_send_t *on_send, + void *ic_opaque); + +/** + * @brief Append an on_acknowledgement() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_acknowledgement Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_acknowledgement ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_acknowledgement_t *on_acknowledgement, + void *ic_opaque); + + +/** + * @brief Append an on_consume() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_consume Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_consume ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_consume_t *on_consume, + void *ic_opaque); + + +/** + * @brief Append an on_commit() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_commit() Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_commit ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_commit_t *on_commit, + void *ic_opaque); + + +/** + * @brief Append an on_request_sent() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_request_sent() Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_request_sent ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_request_sent_t *on_request_sent, + void *ic_opaque); + + +/** + * @brief Append an on_response_received() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_response_received() Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_response_received ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_response_received_t *on_response_received, + void *ic_opaque); + + +/** + * @brief Append an on_thread_start() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_thread_start() Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_thread_start ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_thread_start_t *on_thread_start, + void *ic_opaque); + + +/** + * @brief Append an on_thread_exit() interceptor. + * + * @param rk Client instance. + * @param ic_name Interceptor name, used in logging. + * @param on_thread_exit() Function pointer. + * @param ic_opaque Opaque value that will be passed to the function. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or RD_KAFKA_RESP_ERR__CONFLICT + * if an existing intercepted with the same \p ic_name and function + * has already been added to \p conf. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_interceptor_add_on_thread_exit ( + rd_kafka_t *rk, const char *ic_name, + rd_kafka_interceptor_f_on_thread_exit_t *on_thread_exit, + void *ic_opaque); + + + +/**@}*/ + + + +/** + * @name Auxiliary types + * + * @{ + */ + + + +/** + * @brief Topic result provides per-topic operation result information. + * + */ + +/** + * @returns the error code for the given topic result. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_topic_result_error (const rd_kafka_topic_result_t *topicres); + +/** + * @returns the human readable error string for the given topic result, + * or NULL if there was no error. + * + * @remark lifetime of the returned string is the same as the \p topicres. + */ +RD_EXPORT const char * +rd_kafka_topic_result_error_string (const rd_kafka_topic_result_t *topicres); + +/** + * @returns the name of the topic for the given topic result. + * @remark lifetime of the returned string is the same as the \p topicres. + * + */ +RD_EXPORT const char * +rd_kafka_topic_result_name (const rd_kafka_topic_result_t *topicres); + +/** + * @brief Group result provides per-group operation result information. + * + */ + +/** + * @returns the error for the given group result, or NULL on success. + * @remark lifetime of the returned error is the same as the \p groupres. + */ +RD_EXPORT const rd_kafka_error_t * +rd_kafka_group_result_error (const rd_kafka_group_result_t *groupres); + +/** + * @returns the name of the group for the given group result. + * @remark lifetime of the returned string is the same as the \p groupres. + * + */ +RD_EXPORT const char * +rd_kafka_group_result_name (const rd_kafka_group_result_t *groupres); + +/** + * @returns the partitions/offsets for the given group result, if applicable + * to the request type, else NULL. + * @remark lifetime of the returned list is the same as the \p groupres. + */ +RD_EXPORT const rd_kafka_topic_partition_list_t * +rd_kafka_group_result_partitions (const rd_kafka_group_result_t *groupres); + + +/**@}*/ + + +/** + * @name Admin API + * @{ + * + * @brief The Admin API enables applications to perform administrative + * Apache Kafka tasks, such as creating and deleting topics, + * altering and reading broker configuration, etc. + * + * The Admin API is asynchronous and makes use of librdkafka's standard + * \c rd_kafka_queue_t queues to propagate the result of an admin operation + * back to the application. + * The supplied queue may be any queue, such as a temporary single-call queue, + * a shared queue used for multiple requests, or even the main queue or + * consumer queues. + * + * Use \c rd_kafka_queue_poll() to collect the result of an admin operation + * from the queue of your choice, then extract the admin API-specific result + * type by using the corresponding \c rd_kafka_event_CreateTopics_result, + * \c rd_kafka_event_DescribeConfigs_result, etc, methods. + * Use the getter methods on the \c .._result_t type to extract response + * information and finally destroy the result and event by calling + * \c rd_kafka_event_destroy(). + * + * Use rd_kafka_event_error() and rd_kafka_event_error_string() to acquire + * the request-level error/success for an Admin API request. + * Even if the returned value is \c RD_KAFKA_RESP_ERR_NO_ERROR there + * may be individual objects (topics, resources, etc) that have failed. + * Extract per-object error information with the corresponding + * \c rd_kafka_..._result_topics|resources|..() to check per-object errors. + * + * Locally triggered errors: + * - \c RD_KAFKA_RESP_ERR__TIMED_OUT - (Controller) broker connection did not + * become available in the time allowed by AdminOption_set_request_timeout. + */ + + +/** + * @enum rd_kafka_admin_op_t + * + * @brief Admin operation enum name for use with rd_kafka_AdminOptions_new() + * + * @sa rd_kafka_AdminOptions_new() + */ +typedef enum rd_kafka_admin_op_t { + RD_KAFKA_ADMIN_OP_ANY = 0, /**< Default value */ + RD_KAFKA_ADMIN_OP_CREATETOPICS, /**< CreateTopics */ + RD_KAFKA_ADMIN_OP_DELETETOPICS, /**< DeleteTopics */ + RD_KAFKA_ADMIN_OP_CREATEPARTITIONS, /**< CreatePartitions */ + RD_KAFKA_ADMIN_OP_ALTERCONFIGS, /**< AlterConfigs */ + RD_KAFKA_ADMIN_OP_DESCRIBECONFIGS, /**< DescribeConfigs */ + RD_KAFKA_ADMIN_OP_DELETERECORDS, /**< DeleteRecords */ + RD_KAFKA_ADMIN_OP_DELETEGROUPS, /**< DeleteGroups */ + /** DeleteConsumerGroupOffsets */ + RD_KAFKA_ADMIN_OP_DELETECONSUMERGROUPOFFSETS, + RD_KAFKA_ADMIN_OP__CNT /**< Number of ops defined */ +} rd_kafka_admin_op_t; + +/** + * @brief AdminOptions provides a generic mechanism for setting optional + * parameters for the Admin API requests. + * + * @remark Since AdminOptions is decoupled from the actual request type + * there is no enforcement to prevent setting unrelated properties, + * e.g. setting validate_only on a DescribeConfigs request is allowed + * but is silently ignored by DescribeConfigs. + * Future versions may introduce such enforcement. + */ + + +typedef struct rd_kafka_AdminOptions_s rd_kafka_AdminOptions_t; + +/** + * @brief Create a new AdminOptions object. + * + * The options object is not modified by the Admin API request APIs, + * (e.g. CreateTopics) and may be reused for multiple calls. + * + * @param rk Client instance. + * @param for_api Specifies what Admin API this AdminOptions object will be used + * for, which will enforce what AdminOptions_set_..() calls may + * be used based on the API, causing unsupported set..() calls + * to fail. + * Specifying RD_KAFKA_ADMIN_OP_ANY disables the enforcement + * allowing any option to be set, even if the option + * is not used in a future call to an Admin API method. + * + * @returns a new AdminOptions object (which must be freed with + * rd_kafka_AdminOptions_destroy()), or NULL if \p for_api was set to + * an unknown API op type. + */ +RD_EXPORT rd_kafka_AdminOptions_t * +rd_kafka_AdminOptions_new (rd_kafka_t *rk, rd_kafka_admin_op_t for_api); + + +/** + * @brief Destroy a AdminOptions object. + */ +RD_EXPORT void rd_kafka_AdminOptions_destroy (rd_kafka_AdminOptions_t *options); + + +/** + * @brief Sets the overall request timeout, including broker lookup, + * request transmission, operation time on broker, and response. + * + * @param options Admin options. + * @param timeout_ms Timeout in milliseconds, use -1 for indefinite timeout. + * Defaults to `socket.timeout.ms`. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, or + * RD_KAFKA_RESP_ERR__INVALID_ARG if timeout was out of range in which + * case an error string will be written \p errstr. + * + * @remark This option is valid for all Admin API requests. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_AdminOptions_set_request_timeout (rd_kafka_AdminOptions_t *options, + int timeout_ms, + char *errstr, size_t errstr_size); + + +/** + * @brief Sets the broker's operation timeout, such as the timeout for + * CreateTopics to complete the creation of topics on the controller + * before returning a result to the application. + * + * CreateTopics: values <= 0 will return immediately after triggering topic + * creation, while > 0 will wait this long for topic creation to propagate + * in cluster. Default: 60 seconds. + * + * DeleteTopics: same semantics as CreateTopics. + * CreatePartitions: same semantics as CreateTopics. + * + * @param options Admin options. + * @param timeout_ms Timeout in milliseconds. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, or + * RD_KAFKA_RESP_ERR__INVALID_ARG if timeout was out of range in which + * case an error string will be written \p errstr. + * + * @remark This option is valid for CreateTopics, DeleteTopics, + * CreatePartitions, and DeleteRecords. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_AdminOptions_set_operation_timeout (rd_kafka_AdminOptions_t *options, + int timeout_ms, + char *errstr, size_t errstr_size); + + +/** + * @brief Tell broker to only validate the request, without performing + * the requested operation (create topics, etc). + * + * @param options Admin options. + * @param true_or_false Defaults to false. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an + * error code on failure in which case an error string will + * be written \p errstr. + * + * @remark This option is valid for CreateTopics, + * CreatePartitions, AlterConfigs. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_AdminOptions_set_validate_only (rd_kafka_AdminOptions_t *options, + int true_or_false, + char *errstr, size_t errstr_size); + + +/** + * @brief Override what broker the Admin request will be sent to. + * + * By default, Admin requests are sent to the controller broker, with + * the following exceptions: + * - AlterConfigs with a BROKER resource are sent to the broker id set + * as the resource name. + * - DescribeConfigs with a BROKER resource are sent to the broker id set + * as the resource name. + * + * @param options Admin Options. + * @param broker_id The broker to send the request to. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success or an + * error code on failure in which case an error string will + * be written \p errstr. + * + * @remark This API should typically not be used, but serves as a workaround + * if new resource types are to the broker that the client + * does not know where to send. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_AdminOptions_set_broker (rd_kafka_AdminOptions_t *options, + int32_t broker_id, + char *errstr, size_t errstr_size); + + + +/** + * @brief Set application opaque value that can be extracted from the + * result event using rd_kafka_event_opaque() + */ +RD_EXPORT void +rd_kafka_AdminOptions_set_opaque (rd_kafka_AdminOptions_t *options, + void *ev_opaque); + + + + + + +/* + * CreateTopics - create topics in cluster. + * + */ + + +/*! Defines a new topic to be created. */ +typedef struct rd_kafka_NewTopic_s rd_kafka_NewTopic_t; + +/** + * @brief Create a new NewTopic object. This object is later passed to + * rd_kafka_CreateTopics(). + * + * @param topic Topic name to create. + * @param num_partitions Number of partitions in topic, or -1 to use the + * broker's default partition count (>= 2.4.0). + * @param replication_factor Default replication factor for the topic's + * partitions, or -1 to use the broker's default + * replication factor (>= 2.4.0) or if + * set_replica_assignment() will be used. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * + * @returns a new allocated NewTopic object, or NULL if the input parameters + * are invalid. + * Use rd_kafka_NewTopic_destroy() to free object when done. + */ +RD_EXPORT rd_kafka_NewTopic_t * +rd_kafka_NewTopic_new (const char *topic, int num_partitions, + int replication_factor, + char *errstr, size_t errstr_size); + +/** + * @brief Destroy and free a NewTopic object previously created with + * rd_kafka_NewTopic_new() + */ +RD_EXPORT void +rd_kafka_NewTopic_destroy (rd_kafka_NewTopic_t *new_topic); + + +/** + * @brief Helper function to destroy all NewTopic objects in the \p new_topics + * array (of \p new_topic_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_NewTopic_destroy_array (rd_kafka_NewTopic_t **new_topics, + size_t new_topic_cnt); + + +/** + * @brief Set the replica (broker) assignment for \p partition to the + * replica set in \p broker_ids (of \p broker_id_cnt elements). + * + * @remark When this method is used, rd_kafka_NewTopic_new() must have + * been called with a \c replication_factor of -1. + * + * @remark An application must either set the replica assignment for + * all new partitions, or none. + * + * @remark If called, this function must be called consecutively for each + * partition, starting at 0. + * + * @remark Use rd_kafka_metadata() to retrieve the list of brokers + * in the cluster. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, or an error code + * if the arguments were invalid. + * + * @sa rd_kafka_AdminOptions_set_validate_only() + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_NewTopic_set_replica_assignment (rd_kafka_NewTopic_t *new_topic, + int32_t partition, + int32_t *broker_ids, + size_t broker_id_cnt, + char *errstr, size_t errstr_size); + +/** + * @brief Set (broker-side) topic configuration name/value pair. + * + * @remark The name and value are not validated by the client, the validation + * takes place on the broker. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, or an error code + * if the arguments were invalid. + * + * @sa rd_kafka_AdminOptions_set_validate_only() + * @sa http://kafka.apache.org/documentation.html#topicconfigs + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_NewTopic_set_config (rd_kafka_NewTopic_t *new_topic, + const char *name, const char *value); + + +/** + * @brief Create topics in cluster as specified by the \p new_topics + * array of size \p new_topic_cnt elements. + * + * @param rk Client instance. + * @param new_topics Array of new topics to create. + * @param new_topic_cnt Number of elements in \p new_topics array. + * @param options Optional admin options, or NULL for defaults. + * @param rkqu Queue to emit result on. + * + * Supported admin options: + * - rd_kafka_AdminOptions_set_validate_only() - default false + * - rd_kafka_AdminOptions_set_operation_timeout() - default 60 seconds + * - rd_kafka_AdminOptions_set_request_timeout() - default socket.timeout.ms + * + * @remark The result event type emitted on the supplied queue is of type + * \c RD_KAFKA_EVENT_CREATETOPICS_RESULT + */ +RD_EXPORT void +rd_kafka_CreateTopics (rd_kafka_t *rk, + rd_kafka_NewTopic_t **new_topics, + size_t new_topic_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + +/* + * CreateTopics result type and methods + */ + +/** + * @brief Get an array of topic results from a CreateTopics result. + * + * The returned \p topics life-time is the same as the \p result object. + * + * @param result Result to get topics from. + * @param cntp Updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_topic_result_t ** +rd_kafka_CreateTopics_result_topics ( + const rd_kafka_CreateTopics_result_t *result, + size_t *cntp); + + + + + +/* + * DeleteTopics - delete topics from cluster + * + */ + +/*! Represents a topic to be deleted. */ +typedef struct rd_kafka_DeleteTopic_s rd_kafka_DeleteTopic_t; + +/** + * @brief Create a new DeleteTopic object. This object is later passed to + * rd_kafka_DeleteTopics(). + * + * @param topic Topic name to delete. + * + * @returns a new allocated DeleteTopic object. + * Use rd_kafka_DeleteTopic_destroy() to free object when done. + */ +RD_EXPORT rd_kafka_DeleteTopic_t * +rd_kafka_DeleteTopic_new (const char *topic); + +/** + * @brief Destroy and free a DeleteTopic object previously created with + * rd_kafka_DeleteTopic_new() + */ +RD_EXPORT void +rd_kafka_DeleteTopic_destroy (rd_kafka_DeleteTopic_t *del_topic); + +/** + * @brief Helper function to destroy all DeleteTopic objects in + * the \p del_topics array (of \p del_topic_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_DeleteTopic_destroy_array (rd_kafka_DeleteTopic_t **del_topics, + size_t del_topic_cnt); + +/** + * @brief Delete topics from cluster as specified by the \p topics + * array of size \p topic_cnt elements. + * + * @param rk Client instance. + * @param del_topics Array of topics to delete. + * @param del_topic_cnt Number of elements in \p topics array. + * @param options Optional admin options, or NULL for defaults. + * @param rkqu Queue to emit result on. + * + * @remark The result event type emitted on the supplied queue is of type + * \c RD_KAFKA_EVENT_DELETETOPICS_RESULT + */ +RD_EXPORT +void rd_kafka_DeleteTopics (rd_kafka_t *rk, + rd_kafka_DeleteTopic_t **del_topics, + size_t del_topic_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + + +/* + * DeleteTopics result type and methods + */ + +/** + * @brief Get an array of topic results from a DeleteTopics result. + * + * The returned \p topics life-time is the same as the \p result object. + * + * @param result Result to get topic results from. + * @param cntp is updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_topic_result_t ** +rd_kafka_DeleteTopics_result_topics ( + const rd_kafka_DeleteTopics_result_t *result, + size_t *cntp); + + + + + + +/* + * CreatePartitions - add partitions to topic. + * + */ + +/*! Defines a new partition to be created. */ +typedef struct rd_kafka_NewPartitions_s rd_kafka_NewPartitions_t; + +/** + * @brief Create a new NewPartitions. This object is later passed to + * rd_kafka_CreatePartitions() to increase the number of partitions + * to \p new_total_cnt for an existing topic. + * + * @param topic Topic name to create more partitions for. + * @param new_total_cnt Increase the topic's partition count to this value. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * @returns a new allocated NewPartitions object, or NULL if the + * input parameters are invalid. + * Use rd_kafka_NewPartitions_destroy() to free object when done. + */ +RD_EXPORT rd_kafka_NewPartitions_t * +rd_kafka_NewPartitions_new (const char *topic, size_t new_total_cnt, + char *errstr, size_t errstr_size); + +/** + * @brief Destroy and free a NewPartitions object previously created with + * rd_kafka_NewPartitions_new() + */ +RD_EXPORT void +rd_kafka_NewPartitions_destroy (rd_kafka_NewPartitions_t *new_parts); + +/** + * @brief Helper function to destroy all NewPartitions objects in the + * \p new_parts array (of \p new_parts_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_NewPartitions_destroy_array (rd_kafka_NewPartitions_t **new_parts, + size_t new_parts_cnt); + +/** + * @brief Set the replica (broker id) assignment for \p new_partition_idx to the + * replica set in \p broker_ids (of \p broker_id_cnt elements). + * + * @remark An application must either set the replica assignment for + * all new partitions, or none. + * + * @remark If called, this function must be called consecutively for each + * new partition being created, + * where \p new_partition_idx 0 is the first new partition, + * 1 is the second, and so on. + * + * @remark \p broker_id_cnt should match the topic's replication factor. + * + * @remark Use rd_kafka_metadata() to retrieve the list of brokers + * in the cluster. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR on success, or an error code + * if the arguments were invalid. + * + * @sa rd_kafka_AdminOptions_set_validate_only() + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_NewPartitions_set_replica_assignment (rd_kafka_NewPartitions_t *new_parts, + int32_t new_partition_idx, + int32_t *broker_ids, + size_t broker_id_cnt, + char *errstr, + size_t errstr_size); + + +/** + * @brief Create additional partitions for the given topics, as specified + * by the \p new_parts array of size \p new_parts_cnt elements. + * + * @param rk Client instance. + * @param new_parts Array of topics for which new partitions are to be created. + * @param new_parts_cnt Number of elements in \p new_parts array. + * @param options Optional admin options, or NULL for defaults. + * @param rkqu Queue to emit result on. + * + * Supported admin options: + * - rd_kafka_AdminOptions_set_validate_only() - default false + * - rd_kafka_AdminOptions_set_operation_timeout() - default 60 seconds + * - rd_kafka_AdminOptions_set_request_timeout() - default socket.timeout.ms + * + * @remark The result event type emitted on the supplied queue is of type + * \c RD_KAFKA_EVENT_CREATEPARTITIONS_RESULT + */ +RD_EXPORT void +rd_kafka_CreatePartitions (rd_kafka_t *rk, + rd_kafka_NewPartitions_t **new_parts, + size_t new_parts_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + + +/* + * CreatePartitions result type and methods + */ + +/** + * @brief Get an array of topic results from a CreatePartitions result. + * + * The returned \p topics life-time is the same as the \p result object. + * + * @param result Result o get topic results from. + * @param cntp is updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_topic_result_t ** +rd_kafka_CreatePartitions_result_topics ( + const rd_kafka_CreatePartitions_result_t *result, + size_t *cntp); + + + + + +/* + * Cluster, broker, topic configuration entries, sources, etc. + * + */ + +/** + * @enum rd_kafka_ConfigSource_t + * + * @brief Apache Kafka config sources. + * + * @remark These entities relate to the cluster, not the local client. + * + * @sa rd_kafka_conf_set(), et.al. for local client configuration. + */ +typedef enum rd_kafka_ConfigSource_t { + /** Source unknown, e.g., in the ConfigEntry used for alter requests + * where source is not set */ + RD_KAFKA_CONFIG_SOURCE_UNKNOWN_CONFIG = 0, + /** Dynamic topic config that is configured for a specific topic */ + RD_KAFKA_CONFIG_SOURCE_DYNAMIC_TOPIC_CONFIG = 1, + /** Dynamic broker config that is configured for a specific broker */ + RD_KAFKA_CONFIG_SOURCE_DYNAMIC_BROKER_CONFIG = 2, + /** Dynamic broker config that is configured as default for all + * brokers in the cluster */ + RD_KAFKA_CONFIG_SOURCE_DYNAMIC_DEFAULT_BROKER_CONFIG = 3, + /** Static broker config provided as broker properties at startup + * (e.g. from server.properties file) */ + RD_KAFKA_CONFIG_SOURCE_STATIC_BROKER_CONFIG = 4, + /** Built-in default configuration for configs that have a + * default value */ + RD_KAFKA_CONFIG_SOURCE_DEFAULT_CONFIG = 5, + + /** Number of source types defined */ + RD_KAFKA_CONFIG_SOURCE__CNT, +} rd_kafka_ConfigSource_t; + + +/** + * @returns a string representation of the \p confsource. + */ +RD_EXPORT const char * +rd_kafka_ConfigSource_name (rd_kafka_ConfigSource_t confsource); + + +/*! Apache Kafka configuration entry. */ +typedef struct rd_kafka_ConfigEntry_s rd_kafka_ConfigEntry_t; + +/** + * @returns the configuration property name + */ +RD_EXPORT const char * +rd_kafka_ConfigEntry_name (const rd_kafka_ConfigEntry_t *entry); + +/** + * @returns the configuration value, may be NULL for sensitive or unset + * properties. + */ +RD_EXPORT const char * +rd_kafka_ConfigEntry_value (const rd_kafka_ConfigEntry_t *entry); + +/** + * @returns the config source. + */ +RD_EXPORT rd_kafka_ConfigSource_t +rd_kafka_ConfigEntry_source (const rd_kafka_ConfigEntry_t *entry); + +/** + * @returns 1 if the config property is read-only on the broker, else 0. + * @remark Shall only be used on a DescribeConfigs result, otherwise returns -1. + */ +RD_EXPORT int +rd_kafka_ConfigEntry_is_read_only (const rd_kafka_ConfigEntry_t *entry); + +/** + * @returns 1 if the config property is set to its default value on the broker, + * else 0. + * @remark Shall only be used on a DescribeConfigs result, otherwise returns -1. + */ +RD_EXPORT int +rd_kafka_ConfigEntry_is_default (const rd_kafka_ConfigEntry_t *entry); + +/** + * @returns 1 if the config property contains sensitive information (such as + * security configuration), else 0. + * @remark An application should take care not to include the value of + * sensitive configuration entries in its output. + * @remark Shall only be used on a DescribeConfigs result, otherwise returns -1. + */ +RD_EXPORT int +rd_kafka_ConfigEntry_is_sensitive (const rd_kafka_ConfigEntry_t *entry); + +/** + * @returns 1 if this entry is a synonym, else 0. + */ +RD_EXPORT int +rd_kafka_ConfigEntry_is_synonym (const rd_kafka_ConfigEntry_t *entry); + + +/** + * @returns the synonym config entry array. + * + * @param entry Entry to get synonyms for. + * @param cntp is updated to the number of elements in the array. + * + * @remark The lifetime of the returned entry is the same as \p conf . + * @remark Shall only be used on a DescribeConfigs result, + * otherwise returns NULL. + */ +RD_EXPORT const rd_kafka_ConfigEntry_t ** +rd_kafka_ConfigEntry_synonyms (const rd_kafka_ConfigEntry_t *entry, + size_t *cntp); + + + + +/*! Apache Kafka resource types */ +typedef enum rd_kafka_ResourceType_t { + RD_KAFKA_RESOURCE_UNKNOWN = 0, /**< Unknown */ + RD_KAFKA_RESOURCE_ANY = 1, /**< Any (used for lookups) */ + RD_KAFKA_RESOURCE_TOPIC = 2, /**< Topic */ + RD_KAFKA_RESOURCE_GROUP = 3, /**< Group */ + RD_KAFKA_RESOURCE_BROKER = 4, /**< Broker */ + RD_KAFKA_RESOURCE__CNT, /**< Number of resource types defined */ +} rd_kafka_ResourceType_t; + +/** + * @returns a string representation of the \p restype + */ +RD_EXPORT const char * +rd_kafka_ResourceType_name (rd_kafka_ResourceType_t restype); + +/*! Apache Kafka configuration resource. */ +typedef struct rd_kafka_ConfigResource_s rd_kafka_ConfigResource_t; + + +/** + * @brief Create new ConfigResource object. + * + * @param restype The resource type (e.g., RD_KAFKA_RESOURCE_TOPIC) + * @param resname The resource name (e.g., the topic name) + * + * @returns a newly allocated object + */ +RD_EXPORT rd_kafka_ConfigResource_t * +rd_kafka_ConfigResource_new (rd_kafka_ResourceType_t restype, + const char *resname); + +/** + * @brief Destroy and free a ConfigResource object previously created with + * rd_kafka_ConfigResource_new() + */ +RD_EXPORT void +rd_kafka_ConfigResource_destroy (rd_kafka_ConfigResource_t *config); + + +/** + * @brief Helper function to destroy all ConfigResource objects in + * the \p configs array (of \p config_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_ConfigResource_destroy_array (rd_kafka_ConfigResource_t **config, + size_t config_cnt); + + +/** + * @brief Set configuration name value pair. + * + * @param config ConfigResource to set config property on. + * @param name Configuration name, depends on resource type. + * @param value Configuration value, depends on resource type and \p name. + * Set to \c NULL to revert configuration value to default. + * + * This will overwrite the current value. + * + * @returns RD_KAFKA_RESP_ERR_NO_ERROR if config was added to resource, + * or RD_KAFKA_RESP_ERR__INVALID_ARG on invalid input. + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_ConfigResource_set_config (rd_kafka_ConfigResource_t *config, + const char *name, const char *value); + + +/** + * @brief Get an array of config entries from a ConfigResource object. + * + * The returned object life-times are the same as the \p config object. + * + * @param config ConfigResource to get configs from. + * @param cntp is updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_ConfigEntry_t ** +rd_kafka_ConfigResource_configs (const rd_kafka_ConfigResource_t *config, + size_t *cntp); + + + +/** + * @returns the ResourceType for \p config + */ +RD_EXPORT rd_kafka_ResourceType_t +rd_kafka_ConfigResource_type (const rd_kafka_ConfigResource_t *config); + +/** + * @returns the name for \p config + */ +RD_EXPORT const char * +rd_kafka_ConfigResource_name (const rd_kafka_ConfigResource_t *config); + +/** + * @returns the error for this resource from an AlterConfigs request + */ +RD_EXPORT rd_kafka_resp_err_t +rd_kafka_ConfigResource_error (const rd_kafka_ConfigResource_t *config); + +/** + * @returns the error string for this resource from an AlterConfigs + * request, or NULL if no error. + */ +RD_EXPORT const char * +rd_kafka_ConfigResource_error_string (const rd_kafka_ConfigResource_t *config); + + +/* + * AlterConfigs - alter cluster configuration. + * + */ + + +/** + * @brief Update the configuration for the specified resources. + * Updates are not transactional so they may succeed for a subset + * of the provided resources while the others fail. + * The configuration for a particular resource is updated atomically, + * replacing values using the provided ConfigEntrys and reverting + * unspecified ConfigEntrys to their default values. + * + * @remark Requires broker version >=0.11.0.0 + * + * @warning AlterConfigs will replace all existing configuration for + * the provided resources with the new configuration given, + * reverting all other configuration to their default values. + * + * @remark Multiple resources and resource types may be set, but at most one + * resource of type \c RD_KAFKA_RESOURCE_BROKER is allowed per call + * since these resource requests must be sent to the broker specified + * in the resource. + * + */ +RD_EXPORT +void rd_kafka_AlterConfigs (rd_kafka_t *rk, + rd_kafka_ConfigResource_t **configs, + size_t config_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + +/* + * AlterConfigs result type and methods + */ + +/** + * @brief Get an array of resource results from a AlterConfigs result. + * + * Use \c rd_kafka_ConfigResource_error() and + * \c rd_kafka_ConfigResource_error_string() to extract per-resource error + * results on the returned array elements. + * + * The returned object life-times are the same as the \p result object. + * + * @param result Result object to get resource results from. + * @param cntp is updated to the number of elements in the array. + * + * @returns an array of ConfigResource elements, or NULL if not available. + */ +RD_EXPORT const rd_kafka_ConfigResource_t ** +rd_kafka_AlterConfigs_result_resources ( + const rd_kafka_AlterConfigs_result_t *result, + size_t *cntp); + + + + + + +/* + * DescribeConfigs - retrieve cluster configuration. + * + */ + + +/** + * @brief Get configuration for the specified resources in \p configs. + * + * The returned configuration includes default values and the + * rd_kafka_ConfigEntry_is_default() or rd_kafka_ConfigEntry_source() + * methods may be used to distinguish them from user supplied values. + * + * The value of config entries where rd_kafka_ConfigEntry_is_sensitive() + * is true will always be NULL to avoid disclosing sensitive + * information, such as security settings. + * + * Configuration entries where rd_kafka_ConfigEntry_is_read_only() + * is true can't be updated (with rd_kafka_AlterConfigs()). + * + * Synonym configuration entries are returned if the broker supports + * it (broker version >= 1.1.0). See rd_kafka_ConfigEntry_synonyms(). + * + * @remark Requires broker version >=0.11.0.0 + * + * @remark Multiple resources and resource types may be requested, but at most + * one resource of type \c RD_KAFKA_RESOURCE_BROKER is allowed per call + * since these resource requests must be sent to the broker specified + * in the resource. + */ +RD_EXPORT +void rd_kafka_DescribeConfigs (rd_kafka_t *rk, + rd_kafka_ConfigResource_t **configs, + size_t config_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + + + +/* + * DescribeConfigs result type and methods + */ + +/** + * @brief Get an array of resource results from a DescribeConfigs result. + * + * The returned \p resources life-time is the same as the \p result object. + * + * @param result Result object to get resource results from. + * @param cntp is updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_ConfigResource_t ** +rd_kafka_DescribeConfigs_result_resources ( + const rd_kafka_DescribeConfigs_result_t *result, + size_t *cntp); + + +/* + * DeleteRecords - delete records (messages) from partitions + * + * + */ + +/**! Represents records to be deleted */ +typedef struct rd_kafka_DeleteRecords_s rd_kafka_DeleteRecords_t; + +/** + * @brief Create a new DeleteRecords object. This object is later passed to + * rd_kafka_DeleteRecords(). + * + * \p before_offsets must contain \c topic, \c partition, and + * \c offset is the offset before which the messages will + * be deleted (exclusive). + * Set \c offset to RD_KAFKA_OFFSET_END (high-watermark) in order to + * delete all data in the partition. + * + * @param before_offsets For each partition delete all messages up to but not + * including the specified offset. + * + * @returns a new allocated DeleteRecords object. + * Use rd_kafka_DeleteRecords_destroy() to free object when done. + */ +RD_EXPORT rd_kafka_DeleteRecords_t * +rd_kafka_DeleteRecords_new (const rd_kafka_topic_partition_list_t * + before_offsets); + +/** + * @brief Destroy and free a DeleteRecords object previously created with + * rd_kafka_DeleteRecords_new() + */ +RD_EXPORT void +rd_kafka_DeleteRecords_destroy (rd_kafka_DeleteRecords_t *del_records); + +/** + * @brief Helper function to destroy all DeleteRecords objects in + * the \p del_groups array (of \p del_group_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_DeleteRecords_destroy_array (rd_kafka_DeleteRecords_t **del_records, + size_t del_record_cnt); + +/** + * @brief Delete records (messages) in topic partitions older than the + * offsets provided. + * + * @param rk Client instance. + * @param del_records The offsets to delete (up to). + * Currently only one DeleteRecords_t (but containing + * multiple offsets) is supported. + * @param del_record_cnt The number of elements in del_records, must be 1. + * @param options Optional admin options, or NULL for defaults. + * @param rkqu Queue to emit result on. + * + * Supported admin options: + * - rd_kafka_AdminOptions_set_operation_timeout() - default 60 seconds. + * Controls how long the brokers will wait for records to be deleted. + * - rd_kafka_AdminOptions_set_request_timeout() - default socket.timeout.ms. + * Controls how long \c rdkafka will wait for the request to complete. + * + * @remark The result event type emitted on the supplied queue is of type + * \c RD_KAFKA_EVENT_DELETERECORDS_RESULT + */ +RD_EXPORT void +rd_kafka_DeleteRecords (rd_kafka_t *rk, + rd_kafka_DeleteRecords_t **del_records, + size_t del_record_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + +/* + * DeleteRecords result type and methods + */ + +/** + * @brief Get a list of topic and partition results from a DeleteRecords result. + * The returned objects will contain \c topic, \c partition, \c offset + * and \c err. \c offset will be set to the post-deletion low-watermark + * (smallest available offset of all live replicas). \c err will be set + * per-partition if deletion failed. + * + * The returned object's life-time is the same as the \p result object. + */ +RD_EXPORT const rd_kafka_topic_partition_list_t * +rd_kafka_DeleteRecords_result_offsets ( + const rd_kafka_DeleteRecords_result_t *result); + +/* + * DeleteGroups - delete groups from cluster + * + * + */ + +/*! Represents a group to be deleted. */ +typedef struct rd_kafka_DeleteGroup_s rd_kafka_DeleteGroup_t; + +/** + * @brief Create a new DeleteGroup object. This object is later passed to + * rd_kafka_DeleteGroups(). + * + * @param group Name of group to delete. + * + * @returns a new allocated DeleteGroup object. + * Use rd_kafka_DeleteGroup_destroy() to free object when done. + */ +RD_EXPORT rd_kafka_DeleteGroup_t * +rd_kafka_DeleteGroup_new (const char *group); + +/** + * @brief Destroy and free a DeleteGroup object previously created with + * rd_kafka_DeleteGroup_new() + */ +RD_EXPORT void +rd_kafka_DeleteGroup_destroy (rd_kafka_DeleteGroup_t *del_group); + +/** + * @brief Helper function to destroy all DeleteGroup objects in + * the \p del_groups array (of \p del_group_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_DeleteGroup_destroy_array (rd_kafka_DeleteGroup_t **del_groups, + size_t del_group_cnt); + +/** + * @brief Delete groups from cluster as specified by the \p del_groups + * array of size \p del_group_cnt elements. + * + * @param rk Client instance. + * @param del_groups Array of groups to delete. + * @param del_group_cnt Number of elements in \p del_groups array. + * @param options Optional admin options, or NULL for defaults. + * @param rkqu Queue to emit result on. + * + * @remark The result event type emitted on the supplied queue is of type + * \c RD_KAFKA_EVENT_DELETEGROUPS_RESULT + */ +RD_EXPORT +void rd_kafka_DeleteGroups (rd_kafka_t *rk, + rd_kafka_DeleteGroup_t **del_groups, + size_t del_group_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + + +/* + * DeleteGroups result type and methods + */ + +/** + * @brief Get an array of group results from a DeleteGroups result. + * + * The returned groups life-time is the same as the \p result object. + * + * @param result Result to get group results from. + * @param cntp is updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_group_result_t ** +rd_kafka_DeleteGroups_result_groups ( + const rd_kafka_DeleteGroups_result_t *result, + size_t *cntp); + + +/* + * DeleteConsumerGroupOffsets - delete groups from cluster + * + * + */ + +/*! Represents consumer group committed offsets to be deleted. */ +typedef struct rd_kafka_DeleteConsumerGroupOffsets_s +rd_kafka_DeleteConsumerGroupOffsets_t; + +/** + * @brief Create a new DeleteConsumerGroupOffsets object. + * This object is later passed to rd_kafka_DeleteConsumerGroupOffsets(). + * + * @param group Consumer group id. + * @param partitions Partitions to delete committed offsets for. + * Only the topic and partition fields are used. + * + * @returns a new allocated DeleteConsumerGroupOffsets object. + * Use rd_kafka_DeleteConsumerGroupOffsets_destroy() to free + * object when done. + */ +RD_EXPORT rd_kafka_DeleteConsumerGroupOffsets_t * +rd_kafka_DeleteConsumerGroupOffsets_new (const char *group, + const rd_kafka_topic_partition_list_t + *partitions); + +/** + * @brief Destroy and free a DeleteConsumerGroupOffsets object previously + * created with rd_kafka_DeleteConsumerGroupOffsets_new() + */ +RD_EXPORT void +rd_kafka_DeleteConsumerGroupOffsets_destroy ( + rd_kafka_DeleteConsumerGroupOffsets_t *del_grpoffsets); + +/** + * @brief Helper function to destroy all DeleteConsumerGroupOffsets objects in + * the \p del_grpoffsets array (of \p del_grpoffsets_cnt elements). + * The array itself is not freed. + */ +RD_EXPORT void +rd_kafka_DeleteConsumerGroupOffsets_destroy_array ( + rd_kafka_DeleteConsumerGroupOffsets_t **del_grpoffsets, + size_t del_grpoffset_cnt); + +/** + * @brief Delete committed offsets for a set of partitions in a conusmer + * group. This will succeed at the partition level only if the group + * is not actively subscribed to the corresponding topic. + * + * @param rk Client instance. + * @param del_grpoffsets Array of group committed offsets to delete. + * MUST only be one single element. + * @param del_grpoffsets_cnt Number of elements in \p del_grpoffsets array. + * MUST always be 1. + * @param options Optional admin options, or NULL for defaults. + * @param rkqu Queue to emit result on. + * + * @remark The result event type emitted on the supplied queue is of type + * \c RD_KAFKA_EVENT_DELETECONSUMERGROUPOFFSETS_RESULT + * + * @remark The current implementation only supports one group per invocation. + */ +RD_EXPORT +void rd_kafka_DeleteConsumerGroupOffsets ( + rd_kafka_t *rk, + rd_kafka_DeleteConsumerGroupOffsets_t **del_grpoffsets, + size_t del_grpoffsets_cnt, + const rd_kafka_AdminOptions_t *options, + rd_kafka_queue_t *rkqu); + + + +/* + * DeleteConsumerGroupOffsets result type and methods + */ + +/** + * @brief Get an array of results from a DeleteConsumerGroupOffsets result. + * + * The returned groups life-time is the same as the \p result object. + * + * @param result Result to get group results from. + * @param cntp is updated to the number of elements in the array. + */ +RD_EXPORT const rd_kafka_group_result_t ** +rd_kafka_DeleteConsumerGroupOffsets_result_groups ( + const rd_kafka_DeleteConsumerGroupOffsets_result_t *result, + size_t *cntp); + + +/**@}*/ + + +/** + * @name Security APIs + * @{ + * + */ + +/** + * @brief Set SASL/OAUTHBEARER token and metadata + * + * @param rk Client instance. + * @param token_value the mandatory token value to set, often (but not + * necessarily) a JWS compact serialization as per + * https://tools.ietf.org/html/rfc7515#section-3.1. + * @param md_lifetime_ms when the token expires, in terms of the number of + * milliseconds since the epoch. + * @param md_principal_name the mandatory Kafka principal name associated + * with the token. + * @param extensions optional SASL extensions key-value array with + * \p extensions_size elements (number of keys * 2), where [i] is the key and + * [i+1] is the key's value, to be communicated to the broker + * as additional key-value pairs during the initial client response as per + * https://tools.ietf.org/html/rfc7628#section-3.1. The key-value pairs are + * copied. + * @param extension_size the number of SASL extension keys plus values, + * which must be a non-negative multiple of 2. + * @param errstr A human readable error string (nul-terminated) is written to + * this location that must be of at least \p errstr_size bytes. + * The \p errstr is only written in case of error. + * @param errstr_size Writable size in \p errstr. + * + * The SASL/OAUTHBEARER token refresh callback or event handler should invoke + * this method upon success. The extension keys must not include the reserved + * key "`auth`", and all extension keys and values must conform to the required + * format as per https://tools.ietf.org/html/rfc7628#section-3.1: + * + * key = 1*(ALPHA) + * value = *(VCHAR / SP / HTAB / CR / LF ) + * + * @returns \c RD_KAFKA_RESP_ERR_NO_ERROR on success, otherwise \p errstr set + * and:
+ * \c RD_KAFKA_RESP_ERR__INVALID_ARG if any of the arguments are + * invalid;
+ * \c RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED if SASL/OAUTHBEARER is not + * supported by this build;
+ * \c RD_KAFKA_RESP_ERR__STATE if SASL/OAUTHBEARER is supported but is + * not configured as the client's authentication mechanism.
+ * + * @sa rd_kafka_oauthbearer_set_token_failure + * @sa rd_kafka_conf_set_oauthbearer_token_refresh_cb + */ +RD_EXPORT +rd_kafka_resp_err_t +rd_kafka_oauthbearer_set_token (rd_kafka_t *rk, + const char *token_value, + int64_t md_lifetime_ms, + const char *md_principal_name, + const char **extensions, size_t extension_size, + char *errstr, size_t errstr_size); + +/** + * @brief SASL/OAUTHBEARER token refresh failure indicator. + * + * @param rk Client instance. + * @param errstr mandatory human readable error reason for failing to acquire + * a token. + * + * The SASL/OAUTHBEARER token refresh callback or event handler should invoke + * this method upon failure. + * + * @returns \c RD_KAFKA_RESP_ERR_NO_ERROR on success, otherwise:
+ * \c RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED if SASL/OAUTHBEARER is not + * supported by this build;
+ * \c RD_KAFKA_RESP_ERR__STATE if SASL/OAUTHBEARER is supported but is + * not configured as the client's authentication mechanism,
+ * \c RD_KAFKA_RESP_ERR__INVALID_ARG if no error string is supplied. + * + * @sa rd_kafka_oauthbearer_set_token + * @sa rd_kafka_conf_set_oauthbearer_token_refresh_cb + */ +RD_EXPORT +rd_kafka_resp_err_t +rd_kafka_oauthbearer_set_token_failure (rd_kafka_t *rk, const char *errstr); + +/**@}*/ + + +/** + * @name Transactional producer API + * + * The transactional producer operates on top of the idempotent producer, + * and provides full exactly-once semantics (EOS) for Apache Kafka when used + * with the transaction aware consumer (\c isolation.level=read_committed). + * + * A producer instance is configured for transactions by setting the + * \c transactional.id to an identifier unique for the application. This + * id will be used to fence stale transactions from previous instances of + * the application, typically following an outage or crash. + * + * After creating the transactional producer instance using rd_kafka_new() + * the transactional state must be initialized by calling + * rd_kafka_init_transactions(). This is a blocking call that will + * acquire a runtime producer id from the transaction coordinator broker + * as well as abort any stale transactions and fence any still running producer + * instances with the same \c transactional.id. + * + * Once transactions are initialized the application may begin a new + * transaction by calling rd_kafka_begin_transaction(). + * A producer instance may only have one single on-going transaction. + * + * Any messages produced after the transaction has been started will + * belong to the ongoing transaction and will be committed or aborted + * atomically. + * It is not permitted to produce messages outside a transaction + * boundary, e.g., before rd_kafka_begin_transaction() or after + * rd_kafka_commit_transaction(), rd_kafka_abort_transaction(), or after + * the current transaction has failed. + * + * If consumed messages are used as input to the transaction, the consumer + * instance must be configured with \c enable.auto.commit set to \c false. + * To commit the consumed offsets along with the transaction pass the + * list of consumed partitions and the last offset processed + 1 to + * rd_kafka_send_offsets_to_transaction() prior to committing the transaction. + * This allows an aborted transaction to be restarted using the previously + * committed offsets. + * + * To commit the produced messages, and any consumed offsets, to the + * current transaction, call rd_kafka_commit_transaction(). + * This call will block until the transaction has been fully committed or + * failed (typically due to fencing by a newer producer instance). + * + * Alternatively, if processing fails, or an abortable transaction error is + * raised, the transaction needs to be aborted by calling + * rd_kafka_abort_transaction() which marks any produced messages and + * offset commits as aborted. + * + * After the current transaction has been committed or aborted a new + * transaction may be started by calling rd_kafka_begin_transaction() again. + * + * @par Retriable errors + * Some error cases allow the attempted operation to be retried, this is + * indicated by the error object having the retriable flag set which can + * be detected by calling rd_kafka_error_is_retriable(). + * When this flag is set the application may retry the operation immediately + * or preferably after a shorter grace period (to avoid busy-looping). + * Retriable errors include timeouts, broker transport failures, etc. + * + * @par Abortable errors + * An ongoing transaction may fail permanently due to various errors, + * such as transaction coordinator becoming unavailable, write failures to the + * Apache Kafka log, under-replicated partitions, etc. + * At this point the producer application must abort the current transaction + * using rd_kafka_abort_transaction() and optionally start a new transaction + * by calling rd_kafka_begin_transaction(). + * Whether an error is abortable or not is detected by calling + * rd_kafka_error_txn_requires_abort() on the returned error object. + * + * @par Fatal errors + * While the underlying idempotent producer will typically only raise + * fatal errors for unrecoverable cluster errors where the idempotency + * guarantees can't be maintained, most of these are treated as abortable by + * the transactional producer since transactions may be aborted and retried + * in their entirety; + * The transactional producer on the other hand introduces a set of additional + * fatal errors which the application needs to handle by shutting down the + * producer and terminate. There is no way for a producer instance to recover + * from fatal errors. + * Whether an error is fatal or not is detected by calling + * rd_kafka_error_is_fatal() on the returned error object or by checking + * the global rd_kafka_fatal_error() code. + * Fatal errors are raised by triggering the \c error_cb (see the + * Fatal error chapter in INTRODUCTION.md for more information), and any + * sub-sequent transactional API calls will return RD_KAFKA_RESP_ERR__FATAL + * or have the fatal flag set (see rd_kafka_error_is_fatal()). + * The originating fatal error code can be retrieved by calling + * rd_kafka_fatal_error(). + * + * @par Handling of other errors + * For errors that have neither retriable, abortable or the fatal flag set + * it is not always obvious how to handle them. While some of these errors + * may be indicative of bugs in the application code, such as when + * an invalid parameter is passed to a method, other errors might originate + * from the broker and be passed thru as-is to the application. + * The general recommendation is to treat these errors, that have + * neither the retriable or abortable flags set, as fatal. + * + * @par Error handling example + * @code + * retry: + * rd_kafka_error_t *error; + * + * error = rd_kafka_commit_transaction(producer, 10*1000); + * if (!error) + * return success; + * else if (rd_kafka_error_txn_requires_abort(error)) { + * do_abort_transaction_and_reset_inputs(); + * } else if (rd_kafka_error_is_retriable(error)) { + * rd_kafka_error_destroy(error); + * goto retry; + * } else { // treat all other errors as fatal errors + * fatal_error(rd_kafka_error_string(error)); + * } + * rd_kafka_error_destroy(error); + * @endcode + * + * + * @{ + */ + + +/** + * @brief Initialize transactions for the producer instance. + * + * This function ensures any transactions initiated by previous instances + * of the producer with the same \c transactional.id are completed. + * If the previous instance failed with a transaction in progress the + * previous transaction will be aborted. + * This function needs to be called before any other transactional or + * produce functions are called when the \c transactional.id is configured. + * + * If the last transaction had begun completion (following transaction commit) + * but not yet finished, this function will await the previous transaction's + * completion. + * + * When any previous transactions have been fenced this function + * will acquire the internal producer id and epoch, used in all future + * transactional messages issued by this producer instance. + * + * @param rk Producer instance. + * @param timeout_ms The maximum time to block. On timeout the operation + * may continue in the background, depending on state, + * and it is okay to call init_transactions() again. + * + * @remark This function may block up to \p timeout_ms milliseconds. + * + * @returns NULL on success or an error object on failure. + * Check whether the returned error object permits retrying + * by calling rd_kafka_error_is_retriable(), or whether a fatal + * error has been raised by calling rd_kafka_error_is_fatal(). + * Error codes: + * RD_KAFKA_RESP_ERR__TIMED_OUT if the transaction coordinator + * could be not be contacted within \p timeout_ms (retriable), + * RD_KAFKA_RESP_ERR_COORDINATOR_NOT_AVAILABLE if the transaction + * coordinator is not available (retriable), + * RD_KAFKA_RESP_ERR_CONCURRENT_TRANSACTIONS if a previous transaction + * would not complete within \p timeout_ms (retriable), + * RD_KAFKA_RESP_ERR__STATE if transactions have already been started + * or upon fatal error, + * RD_KAFKA_RESP_ERR__UNSUPPORTED_FEATURE if the broker(s) do not + * support transactions ( +#include +#include "select_rdkafka.h" +#include "glue_rdkafka.h" + +void setup_rkmessage (rd_kafka_message_t *rkmessage, + rd_kafka_topic_t *rkt, int32_t partition, + const void *payload, size_t len, + void *key, size_t keyLen, void *opaque) { + rkmessage->rkt = rkt; + rkmessage->partition = partition; + rkmessage->payload = (void *)payload; + rkmessage->len = len; + rkmessage->key = (void *)key; + rkmessage->key_len = keyLen; + rkmessage->_private = opaque; +} +*/ +import "C" + +// TimestampType is a the Message timestamp type or source +// +type TimestampType int + +const ( + // TimestampNotAvailable indicates no timestamp was set, or not available due to lacking broker support + TimestampNotAvailable = TimestampType(C.RD_KAFKA_TIMESTAMP_NOT_AVAILABLE) + // TimestampCreateTime indicates timestamp set by producer (source time) + TimestampCreateTime = TimestampType(C.RD_KAFKA_TIMESTAMP_CREATE_TIME) + // TimestampLogAppendTime indicates timestamp set set by broker (store time) + TimestampLogAppendTime = TimestampType(C.RD_KAFKA_TIMESTAMP_LOG_APPEND_TIME) +) + +func (t TimestampType) String() string { + switch t { + case TimestampCreateTime: + return "CreateTime" + case TimestampLogAppendTime: + return "LogAppendTime" + case TimestampNotAvailable: + fallthrough + default: + return "NotAvailable" + } +} + +// Message represents a Kafka message +type Message struct { + TopicPartition TopicPartition + Value []byte + Key []byte + Timestamp time.Time + TimestampType TimestampType + Opaque interface{} + Headers []Header +} + +// String returns a human readable representation of a Message. +// Key and payload are not represented. +func (m *Message) String() string { + var topic string + if m.TopicPartition.Topic != nil { + topic = *m.TopicPartition.Topic + } else { + topic = "" + } + return fmt.Sprintf("%s[%d]@%s", topic, m.TopicPartition.Partition, m.TopicPartition.Offset) +} + +func (h *handle) getRktFromMessage(msg *Message) (crkt *C.rd_kafka_topic_t) { + if msg.TopicPartition.Topic == nil { + return nil + } + + return h.getRkt(*msg.TopicPartition.Topic) +} + +// setupHeadersFromGlueMsg converts the C tmp headers in gMsg to +// Go Headers in msg. +// gMsg.tmphdrs will be freed. +func setupHeadersFromGlueMsg(msg *Message, gMsg *C.glue_msg_t) { + msg.Headers = make([]Header, gMsg.tmphdrsCnt) + for n := range msg.Headers { + tmphdr := (*[1 << 30]C.tmphdr_t)(unsafe.Pointer(gMsg.tmphdrs))[n] + msg.Headers[n].Key = C.GoString(tmphdr.key) + if tmphdr.val != nil { + msg.Headers[n].Value = C.GoBytes(unsafe.Pointer(tmphdr.val), C.int(tmphdr.size)) + } else { + msg.Headers[n].Value = nil + } + } + C.free(unsafe.Pointer(gMsg.tmphdrs)) +} + +func (h *handle) newMessageFromGlueMsg(gMsg *C.glue_msg_t) (msg *Message) { + msg = &Message{} + + if gMsg.ts != -1 { + ts := int64(gMsg.ts) + msg.TimestampType = TimestampType(gMsg.tstype) + msg.Timestamp = time.Unix(ts/1000, (ts%1000)*1000000) + } + + if gMsg.tmphdrsCnt > 0 { + setupHeadersFromGlueMsg(msg, gMsg) + } + + h.setupMessageFromC(msg, gMsg.msg) + + return msg +} + +// setupMessageFromC sets up a message object from a C rd_kafka_message_t +func (h *handle) setupMessageFromC(msg *Message, cmsg *C.rd_kafka_message_t) { + if cmsg.rkt != nil { + topic := h.getTopicNameFromRkt(cmsg.rkt) + msg.TopicPartition.Topic = &topic + } + msg.TopicPartition.Partition = int32(cmsg.partition) + if cmsg.payload != nil && h.msgFields.Value { + msg.Value = C.GoBytes(unsafe.Pointer(cmsg.payload), C.int(cmsg.len)) + } + if cmsg.key != nil && h.msgFields.Key { + msg.Key = C.GoBytes(unsafe.Pointer(cmsg.key), C.int(cmsg.key_len)) + } + if h.msgFields.Headers { + var gMsg C.glue_msg_t + gMsg.msg = cmsg + gMsg.want_hdrs = C.int8_t(1) + chdrsToTmphdrs(&gMsg) + if gMsg.tmphdrsCnt > 0 { + setupHeadersFromGlueMsg(msg, &gMsg) + } + } + msg.TopicPartition.Offset = Offset(cmsg.offset) + if cmsg.err != 0 { + msg.TopicPartition.Error = newError(cmsg.err) + } +} + +// newMessageFromC creates a new message object from a C rd_kafka_message_t +// NOTE: For use with Producer: does not set message timestamp fields. +func (h *handle) newMessageFromC(cmsg *C.rd_kafka_message_t) (msg *Message) { + msg = &Message{} + + h.setupMessageFromC(msg, cmsg) + + return msg +} + +// messageToC sets up cmsg as a clone of msg +func (h *handle) messageToC(msg *Message, cmsg *C.rd_kafka_message_t) { + var valp unsafe.Pointer + var keyp unsafe.Pointer + + // to circumvent Cgo constraints we need to allocate C heap memory + // for both Value and Key (one allocation back to back) + // and copy the bytes from Value and Key to the C memory. + // We later tell librdkafka (in produce()) to free the + // C memory pointer when it is done. + var payload unsafe.Pointer + + valueLen := 0 + keyLen := 0 + if msg.Value != nil { + valueLen = len(msg.Value) + } + if msg.Key != nil { + keyLen = len(msg.Key) + } + + allocLen := valueLen + keyLen + if allocLen > 0 { + payload = C.malloc(C.size_t(allocLen)) + if valueLen > 0 { + copy((*[1 << 30]byte)(payload)[0:valueLen], msg.Value) + valp = payload + } + if keyLen > 0 { + copy((*[1 << 30]byte)(payload)[valueLen:allocLen], msg.Key) + keyp = unsafe.Pointer(&((*[1 << 31]byte)(payload)[valueLen])) + } + } + + cmsg.rkt = h.getRktFromMessage(msg) + cmsg.partition = C.int32_t(msg.TopicPartition.Partition) + cmsg.payload = valp + cmsg.len = C.size_t(valueLen) + cmsg.key = keyp + cmsg.key_len = C.size_t(keyLen) + cmsg._private = nil +} + +// used for testing messageToC performance +func (h *handle) messageToCDummy(msg *Message) { + var cmsg C.rd_kafka_message_t + h.messageToC(msg, &cmsg) +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/metadata.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/metadata.go new file mode 100644 index 0000000..3b195f4 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/metadata.go @@ -0,0 +1,180 @@ +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import ( + "unsafe" +) + +/* +#include +#include "select_rdkafka.h" + +struct rd_kafka_metadata_broker *_getMetadata_broker_element(struct rd_kafka_metadata *m, int i) { + return &m->brokers[i]; +} + +struct rd_kafka_metadata_topic *_getMetadata_topic_element(struct rd_kafka_metadata *m, int i) { + return &m->topics[i]; +} + +struct rd_kafka_metadata_partition *_getMetadata_partition_element(struct rd_kafka_metadata *m, int topic_idx, int partition_idx) { + return &m->topics[topic_idx].partitions[partition_idx]; +} + +int32_t _get_int32_element (int32_t *arr, int i) { + return arr[i]; +} + +*/ +import "C" + +// BrokerMetadata contains per-broker metadata +type BrokerMetadata struct { + ID int32 + Host string + Port int +} + +// PartitionMetadata contains per-partition metadata +type PartitionMetadata struct { + ID int32 + Error Error + Leader int32 + Replicas []int32 + Isrs []int32 +} + +// TopicMetadata contains per-topic metadata +type TopicMetadata struct { + Topic string + Partitions []PartitionMetadata + Error Error +} + +// Metadata contains broker and topic metadata for all (matching) topics +type Metadata struct { + Brokers []BrokerMetadata + Topics map[string]TopicMetadata + + OriginatingBroker BrokerMetadata +} + +// getMetadata queries broker for cluster and topic metadata. +// If topic is non-nil only information about that topic is returned, else if +// allTopics is false only information about locally used topics is returned, +// else information about all topics is returned. +func getMetadata(H Handle, topic *string, allTopics bool, timeoutMs int) (*Metadata, error) { + h := H.gethandle() + + var rkt *C.rd_kafka_topic_t + if topic != nil { + rkt = h.getRkt(*topic) + } + + var cMd *C.struct_rd_kafka_metadata + cErr := C.rd_kafka_metadata(h.rk, bool2cint(allTopics), + rkt, &cMd, C.int(timeoutMs)) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cErr) + } + + m := Metadata{} + defer C.rd_kafka_metadata_destroy(cMd) + + m.Brokers = make([]BrokerMetadata, cMd.broker_cnt) + for i := 0; i < int(cMd.broker_cnt); i++ { + b := C._getMetadata_broker_element(cMd, C.int(i)) + m.Brokers[i] = BrokerMetadata{int32(b.id), C.GoString(b.host), + int(b.port)} + } + + m.Topics = make(map[string]TopicMetadata, int(cMd.topic_cnt)) + for i := 0; i < int(cMd.topic_cnt); i++ { + t := C._getMetadata_topic_element(cMd, C.int(i)) + + thisTopic := C.GoString(t.topic) + m.Topics[thisTopic] = TopicMetadata{Topic: thisTopic, + Error: newError(t.err), + Partitions: make([]PartitionMetadata, int(t.partition_cnt))} + + for j := 0; j < int(t.partition_cnt); j++ { + p := C._getMetadata_partition_element(cMd, C.int(i), C.int(j)) + m.Topics[thisTopic].Partitions[j] = PartitionMetadata{ + ID: int32(p.id), + Error: newError(p.err), + Leader: int32(p.leader)} + m.Topics[thisTopic].Partitions[j].Replicas = make([]int32, int(p.replica_cnt)) + for ir := 0; ir < int(p.replica_cnt); ir++ { + m.Topics[thisTopic].Partitions[j].Replicas[ir] = int32(C._get_int32_element(p.replicas, C.int(ir))) + } + + m.Topics[thisTopic].Partitions[j].Isrs = make([]int32, int(p.isr_cnt)) + for ii := 0; ii < int(p.isr_cnt); ii++ { + m.Topics[thisTopic].Partitions[j].Isrs[ii] = int32(C._get_int32_element(p.isrs, C.int(ii))) + } + } + } + + m.OriginatingBroker = BrokerMetadata{int32(cMd.orig_broker_id), + C.GoString(cMd.orig_broker_name), 0} + + return &m, nil +} + +// queryWatermarkOffsets returns the broker's low and high offsets for the given topic +// and partition. +func queryWatermarkOffsets(H Handle, topic string, partition int32, timeoutMs int) (low, high int64, err error) { + h := H.gethandle() + + ctopic := C.CString(topic) + defer C.free(unsafe.Pointer(ctopic)) + + var cLow, cHigh C.int64_t + + e := C.rd_kafka_query_watermark_offsets(h.rk, ctopic, C.int32_t(partition), + &cLow, &cHigh, C.int(timeoutMs)) + if e != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return 0, 0, newError(e) + } + + low = int64(cLow) + high = int64(cHigh) + return low, high, nil +} + +// getWatermarkOffsets returns the clients cached low and high offsets for the given topic +// and partition. +func getWatermarkOffsets(H Handle, topic string, partition int32) (low, high int64, err error) { + h := H.gethandle() + + ctopic := C.CString(topic) + defer C.free(unsafe.Pointer(ctopic)) + + var cLow, cHigh C.int64_t + + e := C.rd_kafka_get_watermark_offsets(h.rk, ctopic, C.int32_t(partition), + &cLow, &cHigh) + if e != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return 0, 0, newError(e) + } + + low = int64(cLow) + high = int64(cHigh) + + return low, high, nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/misc.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/misc.go new file mode 100644 index 0000000..6d602ce --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/misc.go @@ -0,0 +1,35 @@ +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import "C" + +// bool2int converts a bool to a C.int (1 or 0) +func bool2cint(b bool) C.int { + if b { + return 1 + } + return 0 +} + +// cint2bool converts a C.int to a bool +func cint2bool(v C.int) bool { + if v == 0 { + return false + } + return true +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/offset.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/offset.go new file mode 100644 index 0000000..4cb1819 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/offset.go @@ -0,0 +1,145 @@ +/** + * Copyright 2017 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import ( + "fmt" + "strconv" +) + +/* +#include +#include "select_rdkafka.h" + +static int64_t _c_rdkafka_offset_tail(int64_t rel) { + return RD_KAFKA_OFFSET_TAIL(rel); +} +*/ +import "C" + +// Offset type (int64) with support for canonical names +type Offset int64 + +// OffsetBeginning represents the earliest offset (logical) +const OffsetBeginning = Offset(C.RD_KAFKA_OFFSET_BEGINNING) + +// OffsetEnd represents the latest offset (logical) +const OffsetEnd = Offset(C.RD_KAFKA_OFFSET_END) + +// OffsetInvalid represents an invalid/unspecified offset +const OffsetInvalid = Offset(C.RD_KAFKA_OFFSET_INVALID) + +// OffsetStored represents a stored offset +const OffsetStored = Offset(C.RD_KAFKA_OFFSET_STORED) + +func (o Offset) String() string { + switch o { + case OffsetBeginning: + return "beginning" + case OffsetEnd: + return "end" + case OffsetInvalid: + return "unset" + case OffsetStored: + return "stored" + default: + return fmt.Sprintf("%d", int64(o)) + } +} + +// Set offset value, see NewOffset() +func (o *Offset) Set(offset interface{}) error { + n, err := NewOffset(offset) + + if err == nil { + *o = n + } + + return err +} + +// NewOffset creates a new Offset using the provided logical string, or an +// absolute int64 offset value. +// Logical offsets: "beginning", "earliest", "end", "latest", "unset", "invalid", "stored" +func NewOffset(offset interface{}) (Offset, error) { + + switch v := offset.(type) { + case string: + switch v { + case "beginning": + fallthrough + case "earliest": + return Offset(OffsetBeginning), nil + + case "end": + fallthrough + case "latest": + return Offset(OffsetEnd), nil + + case "unset": + fallthrough + case "invalid": + return Offset(OffsetInvalid), nil + + case "stored": + return Offset(OffsetStored), nil + + default: + off, err := strconv.Atoi(v) + return Offset(off), err + } + + case int: + return Offset((int64)(v)), nil + case int64: + return Offset(v), nil + default: + return OffsetInvalid, newErrorFromString(ErrInvalidArg, + fmt.Sprintf("Invalid offset type: %t", v)) + } +} + +// OffsetTail returns the logical offset relativeOffset from current end of partition +func OffsetTail(relativeOffset Offset) Offset { + return Offset(C._c_rdkafka_offset_tail(C.int64_t(relativeOffset))) +} + +// offsetsForTimes looks up offsets by timestamp for the given partitions. +// +// The returned offset for each partition is the earliest offset whose +// timestamp is greater than or equal to the given timestamp in the +// corresponding partition. If the provided timestamp exceeds that of the +// last message in the partition, a value of -1 will be returned. +// +// The timestamps to query are represented as `.Offset` in the `times` +// argument and the looked up offsets are represented as `.Offset` in the returned +// `offsets` list. +// +// The function will block for at most timeoutMs milliseconds. +// +// Duplicate Topic+Partitions are not supported. +// Per-partition errors may be returned in the `.Error` field. +func offsetsForTimes(H Handle, times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) { + cparts := newCPartsFromTopicPartitions(times) + defer C.rd_kafka_topic_partition_list_destroy(cparts) + cerr := C.rd_kafka_offsets_for_times(H.gethandle().rk, cparts, C.int(timeoutMs)) + if cerr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return nil, newError(cerr) + } + + return newTopicPartitionsFromCparts(cparts), nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/producer.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/producer.go new file mode 100644 index 0000000..f68854a --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/producer.go @@ -0,0 +1,918 @@ +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import ( + "context" + "fmt" + "math" + "time" + "unsafe" +) + +/* +#include +#include "select_rdkafka.h" +#include "glue_rdkafka.h" + + +#ifdef RD_KAFKA_V_HEADERS +// Convert tmphdrs to chdrs (created by this function). +// If tmphdr.size == -1: value is considered Null +// tmphdr.size == 0: value is considered empty (ignored) +// tmphdr.size > 0: value is considered non-empty +// +// WARNING: The header keys and values will be freed by this function. +void tmphdrs_to_chdrs (tmphdr_t *tmphdrs, size_t tmphdrsCnt, + rd_kafka_headers_t **chdrs) { + size_t i; + + *chdrs = rd_kafka_headers_new(tmphdrsCnt); + + for (i = 0 ; i < tmphdrsCnt ; i++) { + rd_kafka_header_add(*chdrs, + tmphdrs[i].key, -1, + tmphdrs[i].size == -1 ? NULL : + (tmphdrs[i].size == 0 ? "" : tmphdrs[i].val), + tmphdrs[i].size == -1 ? 0 : tmphdrs[i].size); + if (tmphdrs[i].size > 0) + free((void *)tmphdrs[i].val); + free((void *)tmphdrs[i].key); + } +} + +#else +void free_tmphdrs (tmphdr_t *tmphdrs, size_t tmphdrsCnt) { + size_t i; + for (i = 0 ; i < tmphdrsCnt ; i++) { + if (tmphdrs[i].size > 0) + free((void *)tmphdrs[i].val); + free((void *)tmphdrs[i].key); + } +} +#endif + + +rd_kafka_resp_err_t do_produce (rd_kafka_t *rk, + rd_kafka_topic_t *rkt, int32_t partition, + int msgflags, + int valIsNull, void *val, size_t val_len, + int keyIsNull, void *key, size_t key_len, + int64_t timestamp, + tmphdr_t *tmphdrs, size_t tmphdrsCnt, + uintptr_t cgoid) { + void *valp = valIsNull ? NULL : val; + void *keyp = keyIsNull ? NULL : key; +#ifdef RD_KAFKA_V_TIMESTAMP +rd_kafka_resp_err_t err; +#ifdef RD_KAFKA_V_HEADERS + rd_kafka_headers_t *hdrs = NULL; +#endif +#endif + + + if (tmphdrsCnt > 0) { +#ifdef RD_KAFKA_V_HEADERS + tmphdrs_to_chdrs(tmphdrs, tmphdrsCnt, &hdrs); +#else + free_tmphdrs(tmphdrs, tmphdrsCnt); + return RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED; +#endif + } + + +#ifdef RD_KAFKA_V_TIMESTAMP + err = rd_kafka_producev(rk, + RD_KAFKA_V_RKT(rkt), + RD_KAFKA_V_PARTITION(partition), + RD_KAFKA_V_MSGFLAGS(msgflags), + RD_KAFKA_V_VALUE(valp, val_len), + RD_KAFKA_V_KEY(keyp, key_len), + RD_KAFKA_V_TIMESTAMP(timestamp), +#ifdef RD_KAFKA_V_HEADERS + RD_KAFKA_V_HEADERS(hdrs), +#endif + RD_KAFKA_V_OPAQUE((void *)cgoid), + RD_KAFKA_V_END); +#ifdef RD_KAFKA_V_HEADERS + if (err && hdrs) + rd_kafka_headers_destroy(hdrs); +#endif + return err; +#else + if (timestamp) + return RD_KAFKA_RESP_ERR__NOT_IMPLEMENTED; + if (rd_kafka_produce(rkt, partition, msgflags, + valp, val_len, + keyp, key_len, + (void *)cgoid) == -1) + return rd_kafka_last_error(); + else + return RD_KAFKA_RESP_ERR_NO_ERROR; +#endif +} +*/ +import "C" + +// Producer implements a High-level Apache Kafka Producer instance +type Producer struct { + events chan Event + produceChannel chan *Message + handle handle + + // Terminates the poller() goroutine + pollerTermChan chan bool +} + +// String returns a human readable name for a Producer instance +func (p *Producer) String() string { + return p.handle.String() +} + +// get_handle implements the Handle interface +func (p *Producer) gethandle() *handle { + return &p.handle +} + +func (p *Producer) produce(msg *Message, msgFlags int, deliveryChan chan Event) error { + if msg == nil || msg.TopicPartition.Topic == nil || len(*msg.TopicPartition.Topic) == 0 { + return newErrorFromString(ErrInvalidArg, "") + } + + crkt := p.handle.getRkt(*msg.TopicPartition.Topic) + + // Three problems: + // 1) There's a difference between an empty Value or Key (length 0, proper pointer) and + // a null Value or Key (length 0, null pointer). + // 2) we need to be able to send a null Value or Key, but the unsafe.Pointer(&slice[0]) + // dereference can't be performed on a nil slice. + // 3) cgo's pointer checking requires the unsafe.Pointer(slice..) call to be made + // in the call to the C function. + // + // Solution: + // Keep track of whether the Value or Key were nil (1), but let the valp and keyp pointers + // point to a 1-byte slice (but the length to send is still 0) so that the dereference (2) + // works. + // Then perform the unsafe.Pointer() on the valp and keyp pointers (which now either point + // to the original msg.Value and msg.Key or to the 1-byte slices) in the call to C (3). + // + var valp []byte + var keyp []byte + oneByte := []byte{0} + var valIsNull C.int + var keyIsNull C.int + var valLen int + var keyLen int + + if msg.Value == nil { + valIsNull = 1 + valLen = 0 + valp = oneByte + } else { + valLen = len(msg.Value) + if valLen > 0 { + valp = msg.Value + } else { + valp = oneByte + } + } + + if msg.Key == nil { + keyIsNull = 1 + keyLen = 0 + keyp = oneByte + } else { + keyLen = len(msg.Key) + if keyLen > 0 { + keyp = msg.Key + } else { + keyp = oneByte + } + } + + var cgoid int + + // Per-message state that needs to be retained through the C code: + // delivery channel (if specified) + // message opaque (if specified) + // Since these cant be passed as opaque pointers to the C code, + // due to cgo constraints, we add them to a per-producer map for lookup + // when the C code triggers the callbacks or events. + if deliveryChan != nil || msg.Opaque != nil { + cgoid = p.handle.cgoPut(cgoDr{deliveryChan: deliveryChan, opaque: msg.Opaque}) + } + + var timestamp int64 + if !msg.Timestamp.IsZero() { + timestamp = msg.Timestamp.UnixNano() / 1000000 + } + + // Convert headers to C-friendly tmphdrs + var tmphdrs []C.tmphdr_t + tmphdrsCnt := len(msg.Headers) + + if tmphdrsCnt > 0 { + tmphdrs = make([]C.tmphdr_t, tmphdrsCnt) + + for n, hdr := range msg.Headers { + // Make a copy of the key + // to avoid runtime panic with + // foreign Go pointers in cgo. + tmphdrs[n].key = C.CString(hdr.Key) + if hdr.Value != nil { + tmphdrs[n].size = C.ssize_t(len(hdr.Value)) + if tmphdrs[n].size > 0 { + // Make a copy of the value + // to avoid runtime panic with + // foreign Go pointers in cgo. + tmphdrs[n].val = C.CBytes(hdr.Value) + } + } else { + // null value + tmphdrs[n].size = C.ssize_t(-1) + } + } + } else { + // no headers, need a dummy tmphdrs of size 1 to avoid index + // out of bounds panic in do_produce() call below. + // tmphdrsCnt will be 0. + tmphdrs = []C.tmphdr_t{{nil, nil, 0}} + } + + cErr := C.do_produce(p.handle.rk, crkt, + C.int32_t(msg.TopicPartition.Partition), + C.int(msgFlags)|C.RD_KAFKA_MSG_F_COPY, + valIsNull, unsafe.Pointer(&valp[0]), C.size_t(valLen), + keyIsNull, unsafe.Pointer(&keyp[0]), C.size_t(keyLen), + C.int64_t(timestamp), + (*C.tmphdr_t)(unsafe.Pointer(&tmphdrs[0])), C.size_t(tmphdrsCnt), + (C.uintptr_t)(cgoid)) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + if cgoid != 0 { + p.handle.cgoGet(cgoid) + } + return newError(cErr) + } + + return nil +} + +// Produce single message. +// This is an asynchronous call that enqueues the message on the internal +// transmit queue, thus returning immediately. +// The delivery report will be sent on the provided deliveryChan if specified, +// or on the Producer object's Events() channel if not. +// msg.Timestamp requires librdkafka >= 0.9.4 (else returns ErrNotImplemented), +// api.version.request=true, and broker >= 0.10.0.0. +// msg.Headers requires librdkafka >= 0.11.4 (else returns ErrNotImplemented), +// api.version.request=true, and broker >= 0.11.0.0. +// Returns an error if message could not be enqueued. +func (p *Producer) Produce(msg *Message, deliveryChan chan Event) error { + return p.produce(msg, 0, deliveryChan) +} + +// Produce a batch of messages. +// These batches do not relate to the message batches sent to the broker, the latter +// are collected on the fly internally in librdkafka. +// WARNING: This is an experimental API. +// NOTE: timestamps and headers are not supported with this API. +func (p *Producer) produceBatch(topic string, msgs []*Message, msgFlags int) error { + crkt := p.handle.getRkt(topic) + + cmsgs := make([]C.rd_kafka_message_t, len(msgs)) + for i, m := range msgs { + p.handle.messageToC(m, &cmsgs[i]) + } + r := C.rd_kafka_produce_batch(crkt, C.RD_KAFKA_PARTITION_UA, C.int(msgFlags)|C.RD_KAFKA_MSG_F_FREE, + (*C.rd_kafka_message_t)(&cmsgs[0]), C.int(len(msgs))) + if r == -1 { + return newError(C.rd_kafka_last_error()) + } + + return nil +} + +// Events returns the Events channel (read) +func (p *Producer) Events() chan Event { + return p.events +} + +// Logs returns the Log channel (if enabled), else nil +func (p *Producer) Logs() chan LogEvent { + return p.handle.logs +} + +// ProduceChannel returns the produce *Message channel (write) +func (p *Producer) ProduceChannel() chan *Message { + return p.produceChannel +} + +// Len returns the number of messages and requests waiting to be transmitted to the broker +// as well as delivery reports queued for the application. +// Includes messages on ProduceChannel. +func (p *Producer) Len() int { + return len(p.produceChannel) + len(p.events) + int(C.rd_kafka_outq_len(p.handle.rk)) +} + +// Flush and wait for outstanding messages and requests to complete delivery. +// Includes messages on ProduceChannel. +// Runs until value reaches zero or on timeoutMs. +// Returns the number of outstanding events still un-flushed. +func (p *Producer) Flush(timeoutMs int) int { + termChan := make(chan bool) // unused stand-in termChan + + d, _ := time.ParseDuration(fmt.Sprintf("%dms", timeoutMs)) + tEnd := time.Now().Add(d) + for p.Len() > 0 { + remain := tEnd.Sub(time.Now()).Seconds() + if remain <= 0.0 { + return p.Len() + } + + p.handle.eventPoll(p.events, + int(math.Min(100, remain*1000)), 1000, termChan) + } + + return 0 +} + +// Close a Producer instance. +// The Producer object or its channels are no longer usable after this call. +func (p *Producer) Close() { + // Wait for poller() (signaled by closing pollerTermChan) + // and channel_producer() (signaled by closing ProduceChannel) + close(p.pollerTermChan) + close(p.produceChannel) + p.handle.waitGroup.Wait() + + close(p.events) + + p.handle.cleanup() + + C.rd_kafka_destroy(p.handle.rk) +} + +const ( + // PurgeInFlight purges messages in-flight to or from the broker. + // Purging these messages will void any future acknowledgements from the + // broker, making it impossible for the application to know if these + // messages were successfully delivered or not. + // Retrying these messages may lead to duplicates. + PurgeInFlight = int(C.RD_KAFKA_PURGE_F_INFLIGHT) + + // PurgeQueue Purge messages in internal queues. + PurgeQueue = int(C.RD_KAFKA_PURGE_F_QUEUE) + + // PurgeNonBlocking Don't wait for background thread queue purging to finish. + PurgeNonBlocking = int(C.RD_KAFKA_PURGE_F_NON_BLOCKING) +) + +// Purge messages currently handled by this producer instance. +// +// flags is a combination of PurgeQueue, PurgeInFlight and PurgeNonBlocking. +// +// The application will need to call Poll(), Flush() or read the Events() channel +// after this call to serve delivery reports for the purged messages. +// +// Messages purged from internal queues fail with the delivery report +// error code set to ErrPurgeQueue, while purged messages that +// are in-flight to or from the broker will fail with the error code set to +// ErrPurgeInflight. +// +// Warning: Purging messages that are in-flight to or from the broker +// will ignore any sub-sequent acknowledgement for these messages +// received from the broker, effectively making it impossible +// for the application to know if the messages were successfully +// produced or not. This may result in duplicate messages if the +// application retries these messages at a later time. +// +// Note: This call may block for a short time while background thread +// queues are purged. +// +// Returns nil on success, ErrInvalidArg if the purge flags are invalid or unknown. +func (p *Producer) Purge(flags int) error { + cErr := C.rd_kafka_purge(p.handle.rk, C.int(flags)) + if cErr != C.RD_KAFKA_RESP_ERR_NO_ERROR { + return newError(cErr) + } + + return nil +} + +// NewProducer creates a new high-level Producer instance. +// +// conf is a *ConfigMap with standard librdkafka configuration properties. +// +// Supported special configuration properties (type, default): +// go.batch.producer (bool, false) - EXPERIMENTAL: Enable batch producer (for increased performance). +// These batches do not relate to Kafka message batches in any way. +// Note: timestamps and headers are not supported with this interface. +// go.delivery.reports (bool, true) - Forward per-message delivery reports to the +// Events() channel. +// go.delivery.report.fields (string, "key,value") - Comma separated list of fields to enable for delivery reports. +// Allowed values: all, none (or empty string), key, value, headers +// Warning: There is a performance penalty to include headers in the delivery report. +// go.events.channel.size (int, 1000000) - Events(). +// go.produce.channel.size (int, 1000000) - ProduceChannel() buffer size (in number of messages) +// go.logs.channel.enable (bool, false) - Forward log to Logs() channel. +// go.logs.channel (chan kafka.LogEvent, nil) - Forward logs to application-provided channel instead of Logs(). Requires go.logs.channel.enable=true. +// +func NewProducer(conf *ConfigMap) (*Producer, error) { + + err := versionCheck() + if err != nil { + return nil, err + } + + p := &Producer{} + + // before we do anything with the configuration, create a copy such that + // the original is not mutated. + confCopy := conf.clone() + + v, err := confCopy.extract("delivery.report.only.error", false) + if v == true { + // FIXME: The filtering of successful DRs must be done in + // the Go client to avoid cgoDr memory leaks. + return nil, newErrorFromString(ErrUnsupportedFeature, + "delivery.report.only.error=true is not currently supported by the Go client") + } + + v, err = confCopy.extract("go.batch.producer", false) + if err != nil { + return nil, err + } + batchProducer := v.(bool) + + v, err = confCopy.extract("go.delivery.reports", true) + if err != nil { + return nil, err + } + p.handle.fwdDr = v.(bool) + + v, err = confCopy.extract("go.delivery.report.fields", "key,value") + if err != nil { + return nil, err + } + + p.handle.msgFields, err = newMessageFieldsFrom(v) + if err != nil { + return nil, err + } + + v, err = confCopy.extract("go.events.channel.size", 1000000) + if err != nil { + return nil, err + } + eventsChanSize := v.(int) + + v, err = confCopy.extract("go.produce.channel.size", 1000000) + if err != nil { + return nil, err + } + produceChannelSize := v.(int) + + logsChanEnable, logsChan, err := confCopy.extractLogConfig() + if err != nil { + return nil, err + } + + if int(C.rd_kafka_version()) < 0x01000000 { + // produce.offset.report is no longer used in librdkafka >= v1.0.0 + v, _ = confCopy.extract("{topic}.produce.offset.report", nil) + if v == nil { + // Enable offset reporting by default, unless overriden. + confCopy.SetKey("{topic}.produce.offset.report", true) + } + } + + // Convert ConfigMap to librdkafka conf_t + cConf, err := confCopy.convert() + if err != nil { + return nil, err + } + + cErrstr := (*C.char)(C.malloc(C.size_t(256))) + defer C.free(unsafe.Pointer(cErrstr)) + + C.rd_kafka_conf_set_events(cConf, C.RD_KAFKA_EVENT_DR|C.RD_KAFKA_EVENT_STATS|C.RD_KAFKA_EVENT_ERROR|C.RD_KAFKA_EVENT_OAUTHBEARER_TOKEN_REFRESH) + + // Create librdkafka producer instance + p.handle.rk = C.rd_kafka_new(C.RD_KAFKA_PRODUCER, cConf, cErrstr, 256) + if p.handle.rk == nil { + return nil, newErrorFromCString(C.RD_KAFKA_RESP_ERR__INVALID_ARG, cErrstr) + } + + p.handle.p = p + p.handle.setup() + p.handle.rkq = C.rd_kafka_queue_get_main(p.handle.rk) + p.events = make(chan Event, eventsChanSize) + p.produceChannel = make(chan *Message, produceChannelSize) + p.pollerTermChan = make(chan bool) + + if logsChanEnable { + p.handle.setupLogQueue(logsChan, p.pollerTermChan) + } + + p.handle.waitGroup.Add(1) + go func() { + poller(p, p.pollerTermChan) + p.handle.waitGroup.Done() + }() + + // non-batch or batch producer, only one must be used + var producer func(*Producer) + if batchProducer { + producer = channelBatchProducer + } else { + producer = channelProducer + } + + p.handle.waitGroup.Add(1) + go func() { + producer(p) + p.handle.waitGroup.Done() + }() + + return p, nil +} + +// channel_producer serves the ProduceChannel channel +func channelProducer(p *Producer) { + for m := range p.produceChannel { + err := p.produce(m, C.RD_KAFKA_MSG_F_BLOCK, nil) + if err != nil { + m.TopicPartition.Error = err + p.events <- m + } + } +} + +// channelBatchProducer serves the ProduceChannel channel and attempts to +// improve cgo performance by using the produceBatch() interface. +func channelBatchProducer(p *Producer) { + var buffered = make(map[string][]*Message) + bufferedCnt := 0 + const batchSize int = 1000000 + totMsgCnt := 0 + totBatchCnt := 0 + + for m := range p.produceChannel { + buffered[*m.TopicPartition.Topic] = append(buffered[*m.TopicPartition.Topic], m) + bufferedCnt++ + + loop2: + for true { + select { + case m, ok := <-p.produceChannel: + if !ok { + break loop2 + } + if m == nil { + panic("nil message received on ProduceChannel") + } + if m.TopicPartition.Topic == nil { + panic(fmt.Sprintf("message without Topic received on ProduceChannel: %v", m)) + } + buffered[*m.TopicPartition.Topic] = append(buffered[*m.TopicPartition.Topic], m) + bufferedCnt++ + if bufferedCnt >= batchSize { + break loop2 + } + default: + break loop2 + } + } + + totBatchCnt++ + totMsgCnt += len(buffered) + + for topic, buffered2 := range buffered { + err := p.produceBatch(topic, buffered2, C.RD_KAFKA_MSG_F_BLOCK) + if err != nil { + for _, m = range buffered2 { + m.TopicPartition.Error = err + p.events <- m + } + } + } + + buffered = make(map[string][]*Message) + bufferedCnt = 0 + } +} + +// poller polls the rd_kafka_t handle for events until signalled for termination +func poller(p *Producer, termChan chan bool) { + for { + select { + case _ = <-termChan: + return + + default: + _, term := p.handle.eventPoll(p.events, 100, 1000, termChan) + if term { + return + } + break + } + } +} + +// GetMetadata queries broker for cluster and topic metadata. +// If topic is non-nil only information about that topic is returned, else if +// allTopics is false only information about locally used topics is returned, +// else information about all topics is returned. +// GetMetadata is equivalent to listTopics, describeTopics and describeCluster in the Java API. +func (p *Producer) GetMetadata(topic *string, allTopics bool, timeoutMs int) (*Metadata, error) { + return getMetadata(p, topic, allTopics, timeoutMs) +} + +// QueryWatermarkOffsets returns the broker's low and high offsets for the given topic +// and partition. +func (p *Producer) QueryWatermarkOffsets(topic string, partition int32, timeoutMs int) (low, high int64, err error) { + return queryWatermarkOffsets(p, topic, partition, timeoutMs) +} + +// OffsetsForTimes looks up offsets by timestamp for the given partitions. +// +// The returned offset for each partition is the earliest offset whose +// timestamp is greater than or equal to the given timestamp in the +// corresponding partition. If the provided timestamp exceeds that of the +// last message in the partition, a value of -1 will be returned. +// +// The timestamps to query are represented as `.Offset` in the `times` +// argument and the looked up offsets are represented as `.Offset` in the returned +// `offsets` list. +// +// The function will block for at most timeoutMs milliseconds. +// +// Duplicate Topic+Partitions are not supported. +// Per-partition errors may be returned in the `.Error` field. +func (p *Producer) OffsetsForTimes(times []TopicPartition, timeoutMs int) (offsets []TopicPartition, err error) { + return offsetsForTimes(p, times, timeoutMs) +} + +// GetFatalError returns an Error object if the client instance has raised a fatal error, else nil. +func (p *Producer) GetFatalError() error { + return getFatalError(p) +} + +// TestFatalError triggers a fatal error in the underlying client. +// This is to be used strictly for testing purposes. +func (p *Producer) TestFatalError(code ErrorCode, str string) ErrorCode { + return testFatalError(p, code, str) +} + +// SetOAuthBearerToken sets the the data to be transmitted +// to a broker during SASL/OAUTHBEARER authentication. It will return nil +// on success, otherwise an error if: +// 1) the token data is invalid (meaning an expiration time in the past +// or either a token value or an extension key or value that does not meet +// the regular expression requirements as per +// https://tools.ietf.org/html/rfc7628#section-3.1); +// 2) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +// 3) SASL/OAUTHBEARER is supported but is not configured as the client's +// authentication mechanism. +func (p *Producer) SetOAuthBearerToken(oauthBearerToken OAuthBearerToken) error { + return p.handle.setOAuthBearerToken(oauthBearerToken) +} + +// SetOAuthBearerTokenFailure sets the error message describing why token +// retrieval/setting failed; it also schedules a new token refresh event for 10 +// seconds later so the attempt may be retried. It will return nil on +// success, otherwise an error if: +// 1) SASL/OAUTHBEARER is not supported by the underlying librdkafka build; +// 2) SASL/OAUTHBEARER is supported but is not configured as the client's +// authentication mechanism. +func (p *Producer) SetOAuthBearerTokenFailure(errstr string) error { + return p.handle.setOAuthBearerTokenFailure(errstr) +} + +// Transactional API + +// InitTransactions Initializes transactions for the producer instance. +// +// This function ensures any transactions initiated by previous instances +// of the producer with the same `transactional.id` are completed. +// If the previous instance failed with a transaction in progress the +// previous transaction will be aborted. +// This function needs to be called before any other transactional or +// produce functions are called when the `transactional.id` is configured. +// +// If the last transaction had begun completion (following transaction commit) +// but not yet finished, this function will await the previous transaction's +// completion. +// +// When any previous transactions have been fenced this function +// will acquire the internal producer id and epoch, used in all future +// transactional messages issued by this producer instance. +// +// Upon successful return from this function the application has to perform at +// least one of the following operations within `transaction.timeout.ms` to +// avoid timing out the transaction on the broker: +// * `Produce()` (et.al) +// * `SendOffsetsToTransaction()` +// * `CommitTransaction()` +// * `AbortTransaction()` +// +// Parameters: +// * `ctx` - The maximum time to block, or nil for indefinite. +// On timeout the operation may continue in the background, +// depending on state, and it is okay to call `InitTransactions()` +// again. +// +// Returns nil on success or an error on failure. +// Check whether the returned error object permits retrying +// by calling `err.(kafka.Error).IsRetriable()`, or whether a fatal +// error has been raised by calling `err.(kafka.Error).IsFatal()`. +func (p *Producer) InitTransactions(ctx context.Context) error { + cError := C.rd_kafka_init_transactions(p.handle.rk, + cTimeoutFromContext(ctx)) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} + +// BeginTransaction starts a new transaction. +// +// `InitTransactions()` must have been called successfully (once) +// before this function is called. +// +// Any messages produced, offsets sent (`SendOffsetsToTransaction()`), +// etc, after the successful return of this function will be part of +// the transaction and committed or aborted atomatically. +// +// Finish the transaction by calling `CommitTransaction()` or +// abort the transaction by calling `AbortTransaction()`. +// +// Returns nil on success or an error object on failure. +// Check whether a fatal error has been raised by +// calling `err.(kafka.Error).IsFatal()`. +// +// Note: With the transactional producer, `Produce()`, et.al, are only +// allowed during an on-going transaction, as started with this function. +// Any produce call outside an on-going transaction, or for a failed +// transaction, will fail. +func (p *Producer) BeginTransaction() error { + cError := C.rd_kafka_begin_transaction(p.handle.rk) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} + +// SendOffsetsToTransaction sends a list of topic partition offsets to the +// consumer group coordinator for `consumerMetadata`, and marks the offsets +// as part part of the current transaction. +// These offsets will be considered committed only if the transaction is +// committed successfully. +// +// The offsets should be the next message your application will consume, +// i.e., the last processed message's offset + 1 for each partition. +// Either track the offsets manually during processing or use +// `consumer.Position()` (on the consumer) to get the current offsets for +// the partitions assigned to the consumer. +// +// Use this method at the end of a consume-transform-produce loop prior +// to committing the transaction with `CommitTransaction()`. +// +// Parameters: +// * `ctx` - The maximum amount of time to block, or nil for indefinite. +// * `offsets` - List of offsets to commit to the consumer group upon +// successful commit of the transaction. Offsets should be +// the next message to consume, e.g., last processed message + 1. +// * `consumerMetadata` - The current consumer group metadata as returned by +// `consumer.GetConsumerGroupMetadata()` on the consumer +// instance the provided offsets were consumed from. +// +// Note: The consumer must disable auto commits (set `enable.auto.commit` to false on the consumer). +// +// Note: Logical and invalid offsets (e.g., OffsetInvalid) in +// `offsets` will be ignored. If there are no valid offsets in +// `offsets` the function will return nil and no action will be taken. +// +// Returns nil on success or an error object on failure. +// Check whether the returned error object permits retrying +// by calling `err.(kafka.Error).IsRetriable()`, or whether an abortable +// or fatal error has been raised by calling +// `err.(kafka.Error).TxnRequiresAbort()` or `err.(kafka.Error).IsFatal()` +// respectively. +func (p *Producer) SendOffsetsToTransaction(ctx context.Context, offsets []TopicPartition, consumerMetadata *ConsumerGroupMetadata) error { + var cOffsets *C.rd_kafka_topic_partition_list_t + if offsets != nil { + cOffsets = newCPartsFromTopicPartitions(offsets) + defer C.rd_kafka_topic_partition_list_destroy(cOffsets) + } + + cgmd, err := deserializeConsumerGroupMetadata(consumerMetadata.serialized) + if err != nil { + return err + } + defer C.rd_kafka_consumer_group_metadata_destroy(cgmd) + + cError := C.rd_kafka_send_offsets_to_transaction( + p.handle.rk, + cOffsets, + cgmd, + cTimeoutFromContext(ctx)) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} + +// CommitTransaction commits the current transaction. +// +// Any outstanding messages will be flushed (delivered) before actually +// committing the transaction. +// +// If any of the outstanding messages fail permanently the current +// transaction will enter the abortable error state and this +// function will return an abortable error, in this case the application +// must call `AbortTransaction()` before attempting a new +// transaction with `BeginTransaction()`. +// +// Parameters: +// * `ctx` - The maximum amount of time to block, or nil for indefinite. +// +// Note: This function will block until all outstanding messages are +// delivered and the transaction commit request has been successfully +// handled by the transaction coordinator, or until the `ctx` expires, +// which ever comes first. On timeout the application may +// call the function again. +// +// Note: Will automatically call `Flush()` to ensure all queued +// messages are delivered before attempting to commit the transaction. +// The application MUST serve the `producer.Events()` channel for delivery +// reports in a separate go-routine during this time. +// +// Returns nil on success or an error object on failure. +// Check whether the returned error object permits retrying +// by calling `err.(kafka.Error).IsRetriable()`, or whether an abortable +// or fatal error has been raised by calling +// `err.(kafka.Error).TxnRequiresAbort()` or `err.(kafka.Error).IsFatal()` +// respectively. +func (p *Producer) CommitTransaction(ctx context.Context) error { + cError := C.rd_kafka_commit_transaction(p.handle.rk, + cTimeoutFromContext(ctx)) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} + +// AbortTransaction aborts the ongoing transaction. +// +// This function should also be used to recover from non-fatal abortable +// transaction errors. +// +// Any outstanding messages will be purged and fail with +// `ErrPurgeInflight` or `ErrPurgeQueue`. +// +// Parameters: +// * `ctx` - The maximum amount of time to block, or nil for indefinite. +// +// Note: This function will block until all outstanding messages are purged +// and the transaction abort request has been successfully +// handled by the transaction coordinator, or until the `ctx` expires, +// which ever comes first. On timeout the application may +// call the function again. +// +// Note: Will automatically call `Purge()` and `Flush()` to ensure all queued +// and in-flight messages are purged before attempting to abort the transaction. +// The application MUST serve the `producer.Events()` channel for delivery +// reports in a separate go-routine during this time. +// +// Returns nil on success or an error object on failure. +// Check whether the returned error object permits retrying +// by calling `err.(kafka.Error).IsRetriable()`, or whether a fatal error +// has been raised by calling `err.(kafka.Error).IsFatal()`. +func (p *Producer) AbortTransaction(ctx context.Context) error { + cError := C.rd_kafka_abort_transaction(p.handle.rk, + cTimeoutFromContext(ctx)) + if cError != nil { + return newErrorFromCErrorDestroy(cError) + } + + return nil +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/select_rdkafka.h b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/select_rdkafka.h new file mode 100644 index 0000000..98fe330 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/select_rdkafka.h @@ -0,0 +1,29 @@ +/** + * Copyright 2020 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +// This file uses a preprocessor macro defined by the various build_*.go +// files to determine whether to import the bundled librdkafka header, or +// the system one. +// This is needed because cgo will automatically add -I. to the include +// path, so would find a bundled header instead of +// the system one if it were called librdkafka/rdkafka.h instead of +// librdkafka_vendor/rdkafka.h + +#ifdef USE_VENDORED_LIBRDKAFKA +#include "librdkafka_vendor/rdkafka.h" +#else +#include +#endif diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/testconf-example.json b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/testconf-example.json new file mode 100644 index 0000000..7024a9c --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/testconf-example.json @@ -0,0 +1,8 @@ +{ + "Brokers": "mybroker or $BROKERS env", + "Topic": "test", + "GroupID": "testgroup", + "PerfMsgCount": 1000000, + "PerfMsgSize": 100, + "Config": ["api.version.request=true"] +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/testhelpers.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/testhelpers.go new file mode 100644 index 0000000..f9db896 --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/testhelpers.go @@ -0,0 +1,248 @@ +package kafka + +/** + * Copyright 2016 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +import ( + "context" + "encoding/json" + "fmt" + "math/rand" + "os" + "testing" + "time" +) + +/* +#include "select_rdkafka.h" +*/ +import "C" + +var testconf struct { + Brokers string + Topic string + GroupID string + PerfMsgCount int + PerfMsgSize int + Config []string + conf ConfigMap +} + +// testconf_read reads the test suite config file testconf.json which must +// contain at least Brokers and Topic string properties. +// Returns true if the testconf was found and usable, false if no such file, or panics +// if the file format is wrong. +func testconfRead() bool { + cf, err := os.Open("testconf.json") + if err != nil { + fmt.Fprintf(os.Stderr, "%% testconf.json not found - ignoring test\n") + return false + } + + // Default values + testconf.PerfMsgCount = 2000000 + testconf.PerfMsgSize = 100 + testconf.GroupID = "testgroup" + + jp := json.NewDecoder(cf) + err = jp.Decode(&testconf) + if err != nil { + panic(fmt.Sprintf("Failed to parse testconf: %s", err)) + } + + cf.Close() + + if testconf.Brokers[0] == '$' { + // Read broker list from environment variable + testconf.Brokers = os.Getenv(testconf.Brokers[1:]) + } + + if testconf.Brokers == "" || testconf.Topic == "" { + panic("Missing Brokers or Topic in testconf.json") + } + + return true +} + +// update existing ConfigMap with key=value pairs from testconf.Config +func (cm *ConfigMap) updateFromTestconf() error { + if testconf.Config == nil { + return nil + } + + // Translate "key=value" pairs in Config to ConfigMap + for _, s := range testconf.Config { + err := cm.Set(s) + if err != nil { + return err + } + } + + return nil + +} + +// Return the number of messages available in all partitions of a topic. +// WARNING: This uses watermark offsets so it will be incorrect for compacted topics. +func getMessageCountInTopic(topic string) (int, error) { + + // Create consumer + config := &ConfigMap{"bootstrap.servers": testconf.Brokers, + "group.id": testconf.GroupID} + config.updateFromTestconf() + + c, err := NewConsumer(config) + if err != nil { + return 0, err + } + defer c.Close() + + // get metadata for the topic to find out number of partitions + + metadata, err := c.GetMetadata(&topic, false, 5*1000) + if err != nil { + return 0, err + } + + t, ok := metadata.Topics[topic] + if !ok { + return 0, newError(C.RD_KAFKA_RESP_ERR__UNKNOWN_TOPIC) + } + + cnt := 0 + for _, p := range t.Partitions { + low, high, err := c.QueryWatermarkOffsets(topic, p.ID, 5*1000) + if err != nil { + continue + } + cnt += int(high - low) + } + + return cnt, nil +} + +// getBrokerList returns a list of brokers (ids) in the cluster +func getBrokerList(H Handle) (brokers []int32, err error) { + md, err := getMetadata(H, nil, true, 15*1000) + if err != nil { + return nil, err + } + + brokers = make([]int32, len(md.Brokers)) + for i, mdBroker := range md.Brokers { + brokers[i] = mdBroker.ID + } + + return brokers, nil +} + +// waitTopicInMetadata waits for the given topic to show up in metadata +func waitTopicInMetadata(H Handle, topic string, timeoutMs int) error { + d, _ := time.ParseDuration(fmt.Sprintf("%dms", timeoutMs)) + tEnd := time.Now().Add(d) + + for { + remain := tEnd.Sub(time.Now()).Seconds() + if remain < 0.0 { + return newErrorFromString(ErrTimedOut, + fmt.Sprintf("Timed out waiting for topic %s to appear in metadata", topic)) + } + + md, err := getMetadata(H, nil, true, int(remain*1000)) + if err != nil { + return err + } + + for _, t := range md.Topics { + if t.Topic != topic { + continue + } + if t.Error.Code() != ErrNoError || len(t.Partitions) < 1 { + continue + } + // Proper topic found in metadata + return nil + } + + time.Sleep(500 * 1000) // 500ms + } + +} + +func createAdminClient(t *testing.T) (a *AdminClient) { + numver, strver := LibraryVersion() + if numver < 0x000b0500 { + t.Skipf("Requires librdkafka >=0.11.5 (currently on %s, 0x%x)", strver, numver) + } + + if !testconfRead() { + t.Skipf("Missing testconf.json") + } + + conf := ConfigMap{"bootstrap.servers": testconf.Brokers} + conf.updateFromTestconf() + + /* + * Create producer and produce a couple of messages with and without + * headers. + */ + a, err := NewAdminClient(&conf) + if err != nil { + t.Fatalf("NewAdminClient: %v", err) + } + + return a +} + +func createTestTopic(t *testing.T, suffix string, numPartitions int, replicationFactor int) string { + rand.Seed(time.Now().Unix()) + + topic := fmt.Sprintf("%s-%s-%d", testconf.Topic, suffix, rand.Intn(100000)) + + a := createAdminClient(t) + defer a.Close() + + newTopics := []TopicSpecification{ + { + Topic: topic, + NumPartitions: numPartitions, + ReplicationFactor: replicationFactor, + }, + } + + maxDuration, err := time.ParseDuration("30s") + if err != nil { + t.Fatalf("%s", err) + } + + ctx, cancel := context.WithTimeout(context.Background(), maxDuration) + defer cancel() + result, err := a.CreateTopics(ctx, newTopics, nil) + if err != nil { + t.Fatalf("CreateTopics() failed: %s", err) + } + + for _, res := range result { + if res.Error.Code() != ErrNoError { + t.Errorf("Failed to create topic %s: %s\n", + res.Topic, res.Error) + continue + } + + } + + return topic +} diff --git a/vendor/github.com/confluentinc/confluent-kafka-go/kafka/time.go b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/time.go new file mode 100644 index 0000000..ff93f0a --- /dev/null +++ b/vendor/github.com/confluentinc/confluent-kafka-go/kafka/time.go @@ -0,0 +1,55 @@ +/** + * Copyright 2019 Confluent Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package kafka + +import "C" + +import ( + "context" + "time" +) + +const ( + cTimeoutInfinite = C.int(-1) // Blocks indefinitely until completion. + cTimeoutNoWait = C.int(0) // Returns immediately without blocking. +) + +// cTimeoutFromContext returns the remaining time after which work done on behalf of this context +// should be canceled, in milliseconds. +// +// If no deadline/timeout is set, or if the timeout does not fit in an int32, it returns +// cTimeoutInfinite; +// If there is no time left in this context, it returns cTimeoutNoWait. +func cTimeoutFromContext(ctx context.Context) C.int { + if ctx == nil { + return cTimeoutInfinite + } + timeout, hasTimeout := timeout(ctx) + if !hasTimeout { + return cTimeoutInfinite + } + if timeout <= 0 { + return cTimeoutNoWait + } + + timeoutMs := int64(timeout / time.Millisecond) + if int64(int32(timeoutMs)) < timeoutMs { + return cTimeoutInfinite + } + + return C.int(timeoutMs) +} diff --git a/vendor/github.com/go-stack/stack/LICENSE.md b/vendor/github.com/go-stack/stack/LICENSE.md new file mode 100644 index 0000000..2abf98e --- /dev/null +++ b/vendor/github.com/go-stack/stack/LICENSE.md @@ -0,0 +1,21 @@ +The MIT License (MIT) + +Copyright (c) 2014 Chris Hines + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. diff --git a/vendor/github.com/go-stack/stack/README.md b/vendor/github.com/go-stack/stack/README.md new file mode 100644 index 0000000..f11cccc --- /dev/null +++ b/vendor/github.com/go-stack/stack/README.md @@ -0,0 +1,38 @@ +[![GoDoc](https://godoc.org/github.com/go-stack/stack?status.svg)](https://godoc.org/github.com/go-stack/stack) +[![Go Report Card](https://goreportcard.com/badge/go-stack/stack)](https://goreportcard.com/report/go-stack/stack) +[![TravisCI](https://travis-ci.org/go-stack/stack.svg?branch=master)](https://travis-ci.org/go-stack/stack) +[![Coverage Status](https://coveralls.io/repos/github/go-stack/stack/badge.svg?branch=master)](https://coveralls.io/github/go-stack/stack?branch=master) + +# stack + +Package stack implements utilities to capture, manipulate, and format call +stacks. It provides a simpler API than package runtime. + +The implementation takes care of the minutia and special cases of interpreting +the program counter (pc) values returned by runtime.Callers. + +## Versioning + +Package stack publishes releases via [semver](http://semver.org/) compatible Git +tags prefixed with a single 'v'. The master branch always contains the latest +release. The develop branch contains unreleased commits. + +## Formatting + +Package stack's types implement fmt.Formatter, which provides a simple and +flexible way to declaratively configure formatting when used with logging or +error tracking packages. + +```go +func DoTheThing() { + c := stack.Caller(0) + log.Print(c) // "source.go:10" + log.Printf("%+v", c) // "pkg/path/source.go:10" + log.Printf("%n", c) // "DoTheThing" + + s := stack.Trace().TrimRuntime() + log.Print(s) // "[source.go:15 caller.go:42 main.go:14]" +} +``` + +See the docs for all of the supported formatting options. diff --git a/vendor/github.com/go-stack/stack/stack.go b/vendor/github.com/go-stack/stack/stack.go new file mode 100644 index 0000000..ac3b93b --- /dev/null +++ b/vendor/github.com/go-stack/stack/stack.go @@ -0,0 +1,400 @@ +// +build go1.7 + +// Package stack implements utilities to capture, manipulate, and format call +// stacks. It provides a simpler API than package runtime. +// +// The implementation takes care of the minutia and special cases of +// interpreting the program counter (pc) values returned by runtime.Callers. +// +// Package stack's types implement fmt.Formatter, which provides a simple and +// flexible way to declaratively configure formatting when used with logging +// or error tracking packages. +package stack + +import ( + "bytes" + "errors" + "fmt" + "io" + "runtime" + "strconv" + "strings" +) + +// Call records a single function invocation from a goroutine stack. +type Call struct { + frame runtime.Frame +} + +// Caller returns a Call from the stack of the current goroutine. The argument +// skip is the number of stack frames to ascend, with 0 identifying the +// calling function. +func Caller(skip int) Call { + // As of Go 1.9 we need room for up to three PC entries. + // + // 0. An entry for the stack frame prior to the target to check for + // special handling needed if that prior entry is runtime.sigpanic. + // 1. A possible second entry to hold metadata about skipped inlined + // functions. If inline functions were not skipped the target frame + // PC will be here. + // 2. A third entry for the target frame PC when the second entry + // is used for skipped inline functions. + var pcs [3]uintptr + n := runtime.Callers(skip+1, pcs[:]) + frames := runtime.CallersFrames(pcs[:n]) + frame, _ := frames.Next() + frame, _ = frames.Next() + + return Call{ + frame: frame, + } +} + +// String implements fmt.Stinger. It is equivalent to fmt.Sprintf("%v", c). +func (c Call) String() string { + return fmt.Sprint(c) +} + +// MarshalText implements encoding.TextMarshaler. It formats the Call the same +// as fmt.Sprintf("%v", c). +func (c Call) MarshalText() ([]byte, error) { + if c.frame == (runtime.Frame{}) { + return nil, ErrNoFunc + } + + buf := bytes.Buffer{} + fmt.Fprint(&buf, c) + return buf.Bytes(), nil +} + +// ErrNoFunc means that the Call has a nil *runtime.Func. The most likely +// cause is a Call with the zero value. +var ErrNoFunc = errors.New("no call stack information") + +// Format implements fmt.Formatter with support for the following verbs. +// +// %s source file +// %d line number +// %n function name +// %k last segment of the package path +// %v equivalent to %s:%d +// +// It accepts the '+' and '#' flags for most of the verbs as follows. +// +// %+s path of source file relative to the compile time GOPATH, +// or the module path joined to the path of source file relative +// to module root +// %#s full path of source file +// %+n import path qualified function name +// %+k full package path +// %+v equivalent to %+s:%d +// %#v equivalent to %#s:%d +func (c Call) Format(s fmt.State, verb rune) { + if c.frame == (runtime.Frame{}) { + fmt.Fprintf(s, "%%!%c(NOFUNC)", verb) + return + } + + switch verb { + case 's', 'v': + file := c.frame.File + switch { + case s.Flag('#'): + // done + case s.Flag('+'): + file = pkgFilePath(&c.frame) + default: + const sep = "/" + if i := strings.LastIndex(file, sep); i != -1 { + file = file[i+len(sep):] + } + } + io.WriteString(s, file) + if verb == 'v' { + buf := [7]byte{':'} + s.Write(strconv.AppendInt(buf[:1], int64(c.frame.Line), 10)) + } + + case 'd': + buf := [6]byte{} + s.Write(strconv.AppendInt(buf[:0], int64(c.frame.Line), 10)) + + case 'k': + name := c.frame.Function + const pathSep = "/" + start, end := 0, len(name) + if i := strings.LastIndex(name, pathSep); i != -1 { + start = i + len(pathSep) + } + const pkgSep = "." + if i := strings.Index(name[start:], pkgSep); i != -1 { + end = start + i + } + if s.Flag('+') { + start = 0 + } + io.WriteString(s, name[start:end]) + + case 'n': + name := c.frame.Function + if !s.Flag('+') { + const pathSep = "/" + if i := strings.LastIndex(name, pathSep); i != -1 { + name = name[i+len(pathSep):] + } + const pkgSep = "." + if i := strings.Index(name, pkgSep); i != -1 { + name = name[i+len(pkgSep):] + } + } + io.WriteString(s, name) + } +} + +// Frame returns the call frame infomation for the Call. +func (c Call) Frame() runtime.Frame { + return c.frame +} + +// PC returns the program counter for this call frame; multiple frames may +// have the same PC value. +// +// Deprecated: Use Call.Frame instead. +func (c Call) PC() uintptr { + return c.frame.PC +} + +// CallStack records a sequence of function invocations from a goroutine +// stack. +type CallStack []Call + +// String implements fmt.Stinger. It is equivalent to fmt.Sprintf("%v", cs). +func (cs CallStack) String() string { + return fmt.Sprint(cs) +} + +var ( + openBracketBytes = []byte("[") + closeBracketBytes = []byte("]") + spaceBytes = []byte(" ") +) + +// MarshalText implements encoding.TextMarshaler. It formats the CallStack the +// same as fmt.Sprintf("%v", cs). +func (cs CallStack) MarshalText() ([]byte, error) { + buf := bytes.Buffer{} + buf.Write(openBracketBytes) + for i, pc := range cs { + if i > 0 { + buf.Write(spaceBytes) + } + fmt.Fprint(&buf, pc) + } + buf.Write(closeBracketBytes) + return buf.Bytes(), nil +} + +// Format implements fmt.Formatter by printing the CallStack as square brackets +// ([, ]) surrounding a space separated list of Calls each formatted with the +// supplied verb and options. +func (cs CallStack) Format(s fmt.State, verb rune) { + s.Write(openBracketBytes) + for i, pc := range cs { + if i > 0 { + s.Write(spaceBytes) + } + pc.Format(s, verb) + } + s.Write(closeBracketBytes) +} + +// Trace returns a CallStack for the current goroutine with element 0 +// identifying the calling function. +func Trace() CallStack { + var pcs [512]uintptr + n := runtime.Callers(1, pcs[:]) + + frames := runtime.CallersFrames(pcs[:n]) + cs := make(CallStack, 0, n) + + // Skip extra frame retrieved just to make sure the runtime.sigpanic + // special case is handled. + frame, more := frames.Next() + + for more { + frame, more = frames.Next() + cs = append(cs, Call{frame: frame}) + } + + return cs +} + +// TrimBelow returns a slice of the CallStack with all entries below c +// removed. +func (cs CallStack) TrimBelow(c Call) CallStack { + for len(cs) > 0 && cs[0] != c { + cs = cs[1:] + } + return cs +} + +// TrimAbove returns a slice of the CallStack with all entries above c +// removed. +func (cs CallStack) TrimAbove(c Call) CallStack { + for len(cs) > 0 && cs[len(cs)-1] != c { + cs = cs[:len(cs)-1] + } + return cs +} + +// pkgIndex returns the index that results in file[index:] being the path of +// file relative to the compile time GOPATH, and file[:index] being the +// $GOPATH/src/ portion of file. funcName must be the name of a function in +// file as returned by runtime.Func.Name. +func pkgIndex(file, funcName string) int { + // As of Go 1.6.2 there is no direct way to know the compile time GOPATH + // at runtime, but we can infer the number of path segments in the GOPATH. + // We note that runtime.Func.Name() returns the function name qualified by + // the import path, which does not include the GOPATH. Thus we can trim + // segments from the beginning of the file path until the number of path + // separators remaining is one more than the number of path separators in + // the function name. For example, given: + // + // GOPATH /home/user + // file /home/user/src/pkg/sub/file.go + // fn.Name() pkg/sub.Type.Method + // + // We want to produce: + // + // file[:idx] == /home/user/src/ + // file[idx:] == pkg/sub/file.go + // + // From this we can easily see that fn.Name() has one less path separator + // than our desired result for file[idx:]. We count separators from the + // end of the file path until it finds two more than in the function name + // and then move one character forward to preserve the initial path + // segment without a leading separator. + const sep = "/" + i := len(file) + for n := strings.Count(funcName, sep) + 2; n > 0; n-- { + i = strings.LastIndex(file[:i], sep) + if i == -1 { + i = -len(sep) + break + } + } + // get back to 0 or trim the leading separator + return i + len(sep) +} + +// pkgFilePath returns the frame's filepath relative to the compile-time GOPATH, +// or its module path joined to its path relative to the module root. +// +// As of Go 1.11 there is no direct way to know the compile time GOPATH or +// module paths at runtime, but we can piece together the desired information +// from available information. We note that runtime.Frame.Function contains the +// function name qualified by the package path, which includes the module path +// but not the GOPATH. We can extract the package path from that and append the +// last segments of the file path to arrive at the desired package qualified +// file path. For example, given: +// +// GOPATH /home/user +// import path pkg/sub +// frame.File /home/user/src/pkg/sub/file.go +// frame.Function pkg/sub.Type.Method +// Desired return pkg/sub/file.go +// +// It appears that we simply need to trim ".Type.Method" from frame.Function and +// append "/" + path.Base(file). +// +// But there are other wrinkles. Although it is idiomatic to do so, the internal +// name of a package is not required to match the last segment of its import +// path. In addition, the introduction of modules in Go 1.11 allows working +// without a GOPATH. So we also must make these work right: +// +// GOPATH /home/user +// import path pkg/go-sub +// package name sub +// frame.File /home/user/src/pkg/go-sub/file.go +// frame.Function pkg/sub.Type.Method +// Desired return pkg/go-sub/file.go +// +// Module path pkg/v2 +// import path pkg/v2/go-sub +// package name sub +// frame.File /home/user/cloned-pkg/go-sub/file.go +// frame.Function pkg/v2/sub.Type.Method +// Desired return pkg/v2/go-sub/file.go +// +// We can handle all of these situations by using the package path extracted +// from frame.Function up to, but not including, the last segment as the prefix +// and the last two segments of frame.File as the suffix of the returned path. +// This preserves the existing behavior when working in a GOPATH without modules +// and a semantically equivalent behavior when used in module aware project. +func pkgFilePath(frame *runtime.Frame) string { + pre := pkgPrefix(frame.Function) + post := pathSuffix(frame.File) + if pre == "" { + return post + } + return pre + "/" + post +} + +// pkgPrefix returns the import path of the function's package with the final +// segment removed. +func pkgPrefix(funcName string) string { + const pathSep = "/" + end := strings.LastIndex(funcName, pathSep) + if end == -1 { + return "" + } + return funcName[:end] +} + +// pathSuffix returns the last two segments of path. +func pathSuffix(path string) string { + const pathSep = "/" + lastSep := strings.LastIndex(path, pathSep) + if lastSep == -1 { + return path + } + return path[strings.LastIndex(path[:lastSep], pathSep)+1:] +} + +var runtimePath string + +func init() { + var pcs [3]uintptr + runtime.Callers(0, pcs[:]) + frames := runtime.CallersFrames(pcs[:]) + frame, _ := frames.Next() + file := frame.File + + idx := pkgIndex(frame.File, frame.Function) + + runtimePath = file[:idx] + if runtime.GOOS == "windows" { + runtimePath = strings.ToLower(runtimePath) + } +} + +func inGoroot(c Call) bool { + file := c.frame.File + if len(file) == 0 || file[0] == '?' { + return true + } + if runtime.GOOS == "windows" { + file = strings.ToLower(file) + } + return strings.HasPrefix(file, runtimePath) || strings.HasSuffix(file, "/_testmain.go") +} + +// TrimRuntime returns a slice of the CallStack with the topmost entries from +// the go runtime removed. It considers any calls originating from unknown +// files, files under GOROOT, or _testmain.go as part of the runtime. +func (cs CallStack) TrimRuntime() CallStack { + for len(cs) > 0 && inGoroot(cs[len(cs)-1]) { + cs = cs[:len(cs)-1] + } + return cs +} diff --git a/vendor/github.com/golang/snappy/.gitignore b/vendor/github.com/golang/snappy/.gitignore new file mode 100644 index 0000000..042091d --- /dev/null +++ b/vendor/github.com/golang/snappy/.gitignore @@ -0,0 +1,16 @@ +cmd/snappytool/snappytool +testdata/bench + +# These explicitly listed benchmark data files are for an obsolete version of +# snappy_test.go. +testdata/alice29.txt +testdata/asyoulik.txt +testdata/fireworks.jpeg +testdata/geo.protodata +testdata/html +testdata/html_x_4 +testdata/kppkn.gtb +testdata/lcet10.txt +testdata/paper-100k.pdf +testdata/plrabn12.txt +testdata/urls.10K diff --git a/vendor/github.com/golang/snappy/AUTHORS b/vendor/github.com/golang/snappy/AUTHORS new file mode 100644 index 0000000..52ccb5a --- /dev/null +++ b/vendor/github.com/golang/snappy/AUTHORS @@ -0,0 +1,18 @@ +# This is the official list of Snappy-Go authors for copyright purposes. +# This file is distinct from the CONTRIBUTORS files. +# See the latter for an explanation. + +# Names should be added to this file as +# Name or Organization +# The email address is not required for organizations. + +# Please keep the list sorted. + +Amazon.com, Inc +Damian Gryski +Eric Buth +Google Inc. +Jan Mercl <0xjnml@gmail.com> +Klaus Post +Rodolfo Carvalho +Sebastien Binet diff --git a/vendor/github.com/golang/snappy/CONTRIBUTORS b/vendor/github.com/golang/snappy/CONTRIBUTORS new file mode 100644 index 0000000..ea6524d --- /dev/null +++ b/vendor/github.com/golang/snappy/CONTRIBUTORS @@ -0,0 +1,41 @@ +# This is the official list of people who can contribute +# (and typically have contributed) code to the Snappy-Go repository. +# The AUTHORS file lists the copyright holders; this file +# lists people. For example, Google employees are listed here +# but not in AUTHORS, because Google holds the copyright. +# +# The submission process automatically checks to make sure +# that people submitting code are listed in this file (by email address). +# +# Names should be added to this file only after verifying that +# the individual or the individual's organization has agreed to +# the appropriate Contributor License Agreement, found here: +# +# http://code.google.com/legal/individual-cla-v1.0.html +# http://code.google.com/legal/corporate-cla-v1.0.html +# +# The agreement for individuals can be filled out on the web. +# +# When adding J Random Contributor's name to this file, +# either J's name or J's organization's name should be +# added to the AUTHORS file, depending on whether the +# individual or corporate CLA was used. + +# Names should be added to this file like so: +# Name + +# Please keep the list sorted. + +Alex Legg +Damian Gryski +Eric Buth +Jan Mercl <0xjnml@gmail.com> +Jonathan Swinney +Kai Backman +Klaus Post +Marc-Antoine Ruel +Nigel Tao +Rob Pike +Rodolfo Carvalho +Russ Cox +Sebastien Binet diff --git a/vendor/github.com/golang/snappy/LICENSE b/vendor/github.com/golang/snappy/LICENSE new file mode 100644 index 0000000..6050c10 --- /dev/null +++ b/vendor/github.com/golang/snappy/LICENSE @@ -0,0 +1,27 @@ +Copyright (c) 2011 The Snappy-Go Authors. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/github.com/golang/snappy/README b/vendor/github.com/golang/snappy/README new file mode 100644 index 0000000..cea1287 --- /dev/null +++ b/vendor/github.com/golang/snappy/README @@ -0,0 +1,107 @@ +The Snappy compression format in the Go programming language. + +To download and install from source: +$ go get github.com/golang/snappy + +Unless otherwise noted, the Snappy-Go source files are distributed +under the BSD-style license found in the LICENSE file. + + + +Benchmarks. + +The golang/snappy benchmarks include compressing (Z) and decompressing (U) ten +or so files, the same set used by the C++ Snappy code (github.com/google/snappy +and note the "google", not "golang"). On an "Intel(R) Core(TM) i7-3770 CPU @ +3.40GHz", Go's GOARCH=amd64 numbers as of 2016-05-29: + +"go test -test.bench=." + +_UFlat0-8 2.19GB/s ± 0% html +_UFlat1-8 1.41GB/s ± 0% urls +_UFlat2-8 23.5GB/s ± 2% jpg +_UFlat3-8 1.91GB/s ± 0% jpg_200 +_UFlat4-8 14.0GB/s ± 1% pdf +_UFlat5-8 1.97GB/s ± 0% html4 +_UFlat6-8 814MB/s ± 0% txt1 +_UFlat7-8 785MB/s ± 0% txt2 +_UFlat8-8 857MB/s ± 0% txt3 +_UFlat9-8 719MB/s ± 1% txt4 +_UFlat10-8 2.84GB/s ± 0% pb +_UFlat11-8 1.05GB/s ± 0% gaviota + +_ZFlat0-8 1.04GB/s ± 0% html +_ZFlat1-8 534MB/s ± 0% urls +_ZFlat2-8 15.7GB/s ± 1% jpg +_ZFlat3-8 740MB/s ± 3% jpg_200 +_ZFlat4-8 9.20GB/s ± 1% pdf +_ZFlat5-8 991MB/s ± 0% html4 +_ZFlat6-8 379MB/s ± 0% txt1 +_ZFlat7-8 352MB/s ± 0% txt2 +_ZFlat8-8 396MB/s ± 1% txt3 +_ZFlat9-8 327MB/s ± 1% txt4 +_ZFlat10-8 1.33GB/s ± 1% pb +_ZFlat11-8 605MB/s ± 1% gaviota + + + +"go test -test.bench=. -tags=noasm" + +_UFlat0-8 621MB/s ± 2% html +_UFlat1-8 494MB/s ± 1% urls +_UFlat2-8 23.2GB/s ± 1% jpg +_UFlat3-8 1.12GB/s ± 1% jpg_200 +_UFlat4-8 4.35GB/s ± 1% pdf +_UFlat5-8 609MB/s ± 0% html4 +_UFlat6-8 296MB/s ± 0% txt1 +_UFlat7-8 288MB/s ± 0% txt2 +_UFlat8-8 309MB/s ± 1% txt3 +_UFlat9-8 280MB/s ± 1% txt4 +_UFlat10-8 753MB/s ± 0% pb +_UFlat11-8 400MB/s ± 0% gaviota + +_ZFlat0-8 409MB/s ± 1% html +_ZFlat1-8 250MB/s ± 1% urls +_ZFlat2-8 12.3GB/s ± 1% jpg +_ZFlat3-8 132MB/s ± 0% jpg_200 +_ZFlat4-8 2.92GB/s ± 0% pdf +_ZFlat5-8 405MB/s ± 1% html4 +_ZFlat6-8 179MB/s ± 1% txt1 +_ZFlat7-8 170MB/s ± 1% txt2 +_ZFlat8-8 189MB/s ± 1% txt3 +_ZFlat9-8 164MB/s ± 1% txt4 +_ZFlat10-8 479MB/s ± 1% pb +_ZFlat11-8 270MB/s ± 1% gaviota + + + +For comparison (Go's encoded output is byte-for-byte identical to C++'s), here +are the numbers from C++ Snappy's + +make CXXFLAGS="-O2 -DNDEBUG -g" clean snappy_unittest.log && cat snappy_unittest.log + +BM_UFlat/0 2.4GB/s html +BM_UFlat/1 1.4GB/s urls +BM_UFlat/2 21.8GB/s jpg +BM_UFlat/3 1.5GB/s jpg_200 +BM_UFlat/4 13.3GB/s pdf +BM_UFlat/5 2.1GB/s html4 +BM_UFlat/6 1.0GB/s txt1 +BM_UFlat/7 959.4MB/s txt2 +BM_UFlat/8 1.0GB/s txt3 +BM_UFlat/9 864.5MB/s txt4 +BM_UFlat/10 2.9GB/s pb +BM_UFlat/11 1.2GB/s gaviota + +BM_ZFlat/0 944.3MB/s html (22.31 %) +BM_ZFlat/1 501.6MB/s urls (47.78 %) +BM_ZFlat/2 14.3GB/s jpg (99.95 %) +BM_ZFlat/3 538.3MB/s jpg_200 (73.00 %) +BM_ZFlat/4 8.3GB/s pdf (83.30 %) +BM_ZFlat/5 903.5MB/s html4 (22.52 %) +BM_ZFlat/6 336.0MB/s txt1 (57.88 %) +BM_ZFlat/7 312.3MB/s txt2 (61.91 %) +BM_ZFlat/8 353.1MB/s txt3 (54.99 %) +BM_ZFlat/9 289.9MB/s txt4 (66.26 %) +BM_ZFlat/10 1.2GB/s pb (19.68 %) +BM_ZFlat/11 527.4MB/s gaviota (37.72 %) diff --git a/vendor/github.com/golang/snappy/decode.go b/vendor/github.com/golang/snappy/decode.go new file mode 100644 index 0000000..23c6e26 --- /dev/null +++ b/vendor/github.com/golang/snappy/decode.go @@ -0,0 +1,264 @@ +// Copyright 2011 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package snappy + +import ( + "encoding/binary" + "errors" + "io" +) + +var ( + // ErrCorrupt reports that the input is invalid. + ErrCorrupt = errors.New("snappy: corrupt input") + // ErrTooLarge reports that the uncompressed length is too large. + ErrTooLarge = errors.New("snappy: decoded block is too large") + // ErrUnsupported reports that the input isn't supported. + ErrUnsupported = errors.New("snappy: unsupported input") + + errUnsupportedLiteralLength = errors.New("snappy: unsupported literal length") +) + +// DecodedLen returns the length of the decoded block. +func DecodedLen(src []byte) (int, error) { + v, _, err := decodedLen(src) + return v, err +} + +// decodedLen returns the length of the decoded block and the number of bytes +// that the length header occupied. +func decodedLen(src []byte) (blockLen, headerLen int, err error) { + v, n := binary.Uvarint(src) + if n <= 0 || v > 0xffffffff { + return 0, 0, ErrCorrupt + } + + const wordSize = 32 << (^uint(0) >> 32 & 1) + if wordSize == 32 && v > 0x7fffffff { + return 0, 0, ErrTooLarge + } + return int(v), n, nil +} + +const ( + decodeErrCodeCorrupt = 1 + decodeErrCodeUnsupportedLiteralLength = 2 +) + +// Decode returns the decoded form of src. The returned slice may be a sub- +// slice of dst if dst was large enough to hold the entire decoded block. +// Otherwise, a newly allocated slice will be returned. +// +// The dst and src must not overlap. It is valid to pass a nil dst. +// +// Decode handles the Snappy block format, not the Snappy stream format. +func Decode(dst, src []byte) ([]byte, error) { + dLen, s, err := decodedLen(src) + if err != nil { + return nil, err + } + if dLen <= len(dst) { + dst = dst[:dLen] + } else { + dst = make([]byte, dLen) + } + switch decode(dst, src[s:]) { + case 0: + return dst, nil + case decodeErrCodeUnsupportedLiteralLength: + return nil, errUnsupportedLiteralLength + } + return nil, ErrCorrupt +} + +// NewReader returns a new Reader that decompresses from r, using the framing +// format described at +// https://github.com/google/snappy/blob/master/framing_format.txt +func NewReader(r io.Reader) *Reader { + return &Reader{ + r: r, + decoded: make([]byte, maxBlockSize), + buf: make([]byte, maxEncodedLenOfMaxBlockSize+checksumSize), + } +} + +// Reader is an io.Reader that can read Snappy-compressed bytes. +// +// Reader handles the Snappy stream format, not the Snappy block format. +type Reader struct { + r io.Reader + err error + decoded []byte + buf []byte + // decoded[i:j] contains decoded bytes that have not yet been passed on. + i, j int + readHeader bool +} + +// Reset discards any buffered data, resets all state, and switches the Snappy +// reader to read from r. This permits reusing a Reader rather than allocating +// a new one. +func (r *Reader) Reset(reader io.Reader) { + r.r = reader + r.err = nil + r.i = 0 + r.j = 0 + r.readHeader = false +} + +func (r *Reader) readFull(p []byte, allowEOF bool) (ok bool) { + if _, r.err = io.ReadFull(r.r, p); r.err != nil { + if r.err == io.ErrUnexpectedEOF || (r.err == io.EOF && !allowEOF) { + r.err = ErrCorrupt + } + return false + } + return true +} + +func (r *Reader) fill() error { + for r.i >= r.j { + if !r.readFull(r.buf[:4], true) { + return r.err + } + chunkType := r.buf[0] + if !r.readHeader { + if chunkType != chunkTypeStreamIdentifier { + r.err = ErrCorrupt + return r.err + } + r.readHeader = true + } + chunkLen := int(r.buf[1]) | int(r.buf[2])<<8 | int(r.buf[3])<<16 + if chunkLen > len(r.buf) { + r.err = ErrUnsupported + return r.err + } + + // The chunk types are specified at + // https://github.com/google/snappy/blob/master/framing_format.txt + switch chunkType { + case chunkTypeCompressedData: + // Section 4.2. Compressed data (chunk type 0x00). + if chunkLen < checksumSize { + r.err = ErrCorrupt + return r.err + } + buf := r.buf[:chunkLen] + if !r.readFull(buf, false) { + return r.err + } + checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24 + buf = buf[checksumSize:] + + n, err := DecodedLen(buf) + if err != nil { + r.err = err + return r.err + } + if n > len(r.decoded) { + r.err = ErrCorrupt + return r.err + } + if _, err := Decode(r.decoded, buf); err != nil { + r.err = err + return r.err + } + if crc(r.decoded[:n]) != checksum { + r.err = ErrCorrupt + return r.err + } + r.i, r.j = 0, n + continue + + case chunkTypeUncompressedData: + // Section 4.3. Uncompressed data (chunk type 0x01). + if chunkLen < checksumSize { + r.err = ErrCorrupt + return r.err + } + buf := r.buf[:checksumSize] + if !r.readFull(buf, false) { + return r.err + } + checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24 + // Read directly into r.decoded instead of via r.buf. + n := chunkLen - checksumSize + if n > len(r.decoded) { + r.err = ErrCorrupt + return r.err + } + if !r.readFull(r.decoded[:n], false) { + return r.err + } + if crc(r.decoded[:n]) != checksum { + r.err = ErrCorrupt + return r.err + } + r.i, r.j = 0, n + continue + + case chunkTypeStreamIdentifier: + // Section 4.1. Stream identifier (chunk type 0xff). + if chunkLen != len(magicBody) { + r.err = ErrCorrupt + return r.err + } + if !r.readFull(r.buf[:len(magicBody)], false) { + return r.err + } + for i := 0; i < len(magicBody); i++ { + if r.buf[i] != magicBody[i] { + r.err = ErrCorrupt + return r.err + } + } + continue + } + + if chunkType <= 0x7f { + // Section 4.5. Reserved unskippable chunks (chunk types 0x02-0x7f). + r.err = ErrUnsupported + return r.err + } + // Section 4.4 Padding (chunk type 0xfe). + // Section 4.6. Reserved skippable chunks (chunk types 0x80-0xfd). + if !r.readFull(r.buf[:chunkLen], false) { + return r.err + } + } + + return nil +} + +// Read satisfies the io.Reader interface. +func (r *Reader) Read(p []byte) (int, error) { + if r.err != nil { + return 0, r.err + } + + if err := r.fill(); err != nil { + return 0, err + } + + n := copy(p, r.decoded[r.i:r.j]) + r.i += n + return n, nil +} + +// ReadByte satisfies the io.ByteReader interface. +func (r *Reader) ReadByte() (byte, error) { + if r.err != nil { + return 0, r.err + } + + if err := r.fill(); err != nil { + return 0, err + } + + c := r.decoded[r.i] + r.i++ + return c, nil +} diff --git a/vendor/github.com/golang/snappy/decode_amd64.s b/vendor/github.com/golang/snappy/decode_amd64.s new file mode 100644 index 0000000..e6179f6 --- /dev/null +++ b/vendor/github.com/golang/snappy/decode_amd64.s @@ -0,0 +1,490 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" + +// The asm code generally follows the pure Go code in decode_other.go, except +// where marked with a "!!!". + +// func decode(dst, src []byte) int +// +// All local variables fit into registers. The non-zero stack size is only to +// spill registers and push args when issuing a CALL. The register allocation: +// - AX scratch +// - BX scratch +// - CX length or x +// - DX offset +// - SI &src[s] +// - DI &dst[d] +// + R8 dst_base +// + R9 dst_len +// + R10 dst_base + dst_len +// + R11 src_base +// + R12 src_len +// + R13 src_base + src_len +// - R14 used by doCopy +// - R15 used by doCopy +// +// The registers R8-R13 (marked with a "+") are set at the start of the +// function, and after a CALL returns, and are not otherwise modified. +// +// The d variable is implicitly DI - R8, and len(dst)-d is R10 - DI. +// The s variable is implicitly SI - R11, and len(src)-s is R13 - SI. +TEXT ·decode(SB), NOSPLIT, $48-56 + // Initialize SI, DI and R8-R13. + MOVQ dst_base+0(FP), R8 + MOVQ dst_len+8(FP), R9 + MOVQ R8, DI + MOVQ R8, R10 + ADDQ R9, R10 + MOVQ src_base+24(FP), R11 + MOVQ src_len+32(FP), R12 + MOVQ R11, SI + MOVQ R11, R13 + ADDQ R12, R13 + +loop: + // for s < len(src) + CMPQ SI, R13 + JEQ end + + // CX = uint32(src[s]) + // + // switch src[s] & 0x03 + MOVBLZX (SI), CX + MOVL CX, BX + ANDL $3, BX + CMPL BX, $1 + JAE tagCopy + + // ---------------------------------------- + // The code below handles literal tags. + + // case tagLiteral: + // x := uint32(src[s] >> 2) + // switch + SHRL $2, CX + CMPL CX, $60 + JAE tagLit60Plus + + // case x < 60: + // s++ + INCQ SI + +doLit: + // This is the end of the inner "switch", when we have a literal tag. + // + // We assume that CX == x and x fits in a uint32, where x is the variable + // used in the pure Go decode_other.go code. + + // length = int(x) + 1 + // + // Unlike the pure Go code, we don't need to check if length <= 0 because + // CX can hold 64 bits, so the increment cannot overflow. + INCQ CX + + // Prepare to check if copying length bytes will run past the end of dst or + // src. + // + // AX = len(dst) - d + // BX = len(src) - s + MOVQ R10, AX + SUBQ DI, AX + MOVQ R13, BX + SUBQ SI, BX + + // !!! Try a faster technique for short (16 or fewer bytes) copies. + // + // if length > 16 || len(dst)-d < 16 || len(src)-s < 16 { + // goto callMemmove // Fall back on calling runtime·memmove. + // } + // + // The C++ snappy code calls this TryFastAppend. It also checks len(src)-s + // against 21 instead of 16, because it cannot assume that all of its input + // is contiguous in memory and so it needs to leave enough source bytes to + // read the next tag without refilling buffers, but Go's Decode assumes + // contiguousness (the src argument is a []byte). + CMPQ CX, $16 + JGT callMemmove + CMPQ AX, $16 + JLT callMemmove + CMPQ BX, $16 + JLT callMemmove + + // !!! Implement the copy from src to dst as a 16-byte load and store. + // (Decode's documentation says that dst and src must not overlap.) + // + // This always copies 16 bytes, instead of only length bytes, but that's + // OK. If the input is a valid Snappy encoding then subsequent iterations + // will fix up the overrun. Otherwise, Decode returns a nil []byte (and a + // non-nil error), so the overrun will be ignored. + // + // Note that on amd64, it is legal and cheap to issue unaligned 8-byte or + // 16-byte loads and stores. This technique probably wouldn't be as + // effective on architectures that are fussier about alignment. + MOVOU 0(SI), X0 + MOVOU X0, 0(DI) + + // d += length + // s += length + ADDQ CX, DI + ADDQ CX, SI + JMP loop + +callMemmove: + // if length > len(dst)-d || length > len(src)-s { etc } + CMPQ CX, AX + JGT errCorrupt + CMPQ CX, BX + JGT errCorrupt + + // copy(dst[d:], src[s:s+length]) + // + // This means calling runtime·memmove(&dst[d], &src[s], length), so we push + // DI, SI and CX as arguments. Coincidentally, we also need to spill those + // three registers to the stack, to save local variables across the CALL. + MOVQ DI, 0(SP) + MOVQ SI, 8(SP) + MOVQ CX, 16(SP) + MOVQ DI, 24(SP) + MOVQ SI, 32(SP) + MOVQ CX, 40(SP) + CALL runtime·memmove(SB) + + // Restore local variables: unspill registers from the stack and + // re-calculate R8-R13. + MOVQ 24(SP), DI + MOVQ 32(SP), SI + MOVQ 40(SP), CX + MOVQ dst_base+0(FP), R8 + MOVQ dst_len+8(FP), R9 + MOVQ R8, R10 + ADDQ R9, R10 + MOVQ src_base+24(FP), R11 + MOVQ src_len+32(FP), R12 + MOVQ R11, R13 + ADDQ R12, R13 + + // d += length + // s += length + ADDQ CX, DI + ADDQ CX, SI + JMP loop + +tagLit60Plus: + // !!! This fragment does the + // + // s += x - 58; if uint(s) > uint(len(src)) { etc } + // + // checks. In the asm version, we code it once instead of once per switch case. + ADDQ CX, SI + SUBQ $58, SI + MOVQ SI, BX + SUBQ R11, BX + CMPQ BX, R12 + JA errCorrupt + + // case x == 60: + CMPL CX, $61 + JEQ tagLit61 + JA tagLit62Plus + + // x = uint32(src[s-1]) + MOVBLZX -1(SI), CX + JMP doLit + +tagLit61: + // case x == 61: + // x = uint32(src[s-2]) | uint32(src[s-1])<<8 + MOVWLZX -2(SI), CX + JMP doLit + +tagLit62Plus: + CMPL CX, $62 + JA tagLit63 + + // case x == 62: + // x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16 + MOVWLZX -3(SI), CX + MOVBLZX -1(SI), BX + SHLL $16, BX + ORL BX, CX + JMP doLit + +tagLit63: + // case x == 63: + // x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24 + MOVL -4(SI), CX + JMP doLit + +// The code above handles literal tags. +// ---------------------------------------- +// The code below handles copy tags. + +tagCopy4: + // case tagCopy4: + // s += 5 + ADDQ $5, SI + + // if uint(s) > uint(len(src)) { etc } + MOVQ SI, BX + SUBQ R11, BX + CMPQ BX, R12 + JA errCorrupt + + // length = 1 + int(src[s-5])>>2 + SHRQ $2, CX + INCQ CX + + // offset = int(uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24) + MOVLQZX -4(SI), DX + JMP doCopy + +tagCopy2: + // case tagCopy2: + // s += 3 + ADDQ $3, SI + + // if uint(s) > uint(len(src)) { etc } + MOVQ SI, BX + SUBQ R11, BX + CMPQ BX, R12 + JA errCorrupt + + // length = 1 + int(src[s-3])>>2 + SHRQ $2, CX + INCQ CX + + // offset = int(uint32(src[s-2]) | uint32(src[s-1])<<8) + MOVWQZX -2(SI), DX + JMP doCopy + +tagCopy: + // We have a copy tag. We assume that: + // - BX == src[s] & 0x03 + // - CX == src[s] + CMPQ BX, $2 + JEQ tagCopy2 + JA tagCopy4 + + // case tagCopy1: + // s += 2 + ADDQ $2, SI + + // if uint(s) > uint(len(src)) { etc } + MOVQ SI, BX + SUBQ R11, BX + CMPQ BX, R12 + JA errCorrupt + + // offset = int(uint32(src[s-2])&0xe0<<3 | uint32(src[s-1])) + MOVQ CX, DX + ANDQ $0xe0, DX + SHLQ $3, DX + MOVBQZX -1(SI), BX + ORQ BX, DX + + // length = 4 + int(src[s-2])>>2&0x7 + SHRQ $2, CX + ANDQ $7, CX + ADDQ $4, CX + +doCopy: + // This is the end of the outer "switch", when we have a copy tag. + // + // We assume that: + // - CX == length && CX > 0 + // - DX == offset + + // if offset <= 0 { etc } + CMPQ DX, $0 + JLE errCorrupt + + // if d < offset { etc } + MOVQ DI, BX + SUBQ R8, BX + CMPQ BX, DX + JLT errCorrupt + + // if length > len(dst)-d { etc } + MOVQ R10, BX + SUBQ DI, BX + CMPQ CX, BX + JGT errCorrupt + + // forwardCopy(dst[d:d+length], dst[d-offset:]); d += length + // + // Set: + // - R14 = len(dst)-d + // - R15 = &dst[d-offset] + MOVQ R10, R14 + SUBQ DI, R14 + MOVQ DI, R15 + SUBQ DX, R15 + + // !!! Try a faster technique for short (16 or fewer bytes) forward copies. + // + // First, try using two 8-byte load/stores, similar to the doLit technique + // above. Even if dst[d:d+length] and dst[d-offset:] can overlap, this is + // still OK if offset >= 8. Note that this has to be two 8-byte load/stores + // and not one 16-byte load/store, and the first store has to be before the + // second load, due to the overlap if offset is in the range [8, 16). + // + // if length > 16 || offset < 8 || len(dst)-d < 16 { + // goto slowForwardCopy + // } + // copy 16 bytes + // d += length + CMPQ CX, $16 + JGT slowForwardCopy + CMPQ DX, $8 + JLT slowForwardCopy + CMPQ R14, $16 + JLT slowForwardCopy + MOVQ 0(R15), AX + MOVQ AX, 0(DI) + MOVQ 8(R15), BX + MOVQ BX, 8(DI) + ADDQ CX, DI + JMP loop + +slowForwardCopy: + // !!! If the forward copy is longer than 16 bytes, or if offset < 8, we + // can still try 8-byte load stores, provided we can overrun up to 10 extra + // bytes. As above, the overrun will be fixed up by subsequent iterations + // of the outermost loop. + // + // The C++ snappy code calls this technique IncrementalCopyFastPath. Its + // commentary says: + // + // ---- + // + // The main part of this loop is a simple copy of eight bytes at a time + // until we've copied (at least) the requested amount of bytes. However, + // if d and d-offset are less than eight bytes apart (indicating a + // repeating pattern of length < 8), we first need to expand the pattern in + // order to get the correct results. For instance, if the buffer looks like + // this, with the eight-byte and patterns marked as + // intervals: + // + // abxxxxxxxxxxxx + // [------] d-offset + // [------] d + // + // a single eight-byte copy from to will repeat the pattern + // once, after which we can move two bytes without moving : + // + // ababxxxxxxxxxx + // [------] d-offset + // [------] d + // + // and repeat the exercise until the two no longer overlap. + // + // This allows us to do very well in the special case of one single byte + // repeated many times, without taking a big hit for more general cases. + // + // The worst case of extra writing past the end of the match occurs when + // offset == 1 and length == 1; the last copy will read from byte positions + // [0..7] and write to [4..11], whereas it was only supposed to write to + // position 1. Thus, ten excess bytes. + // + // ---- + // + // That "10 byte overrun" worst case is confirmed by Go's + // TestSlowForwardCopyOverrun, which also tests the fixUpSlowForwardCopy + // and finishSlowForwardCopy algorithm. + // + // if length > len(dst)-d-10 { + // goto verySlowForwardCopy + // } + SUBQ $10, R14 + CMPQ CX, R14 + JGT verySlowForwardCopy + +makeOffsetAtLeast8: + // !!! As above, expand the pattern so that offset >= 8 and we can use + // 8-byte load/stores. + // + // for offset < 8 { + // copy 8 bytes from dst[d-offset:] to dst[d:] + // length -= offset + // d += offset + // offset += offset + // // The two previous lines together means that d-offset, and therefore + // // R15, is unchanged. + // } + CMPQ DX, $8 + JGE fixUpSlowForwardCopy + MOVQ (R15), BX + MOVQ BX, (DI) + SUBQ DX, CX + ADDQ DX, DI + ADDQ DX, DX + JMP makeOffsetAtLeast8 + +fixUpSlowForwardCopy: + // !!! Add length (which might be negative now) to d (implied by DI being + // &dst[d]) so that d ends up at the right place when we jump back to the + // top of the loop. Before we do that, though, we save DI to AX so that, if + // length is positive, copying the remaining length bytes will write to the + // right place. + MOVQ DI, AX + ADDQ CX, DI + +finishSlowForwardCopy: + // !!! Repeat 8-byte load/stores until length <= 0. Ending with a negative + // length means that we overrun, but as above, that will be fixed up by + // subsequent iterations of the outermost loop. + CMPQ CX, $0 + JLE loop + MOVQ (R15), BX + MOVQ BX, (AX) + ADDQ $8, R15 + ADDQ $8, AX + SUBQ $8, CX + JMP finishSlowForwardCopy + +verySlowForwardCopy: + // verySlowForwardCopy is a simple implementation of forward copy. In C + // parlance, this is a do/while loop instead of a while loop, since we know + // that length > 0. In Go syntax: + // + // for { + // dst[d] = dst[d - offset] + // d++ + // length-- + // if length == 0 { + // break + // } + // } + MOVB (R15), BX + MOVB BX, (DI) + INCQ R15 + INCQ DI + DECQ CX + JNZ verySlowForwardCopy + JMP loop + +// The code above handles copy tags. +// ---------------------------------------- + +end: + // This is the end of the "for s < len(src)". + // + // if d != len(dst) { etc } + CMPQ DI, R10 + JNE errCorrupt + + // return 0 + MOVQ $0, ret+48(FP) + RET + +errCorrupt: + // return decodeErrCodeCorrupt + MOVQ $1, ret+48(FP) + RET diff --git a/vendor/github.com/golang/snappy/decode_arm64.s b/vendor/github.com/golang/snappy/decode_arm64.s new file mode 100644 index 0000000..7a3ead1 --- /dev/null +++ b/vendor/github.com/golang/snappy/decode_arm64.s @@ -0,0 +1,494 @@ +// Copyright 2020 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" + +// The asm code generally follows the pure Go code in decode_other.go, except +// where marked with a "!!!". + +// func decode(dst, src []byte) int +// +// All local variables fit into registers. The non-zero stack size is only to +// spill registers and push args when issuing a CALL. The register allocation: +// - R2 scratch +// - R3 scratch +// - R4 length or x +// - R5 offset +// - R6 &src[s] +// - R7 &dst[d] +// + R8 dst_base +// + R9 dst_len +// + R10 dst_base + dst_len +// + R11 src_base +// + R12 src_len +// + R13 src_base + src_len +// - R14 used by doCopy +// - R15 used by doCopy +// +// The registers R8-R13 (marked with a "+") are set at the start of the +// function, and after a CALL returns, and are not otherwise modified. +// +// The d variable is implicitly R7 - R8, and len(dst)-d is R10 - R7. +// The s variable is implicitly R6 - R11, and len(src)-s is R13 - R6. +TEXT ·decode(SB), NOSPLIT, $56-56 + // Initialize R6, R7 and R8-R13. + MOVD dst_base+0(FP), R8 + MOVD dst_len+8(FP), R9 + MOVD R8, R7 + MOVD R8, R10 + ADD R9, R10, R10 + MOVD src_base+24(FP), R11 + MOVD src_len+32(FP), R12 + MOVD R11, R6 + MOVD R11, R13 + ADD R12, R13, R13 + +loop: + // for s < len(src) + CMP R13, R6 + BEQ end + + // R4 = uint32(src[s]) + // + // switch src[s] & 0x03 + MOVBU (R6), R4 + MOVW R4, R3 + ANDW $3, R3 + MOVW $1, R1 + CMPW R1, R3 + BGE tagCopy + + // ---------------------------------------- + // The code below handles literal tags. + + // case tagLiteral: + // x := uint32(src[s] >> 2) + // switch + MOVW $60, R1 + LSRW $2, R4, R4 + CMPW R4, R1 + BLS tagLit60Plus + + // case x < 60: + // s++ + ADD $1, R6, R6 + +doLit: + // This is the end of the inner "switch", when we have a literal tag. + // + // We assume that R4 == x and x fits in a uint32, where x is the variable + // used in the pure Go decode_other.go code. + + // length = int(x) + 1 + // + // Unlike the pure Go code, we don't need to check if length <= 0 because + // R4 can hold 64 bits, so the increment cannot overflow. + ADD $1, R4, R4 + + // Prepare to check if copying length bytes will run past the end of dst or + // src. + // + // R2 = len(dst) - d + // R3 = len(src) - s + MOVD R10, R2 + SUB R7, R2, R2 + MOVD R13, R3 + SUB R6, R3, R3 + + // !!! Try a faster technique for short (16 or fewer bytes) copies. + // + // if length > 16 || len(dst)-d < 16 || len(src)-s < 16 { + // goto callMemmove // Fall back on calling runtime·memmove. + // } + // + // The C++ snappy code calls this TryFastAppend. It also checks len(src)-s + // against 21 instead of 16, because it cannot assume that all of its input + // is contiguous in memory and so it needs to leave enough source bytes to + // read the next tag without refilling buffers, but Go's Decode assumes + // contiguousness (the src argument is a []byte). + CMP $16, R4 + BGT callMemmove + CMP $16, R2 + BLT callMemmove + CMP $16, R3 + BLT callMemmove + + // !!! Implement the copy from src to dst as a 16-byte load and store. + // (Decode's documentation says that dst and src must not overlap.) + // + // This always copies 16 bytes, instead of only length bytes, but that's + // OK. If the input is a valid Snappy encoding then subsequent iterations + // will fix up the overrun. Otherwise, Decode returns a nil []byte (and a + // non-nil error), so the overrun will be ignored. + // + // Note that on arm64, it is legal and cheap to issue unaligned 8-byte or + // 16-byte loads and stores. This technique probably wouldn't be as + // effective on architectures that are fussier about alignment. + LDP 0(R6), (R14, R15) + STP (R14, R15), 0(R7) + + // d += length + // s += length + ADD R4, R7, R7 + ADD R4, R6, R6 + B loop + +callMemmove: + // if length > len(dst)-d || length > len(src)-s { etc } + CMP R2, R4 + BGT errCorrupt + CMP R3, R4 + BGT errCorrupt + + // copy(dst[d:], src[s:s+length]) + // + // This means calling runtime·memmove(&dst[d], &src[s], length), so we push + // R7, R6 and R4 as arguments. Coincidentally, we also need to spill those + // three registers to the stack, to save local variables across the CALL. + MOVD R7, 8(RSP) + MOVD R6, 16(RSP) + MOVD R4, 24(RSP) + MOVD R7, 32(RSP) + MOVD R6, 40(RSP) + MOVD R4, 48(RSP) + CALL runtime·memmove(SB) + + // Restore local variables: unspill registers from the stack and + // re-calculate R8-R13. + MOVD 32(RSP), R7 + MOVD 40(RSP), R6 + MOVD 48(RSP), R4 + MOVD dst_base+0(FP), R8 + MOVD dst_len+8(FP), R9 + MOVD R8, R10 + ADD R9, R10, R10 + MOVD src_base+24(FP), R11 + MOVD src_len+32(FP), R12 + MOVD R11, R13 + ADD R12, R13, R13 + + // d += length + // s += length + ADD R4, R7, R7 + ADD R4, R6, R6 + B loop + +tagLit60Plus: + // !!! This fragment does the + // + // s += x - 58; if uint(s) > uint(len(src)) { etc } + // + // checks. In the asm version, we code it once instead of once per switch case. + ADD R4, R6, R6 + SUB $58, R6, R6 + MOVD R6, R3 + SUB R11, R3, R3 + CMP R12, R3 + BGT errCorrupt + + // case x == 60: + MOVW $61, R1 + CMPW R1, R4 + BEQ tagLit61 + BGT tagLit62Plus + + // x = uint32(src[s-1]) + MOVBU -1(R6), R4 + B doLit + +tagLit61: + // case x == 61: + // x = uint32(src[s-2]) | uint32(src[s-1])<<8 + MOVHU -2(R6), R4 + B doLit + +tagLit62Plus: + CMPW $62, R4 + BHI tagLit63 + + // case x == 62: + // x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16 + MOVHU -3(R6), R4 + MOVBU -1(R6), R3 + ORR R3<<16, R4 + B doLit + +tagLit63: + // case x == 63: + // x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24 + MOVWU -4(R6), R4 + B doLit + + // The code above handles literal tags. + // ---------------------------------------- + // The code below handles copy tags. + +tagCopy4: + // case tagCopy4: + // s += 5 + ADD $5, R6, R6 + + // if uint(s) > uint(len(src)) { etc } + MOVD R6, R3 + SUB R11, R3, R3 + CMP R12, R3 + BGT errCorrupt + + // length = 1 + int(src[s-5])>>2 + MOVD $1, R1 + ADD R4>>2, R1, R4 + + // offset = int(uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24) + MOVWU -4(R6), R5 + B doCopy + +tagCopy2: + // case tagCopy2: + // s += 3 + ADD $3, R6, R6 + + // if uint(s) > uint(len(src)) { etc } + MOVD R6, R3 + SUB R11, R3, R3 + CMP R12, R3 + BGT errCorrupt + + // length = 1 + int(src[s-3])>>2 + MOVD $1, R1 + ADD R4>>2, R1, R4 + + // offset = int(uint32(src[s-2]) | uint32(src[s-1])<<8) + MOVHU -2(R6), R5 + B doCopy + +tagCopy: + // We have a copy tag. We assume that: + // - R3 == src[s] & 0x03 + // - R4 == src[s] + CMP $2, R3 + BEQ tagCopy2 + BGT tagCopy4 + + // case tagCopy1: + // s += 2 + ADD $2, R6, R6 + + // if uint(s) > uint(len(src)) { etc } + MOVD R6, R3 + SUB R11, R3, R3 + CMP R12, R3 + BGT errCorrupt + + // offset = int(uint32(src[s-2])&0xe0<<3 | uint32(src[s-1])) + MOVD R4, R5 + AND $0xe0, R5 + MOVBU -1(R6), R3 + ORR R5<<3, R3, R5 + + // length = 4 + int(src[s-2])>>2&0x7 + MOVD $7, R1 + AND R4>>2, R1, R4 + ADD $4, R4, R4 + +doCopy: + // This is the end of the outer "switch", when we have a copy tag. + // + // We assume that: + // - R4 == length && R4 > 0 + // - R5 == offset + + // if offset <= 0 { etc } + MOVD $0, R1 + CMP R1, R5 + BLE errCorrupt + + // if d < offset { etc } + MOVD R7, R3 + SUB R8, R3, R3 + CMP R5, R3 + BLT errCorrupt + + // if length > len(dst)-d { etc } + MOVD R10, R3 + SUB R7, R3, R3 + CMP R3, R4 + BGT errCorrupt + + // forwardCopy(dst[d:d+length], dst[d-offset:]); d += length + // + // Set: + // - R14 = len(dst)-d + // - R15 = &dst[d-offset] + MOVD R10, R14 + SUB R7, R14, R14 + MOVD R7, R15 + SUB R5, R15, R15 + + // !!! Try a faster technique for short (16 or fewer bytes) forward copies. + // + // First, try using two 8-byte load/stores, similar to the doLit technique + // above. Even if dst[d:d+length] and dst[d-offset:] can overlap, this is + // still OK if offset >= 8. Note that this has to be two 8-byte load/stores + // and not one 16-byte load/store, and the first store has to be before the + // second load, due to the overlap if offset is in the range [8, 16). + // + // if length > 16 || offset < 8 || len(dst)-d < 16 { + // goto slowForwardCopy + // } + // copy 16 bytes + // d += length + CMP $16, R4 + BGT slowForwardCopy + CMP $8, R5 + BLT slowForwardCopy + CMP $16, R14 + BLT slowForwardCopy + MOVD 0(R15), R2 + MOVD R2, 0(R7) + MOVD 8(R15), R3 + MOVD R3, 8(R7) + ADD R4, R7, R7 + B loop + +slowForwardCopy: + // !!! If the forward copy is longer than 16 bytes, or if offset < 8, we + // can still try 8-byte load stores, provided we can overrun up to 10 extra + // bytes. As above, the overrun will be fixed up by subsequent iterations + // of the outermost loop. + // + // The C++ snappy code calls this technique IncrementalCopyFastPath. Its + // commentary says: + // + // ---- + // + // The main part of this loop is a simple copy of eight bytes at a time + // until we've copied (at least) the requested amount of bytes. However, + // if d and d-offset are less than eight bytes apart (indicating a + // repeating pattern of length < 8), we first need to expand the pattern in + // order to get the correct results. For instance, if the buffer looks like + // this, with the eight-byte and patterns marked as + // intervals: + // + // abxxxxxxxxxxxx + // [------] d-offset + // [------] d + // + // a single eight-byte copy from to will repeat the pattern + // once, after which we can move two bytes without moving : + // + // ababxxxxxxxxxx + // [------] d-offset + // [------] d + // + // and repeat the exercise until the two no longer overlap. + // + // This allows us to do very well in the special case of one single byte + // repeated many times, without taking a big hit for more general cases. + // + // The worst case of extra writing past the end of the match occurs when + // offset == 1 and length == 1; the last copy will read from byte positions + // [0..7] and write to [4..11], whereas it was only supposed to write to + // position 1. Thus, ten excess bytes. + // + // ---- + // + // That "10 byte overrun" worst case is confirmed by Go's + // TestSlowForwardCopyOverrun, which also tests the fixUpSlowForwardCopy + // and finishSlowForwardCopy algorithm. + // + // if length > len(dst)-d-10 { + // goto verySlowForwardCopy + // } + SUB $10, R14, R14 + CMP R14, R4 + BGT verySlowForwardCopy + +makeOffsetAtLeast8: + // !!! As above, expand the pattern so that offset >= 8 and we can use + // 8-byte load/stores. + // + // for offset < 8 { + // copy 8 bytes from dst[d-offset:] to dst[d:] + // length -= offset + // d += offset + // offset += offset + // // The two previous lines together means that d-offset, and therefore + // // R15, is unchanged. + // } + CMP $8, R5 + BGE fixUpSlowForwardCopy + MOVD (R15), R3 + MOVD R3, (R7) + SUB R5, R4, R4 + ADD R5, R7, R7 + ADD R5, R5, R5 + B makeOffsetAtLeast8 + +fixUpSlowForwardCopy: + // !!! Add length (which might be negative now) to d (implied by R7 being + // &dst[d]) so that d ends up at the right place when we jump back to the + // top of the loop. Before we do that, though, we save R7 to R2 so that, if + // length is positive, copying the remaining length bytes will write to the + // right place. + MOVD R7, R2 + ADD R4, R7, R7 + +finishSlowForwardCopy: + // !!! Repeat 8-byte load/stores until length <= 0. Ending with a negative + // length means that we overrun, but as above, that will be fixed up by + // subsequent iterations of the outermost loop. + MOVD $0, R1 + CMP R1, R4 + BLE loop + MOVD (R15), R3 + MOVD R3, (R2) + ADD $8, R15, R15 + ADD $8, R2, R2 + SUB $8, R4, R4 + B finishSlowForwardCopy + +verySlowForwardCopy: + // verySlowForwardCopy is a simple implementation of forward copy. In C + // parlance, this is a do/while loop instead of a while loop, since we know + // that length > 0. In Go syntax: + // + // for { + // dst[d] = dst[d - offset] + // d++ + // length-- + // if length == 0 { + // break + // } + // } + MOVB (R15), R3 + MOVB R3, (R7) + ADD $1, R15, R15 + ADD $1, R7, R7 + SUB $1, R4, R4 + CBNZ R4, verySlowForwardCopy + B loop + + // The code above handles copy tags. + // ---------------------------------------- + +end: + // This is the end of the "for s < len(src)". + // + // if d != len(dst) { etc } + CMP R10, R7 + BNE errCorrupt + + // return 0 + MOVD $0, ret+48(FP) + RET + +errCorrupt: + // return decodeErrCodeCorrupt + MOVD $1, R2 + MOVD R2, ret+48(FP) + RET diff --git a/vendor/github.com/golang/snappy/decode_asm.go b/vendor/github.com/golang/snappy/decode_asm.go new file mode 100644 index 0000000..7082b34 --- /dev/null +++ b/vendor/github.com/golang/snappy/decode_asm.go @@ -0,0 +1,15 @@ +// Copyright 2016 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine +// +build gc +// +build !noasm +// +build amd64 arm64 + +package snappy + +// decode has the same semantics as in decode_other.go. +// +//go:noescape +func decode(dst, src []byte) int diff --git a/vendor/github.com/golang/snappy/decode_other.go b/vendor/github.com/golang/snappy/decode_other.go new file mode 100644 index 0000000..2f672be --- /dev/null +++ b/vendor/github.com/golang/snappy/decode_other.go @@ -0,0 +1,115 @@ +// Copyright 2016 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64,!arm64 appengine !gc noasm + +package snappy + +// decode writes the decoding of src to dst. It assumes that the varint-encoded +// length of the decompressed bytes has already been read, and that len(dst) +// equals that length. +// +// It returns 0 on success or a decodeErrCodeXxx error code on failure. +func decode(dst, src []byte) int { + var d, s, offset, length int + for s < len(src) { + switch src[s] & 0x03 { + case tagLiteral: + x := uint32(src[s] >> 2) + switch { + case x < 60: + s++ + case x == 60: + s += 2 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-1]) + case x == 61: + s += 3 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-2]) | uint32(src[s-1])<<8 + case x == 62: + s += 4 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16 + case x == 63: + s += 5 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24 + } + length = int(x) + 1 + if length <= 0 { + return decodeErrCodeUnsupportedLiteralLength + } + if length > len(dst)-d || length > len(src)-s { + return decodeErrCodeCorrupt + } + copy(dst[d:], src[s:s+length]) + d += length + s += length + continue + + case tagCopy1: + s += 2 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + length = 4 + int(src[s-2])>>2&0x7 + offset = int(uint32(src[s-2])&0xe0<<3 | uint32(src[s-1])) + + case tagCopy2: + s += 3 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + length = 1 + int(src[s-3])>>2 + offset = int(uint32(src[s-2]) | uint32(src[s-1])<<8) + + case tagCopy4: + s += 5 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + length = 1 + int(src[s-5])>>2 + offset = int(uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24) + } + + if offset <= 0 || d < offset || length > len(dst)-d { + return decodeErrCodeCorrupt + } + // Copy from an earlier sub-slice of dst to a later sub-slice. + // If no overlap, use the built-in copy: + if offset >= length { + copy(dst[d:d+length], dst[d-offset:]) + d += length + continue + } + + // Unlike the built-in copy function, this byte-by-byte copy always runs + // forwards, even if the slices overlap. Conceptually, this is: + // + // d += forwardCopy(dst[d:d+length], dst[d-offset:]) + // + // We align the slices into a and b and show the compiler they are the same size. + // This allows the loop to run without bounds checks. + a := dst[d : d+length] + b := dst[d-offset:] + b = b[:len(a)] + for i := range a { + a[i] = b[i] + } + d += length + } + if d != len(dst) { + return decodeErrCodeCorrupt + } + return 0 +} diff --git a/vendor/github.com/golang/snappy/encode.go b/vendor/github.com/golang/snappy/encode.go new file mode 100644 index 0000000..7f23657 --- /dev/null +++ b/vendor/github.com/golang/snappy/encode.go @@ -0,0 +1,289 @@ +// Copyright 2011 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package snappy + +import ( + "encoding/binary" + "errors" + "io" +) + +// Encode returns the encoded form of src. The returned slice may be a sub- +// slice of dst if dst was large enough to hold the entire encoded block. +// Otherwise, a newly allocated slice will be returned. +// +// The dst and src must not overlap. It is valid to pass a nil dst. +// +// Encode handles the Snappy block format, not the Snappy stream format. +func Encode(dst, src []byte) []byte { + if n := MaxEncodedLen(len(src)); n < 0 { + panic(ErrTooLarge) + } else if len(dst) < n { + dst = make([]byte, n) + } + + // The block starts with the varint-encoded length of the decompressed bytes. + d := binary.PutUvarint(dst, uint64(len(src))) + + for len(src) > 0 { + p := src + src = nil + if len(p) > maxBlockSize { + p, src = p[:maxBlockSize], p[maxBlockSize:] + } + if len(p) < minNonLiteralBlockSize { + d += emitLiteral(dst[d:], p) + } else { + d += encodeBlock(dst[d:], p) + } + } + return dst[:d] +} + +// inputMargin is the minimum number of extra input bytes to keep, inside +// encodeBlock's inner loop. On some architectures, this margin lets us +// implement a fast path for emitLiteral, where the copy of short (<= 16 byte) +// literals can be implemented as a single load to and store from a 16-byte +// register. That literal's actual length can be as short as 1 byte, so this +// can copy up to 15 bytes too much, but that's OK as subsequent iterations of +// the encoding loop will fix up the copy overrun, and this inputMargin ensures +// that we don't overrun the dst and src buffers. +const inputMargin = 16 - 1 + +// minNonLiteralBlockSize is the minimum size of the input to encodeBlock that +// could be encoded with a copy tag. This is the minimum with respect to the +// algorithm used by encodeBlock, not a minimum enforced by the file format. +// +// The encoded output must start with at least a 1 byte literal, as there are +// no previous bytes to copy. A minimal (1 byte) copy after that, generated +// from an emitCopy call in encodeBlock's main loop, would require at least +// another inputMargin bytes, for the reason above: we want any emitLiteral +// calls inside encodeBlock's main loop to use the fast path if possible, which +// requires being able to overrun by inputMargin bytes. Thus, +// minNonLiteralBlockSize equals 1 + 1 + inputMargin. +// +// The C++ code doesn't use this exact threshold, but it could, as discussed at +// https://groups.google.com/d/topic/snappy-compression/oGbhsdIJSJ8/discussion +// The difference between Go (2+inputMargin) and C++ (inputMargin) is purely an +// optimization. It should not affect the encoded form. This is tested by +// TestSameEncodingAsCppShortCopies. +const minNonLiteralBlockSize = 1 + 1 + inputMargin + +// MaxEncodedLen returns the maximum length of a snappy block, given its +// uncompressed length. +// +// It will return a negative value if srcLen is too large to encode. +func MaxEncodedLen(srcLen int) int { + n := uint64(srcLen) + if n > 0xffffffff { + return -1 + } + // Compressed data can be defined as: + // compressed := item* literal* + // item := literal* copy + // + // The trailing literal sequence has a space blowup of at most 62/60 + // since a literal of length 60 needs one tag byte + one extra byte + // for length information. + // + // Item blowup is trickier to measure. Suppose the "copy" op copies + // 4 bytes of data. Because of a special check in the encoding code, + // we produce a 4-byte copy only if the offset is < 65536. Therefore + // the copy op takes 3 bytes to encode, and this type of item leads + // to at most the 62/60 blowup for representing literals. + // + // Suppose the "copy" op copies 5 bytes of data. If the offset is big + // enough, it will take 5 bytes to encode the copy op. Therefore the + // worst case here is a one-byte literal followed by a five-byte copy. + // That is, 6 bytes of input turn into 7 bytes of "compressed" data. + // + // This last factor dominates the blowup, so the final estimate is: + n = 32 + n + n/6 + if n > 0xffffffff { + return -1 + } + return int(n) +} + +var errClosed = errors.New("snappy: Writer is closed") + +// NewWriter returns a new Writer that compresses to w. +// +// The Writer returned does not buffer writes. There is no need to Flush or +// Close such a Writer. +// +// Deprecated: the Writer returned is not suitable for many small writes, only +// for few large writes. Use NewBufferedWriter instead, which is efficient +// regardless of the frequency and shape of the writes, and remember to Close +// that Writer when done. +func NewWriter(w io.Writer) *Writer { + return &Writer{ + w: w, + obuf: make([]byte, obufLen), + } +} + +// NewBufferedWriter returns a new Writer that compresses to w, using the +// framing format described at +// https://github.com/google/snappy/blob/master/framing_format.txt +// +// The Writer returned buffers writes. Users must call Close to guarantee all +// data has been forwarded to the underlying io.Writer. They may also call +// Flush zero or more times before calling Close. +func NewBufferedWriter(w io.Writer) *Writer { + return &Writer{ + w: w, + ibuf: make([]byte, 0, maxBlockSize), + obuf: make([]byte, obufLen), + } +} + +// Writer is an io.Writer that can write Snappy-compressed bytes. +// +// Writer handles the Snappy stream format, not the Snappy block format. +type Writer struct { + w io.Writer + err error + + // ibuf is a buffer for the incoming (uncompressed) bytes. + // + // Its use is optional. For backwards compatibility, Writers created by the + // NewWriter function have ibuf == nil, do not buffer incoming bytes, and + // therefore do not need to be Flush'ed or Close'd. + ibuf []byte + + // obuf is a buffer for the outgoing (compressed) bytes. + obuf []byte + + // wroteStreamHeader is whether we have written the stream header. + wroteStreamHeader bool +} + +// Reset discards the writer's state and switches the Snappy writer to write to +// w. This permits reusing a Writer rather than allocating a new one. +func (w *Writer) Reset(writer io.Writer) { + w.w = writer + w.err = nil + if w.ibuf != nil { + w.ibuf = w.ibuf[:0] + } + w.wroteStreamHeader = false +} + +// Write satisfies the io.Writer interface. +func (w *Writer) Write(p []byte) (nRet int, errRet error) { + if w.ibuf == nil { + // Do not buffer incoming bytes. This does not perform or compress well + // if the caller of Writer.Write writes many small slices. This + // behavior is therefore deprecated, but still supported for backwards + // compatibility with code that doesn't explicitly Flush or Close. + return w.write(p) + } + + // The remainder of this method is based on bufio.Writer.Write from the + // standard library. + + for len(p) > (cap(w.ibuf)-len(w.ibuf)) && w.err == nil { + var n int + if len(w.ibuf) == 0 { + // Large write, empty buffer. + // Write directly from p to avoid copy. + n, _ = w.write(p) + } else { + n = copy(w.ibuf[len(w.ibuf):cap(w.ibuf)], p) + w.ibuf = w.ibuf[:len(w.ibuf)+n] + w.Flush() + } + nRet += n + p = p[n:] + } + if w.err != nil { + return nRet, w.err + } + n := copy(w.ibuf[len(w.ibuf):cap(w.ibuf)], p) + w.ibuf = w.ibuf[:len(w.ibuf)+n] + nRet += n + return nRet, nil +} + +func (w *Writer) write(p []byte) (nRet int, errRet error) { + if w.err != nil { + return 0, w.err + } + for len(p) > 0 { + obufStart := len(magicChunk) + if !w.wroteStreamHeader { + w.wroteStreamHeader = true + copy(w.obuf, magicChunk) + obufStart = 0 + } + + var uncompressed []byte + if len(p) > maxBlockSize { + uncompressed, p = p[:maxBlockSize], p[maxBlockSize:] + } else { + uncompressed, p = p, nil + } + checksum := crc(uncompressed) + + // Compress the buffer, discarding the result if the improvement + // isn't at least 12.5%. + compressed := Encode(w.obuf[obufHeaderLen:], uncompressed) + chunkType := uint8(chunkTypeCompressedData) + chunkLen := 4 + len(compressed) + obufEnd := obufHeaderLen + len(compressed) + if len(compressed) >= len(uncompressed)-len(uncompressed)/8 { + chunkType = chunkTypeUncompressedData + chunkLen = 4 + len(uncompressed) + obufEnd = obufHeaderLen + } + + // Fill in the per-chunk header that comes before the body. + w.obuf[len(magicChunk)+0] = chunkType + w.obuf[len(magicChunk)+1] = uint8(chunkLen >> 0) + w.obuf[len(magicChunk)+2] = uint8(chunkLen >> 8) + w.obuf[len(magicChunk)+3] = uint8(chunkLen >> 16) + w.obuf[len(magicChunk)+4] = uint8(checksum >> 0) + w.obuf[len(magicChunk)+5] = uint8(checksum >> 8) + w.obuf[len(magicChunk)+6] = uint8(checksum >> 16) + w.obuf[len(magicChunk)+7] = uint8(checksum >> 24) + + if _, err := w.w.Write(w.obuf[obufStart:obufEnd]); err != nil { + w.err = err + return nRet, err + } + if chunkType == chunkTypeUncompressedData { + if _, err := w.w.Write(uncompressed); err != nil { + w.err = err + return nRet, err + } + } + nRet += len(uncompressed) + } + return nRet, nil +} + +// Flush flushes the Writer to its underlying io.Writer. +func (w *Writer) Flush() error { + if w.err != nil { + return w.err + } + if len(w.ibuf) == 0 { + return nil + } + w.write(w.ibuf) + w.ibuf = w.ibuf[:0] + return w.err +} + +// Close calls Flush and then closes the Writer. +func (w *Writer) Close() error { + w.Flush() + ret := w.err + if w.err == nil { + w.err = errClosed + } + return ret +} diff --git a/vendor/github.com/golang/snappy/encode_amd64.s b/vendor/github.com/golang/snappy/encode_amd64.s new file mode 100644 index 0000000..adfd979 --- /dev/null +++ b/vendor/github.com/golang/snappy/encode_amd64.s @@ -0,0 +1,730 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" + +// The XXX lines assemble on Go 1.4, 1.5 and 1.7, but not 1.6, due to a +// Go toolchain regression. See https://github.com/golang/go/issues/15426 and +// https://github.com/golang/snappy/issues/29 +// +// As a workaround, the package was built with a known good assembler, and +// those instructions were disassembled by "objdump -d" to yield the +// 4e 0f b7 7c 5c 78 movzwq 0x78(%rsp,%r11,2),%r15 +// style comments, in AT&T asm syntax. Note that rsp here is a physical +// register, not Go/asm's SP pseudo-register (see https://golang.org/doc/asm). +// The instructions were then encoded as "BYTE $0x.." sequences, which assemble +// fine on Go 1.6. + +// The asm code generally follows the pure Go code in encode_other.go, except +// where marked with a "!!!". + +// ---------------------------------------------------------------------------- + +// func emitLiteral(dst, lit []byte) int +// +// All local variables fit into registers. The register allocation: +// - AX len(lit) +// - BX n +// - DX return value +// - DI &dst[i] +// - R10 &lit[0] +// +// The 24 bytes of stack space is to call runtime·memmove. +// +// The unusual register allocation of local variables, such as R10 for the +// source pointer, matches the allocation used at the call site in encodeBlock, +// which makes it easier to manually inline this function. +TEXT ·emitLiteral(SB), NOSPLIT, $24-56 + MOVQ dst_base+0(FP), DI + MOVQ lit_base+24(FP), R10 + MOVQ lit_len+32(FP), AX + MOVQ AX, DX + MOVL AX, BX + SUBL $1, BX + + CMPL BX, $60 + JLT oneByte + CMPL BX, $256 + JLT twoBytes + +threeBytes: + MOVB $0xf4, 0(DI) + MOVW BX, 1(DI) + ADDQ $3, DI + ADDQ $3, DX + JMP memmove + +twoBytes: + MOVB $0xf0, 0(DI) + MOVB BX, 1(DI) + ADDQ $2, DI + ADDQ $2, DX + JMP memmove + +oneByte: + SHLB $2, BX + MOVB BX, 0(DI) + ADDQ $1, DI + ADDQ $1, DX + +memmove: + MOVQ DX, ret+48(FP) + + // copy(dst[i:], lit) + // + // This means calling runtime·memmove(&dst[i], &lit[0], len(lit)), so we push + // DI, R10 and AX as arguments. + MOVQ DI, 0(SP) + MOVQ R10, 8(SP) + MOVQ AX, 16(SP) + CALL runtime·memmove(SB) + RET + +// ---------------------------------------------------------------------------- + +// func emitCopy(dst []byte, offset, length int) int +// +// All local variables fit into registers. The register allocation: +// - AX length +// - SI &dst[0] +// - DI &dst[i] +// - R11 offset +// +// The unusual register allocation of local variables, such as R11 for the +// offset, matches the allocation used at the call site in encodeBlock, which +// makes it easier to manually inline this function. +TEXT ·emitCopy(SB), NOSPLIT, $0-48 + MOVQ dst_base+0(FP), DI + MOVQ DI, SI + MOVQ offset+24(FP), R11 + MOVQ length+32(FP), AX + +loop0: + // for length >= 68 { etc } + CMPL AX, $68 + JLT step1 + + // Emit a length 64 copy, encoded as 3 bytes. + MOVB $0xfe, 0(DI) + MOVW R11, 1(DI) + ADDQ $3, DI + SUBL $64, AX + JMP loop0 + +step1: + // if length > 64 { etc } + CMPL AX, $64 + JLE step2 + + // Emit a length 60 copy, encoded as 3 bytes. + MOVB $0xee, 0(DI) + MOVW R11, 1(DI) + ADDQ $3, DI + SUBL $60, AX + +step2: + // if length >= 12 || offset >= 2048 { goto step3 } + CMPL AX, $12 + JGE step3 + CMPL R11, $2048 + JGE step3 + + // Emit the remaining copy, encoded as 2 bytes. + MOVB R11, 1(DI) + SHRL $8, R11 + SHLB $5, R11 + SUBB $4, AX + SHLB $2, AX + ORB AX, R11 + ORB $1, R11 + MOVB R11, 0(DI) + ADDQ $2, DI + + // Return the number of bytes written. + SUBQ SI, DI + MOVQ DI, ret+40(FP) + RET + +step3: + // Emit the remaining copy, encoded as 3 bytes. + SUBL $1, AX + SHLB $2, AX + ORB $2, AX + MOVB AX, 0(DI) + MOVW R11, 1(DI) + ADDQ $3, DI + + // Return the number of bytes written. + SUBQ SI, DI + MOVQ DI, ret+40(FP) + RET + +// ---------------------------------------------------------------------------- + +// func extendMatch(src []byte, i, j int) int +// +// All local variables fit into registers. The register allocation: +// - DX &src[0] +// - SI &src[j] +// - R13 &src[len(src) - 8] +// - R14 &src[len(src)] +// - R15 &src[i] +// +// The unusual register allocation of local variables, such as R15 for a source +// pointer, matches the allocation used at the call site in encodeBlock, which +// makes it easier to manually inline this function. +TEXT ·extendMatch(SB), NOSPLIT, $0-48 + MOVQ src_base+0(FP), DX + MOVQ src_len+8(FP), R14 + MOVQ i+24(FP), R15 + MOVQ j+32(FP), SI + ADDQ DX, R14 + ADDQ DX, R15 + ADDQ DX, SI + MOVQ R14, R13 + SUBQ $8, R13 + +cmp8: + // As long as we are 8 or more bytes before the end of src, we can load and + // compare 8 bytes at a time. If those 8 bytes are equal, repeat. + CMPQ SI, R13 + JA cmp1 + MOVQ (R15), AX + MOVQ (SI), BX + CMPQ AX, BX + JNE bsf + ADDQ $8, R15 + ADDQ $8, SI + JMP cmp8 + +bsf: + // If those 8 bytes were not equal, XOR the two 8 byte values, and return + // the index of the first byte that differs. The BSF instruction finds the + // least significant 1 bit, the amd64 architecture is little-endian, and + // the shift by 3 converts a bit index to a byte index. + XORQ AX, BX + BSFQ BX, BX + SHRQ $3, BX + ADDQ BX, SI + + // Convert from &src[ret] to ret. + SUBQ DX, SI + MOVQ SI, ret+40(FP) + RET + +cmp1: + // In src's tail, compare 1 byte at a time. + CMPQ SI, R14 + JAE extendMatchEnd + MOVB (R15), AX + MOVB (SI), BX + CMPB AX, BX + JNE extendMatchEnd + ADDQ $1, R15 + ADDQ $1, SI + JMP cmp1 + +extendMatchEnd: + // Convert from &src[ret] to ret. + SUBQ DX, SI + MOVQ SI, ret+40(FP) + RET + +// ---------------------------------------------------------------------------- + +// func encodeBlock(dst, src []byte) (d int) +// +// All local variables fit into registers, other than "var table". The register +// allocation: +// - AX . . +// - BX . . +// - CX 56 shift (note that amd64 shifts by non-immediates must use CX). +// - DX 64 &src[0], tableSize +// - SI 72 &src[s] +// - DI 80 &dst[d] +// - R9 88 sLimit +// - R10 . &src[nextEmit] +// - R11 96 prevHash, currHash, nextHash, offset +// - R12 104 &src[base], skip +// - R13 . &src[nextS], &src[len(src) - 8] +// - R14 . len(src), bytesBetweenHashLookups, &src[len(src)], x +// - R15 112 candidate +// +// The second column (56, 64, etc) is the stack offset to spill the registers +// when calling other functions. We could pack this slightly tighter, but it's +// simpler to have a dedicated spill map independent of the function called. +// +// "var table [maxTableSize]uint16" takes up 32768 bytes of stack space. An +// extra 56 bytes, to call other functions, and an extra 64 bytes, to spill +// local variables (registers) during calls gives 32768 + 56 + 64 = 32888. +TEXT ·encodeBlock(SB), 0, $32888-56 + MOVQ dst_base+0(FP), DI + MOVQ src_base+24(FP), SI + MOVQ src_len+32(FP), R14 + + // shift, tableSize := uint32(32-8), 1<<8 + MOVQ $24, CX + MOVQ $256, DX + +calcShift: + // for ; tableSize < maxTableSize && tableSize < len(src); tableSize *= 2 { + // shift-- + // } + CMPQ DX, $16384 + JGE varTable + CMPQ DX, R14 + JGE varTable + SUBQ $1, CX + SHLQ $1, DX + JMP calcShift + +varTable: + // var table [maxTableSize]uint16 + // + // In the asm code, unlike the Go code, we can zero-initialize only the + // first tableSize elements. Each uint16 element is 2 bytes and each MOVOU + // writes 16 bytes, so we can do only tableSize/8 writes instead of the + // 2048 writes that would zero-initialize all of table's 32768 bytes. + SHRQ $3, DX + LEAQ table-32768(SP), BX + PXOR X0, X0 + +memclr: + MOVOU X0, 0(BX) + ADDQ $16, BX + SUBQ $1, DX + JNZ memclr + + // !!! DX = &src[0] + MOVQ SI, DX + + // sLimit := len(src) - inputMargin + MOVQ R14, R9 + SUBQ $15, R9 + + // !!! Pre-emptively spill CX, DX and R9 to the stack. Their values don't + // change for the rest of the function. + MOVQ CX, 56(SP) + MOVQ DX, 64(SP) + MOVQ R9, 88(SP) + + // nextEmit := 0 + MOVQ DX, R10 + + // s := 1 + ADDQ $1, SI + + // nextHash := hash(load32(src, s), shift) + MOVL 0(SI), R11 + IMULL $0x1e35a7bd, R11 + SHRL CX, R11 + +outer: + // for { etc } + + // skip := 32 + MOVQ $32, R12 + + // nextS := s + MOVQ SI, R13 + + // candidate := 0 + MOVQ $0, R15 + +inner0: + // for { etc } + + // s := nextS + MOVQ R13, SI + + // bytesBetweenHashLookups := skip >> 5 + MOVQ R12, R14 + SHRQ $5, R14 + + // nextS = s + bytesBetweenHashLookups + ADDQ R14, R13 + + // skip += bytesBetweenHashLookups + ADDQ R14, R12 + + // if nextS > sLimit { goto emitRemainder } + MOVQ R13, AX + SUBQ DX, AX + CMPQ AX, R9 + JA emitRemainder + + // candidate = int(table[nextHash]) + // XXX: MOVWQZX table-32768(SP)(R11*2), R15 + // XXX: 4e 0f b7 7c 5c 78 movzwq 0x78(%rsp,%r11,2),%r15 + BYTE $0x4e + BYTE $0x0f + BYTE $0xb7 + BYTE $0x7c + BYTE $0x5c + BYTE $0x78 + + // table[nextHash] = uint16(s) + MOVQ SI, AX + SUBQ DX, AX + + // XXX: MOVW AX, table-32768(SP)(R11*2) + // XXX: 66 42 89 44 5c 78 mov %ax,0x78(%rsp,%r11,2) + BYTE $0x66 + BYTE $0x42 + BYTE $0x89 + BYTE $0x44 + BYTE $0x5c + BYTE $0x78 + + // nextHash = hash(load32(src, nextS), shift) + MOVL 0(R13), R11 + IMULL $0x1e35a7bd, R11 + SHRL CX, R11 + + // if load32(src, s) != load32(src, candidate) { continue } break + MOVL 0(SI), AX + MOVL (DX)(R15*1), BX + CMPL AX, BX + JNE inner0 + +fourByteMatch: + // As per the encode_other.go code: + // + // A 4-byte match has been found. We'll later see etc. + + // !!! Jump to a fast path for short (<= 16 byte) literals. See the comment + // on inputMargin in encode.go. + MOVQ SI, AX + SUBQ R10, AX + CMPQ AX, $16 + JLE emitLiteralFastPath + + // ---------------------------------------- + // Begin inline of the emitLiteral call. + // + // d += emitLiteral(dst[d:], src[nextEmit:s]) + + MOVL AX, BX + SUBL $1, BX + + CMPL BX, $60 + JLT inlineEmitLiteralOneByte + CMPL BX, $256 + JLT inlineEmitLiteralTwoBytes + +inlineEmitLiteralThreeBytes: + MOVB $0xf4, 0(DI) + MOVW BX, 1(DI) + ADDQ $3, DI + JMP inlineEmitLiteralMemmove + +inlineEmitLiteralTwoBytes: + MOVB $0xf0, 0(DI) + MOVB BX, 1(DI) + ADDQ $2, DI + JMP inlineEmitLiteralMemmove + +inlineEmitLiteralOneByte: + SHLB $2, BX + MOVB BX, 0(DI) + ADDQ $1, DI + +inlineEmitLiteralMemmove: + // Spill local variables (registers) onto the stack; call; unspill. + // + // copy(dst[i:], lit) + // + // This means calling runtime·memmove(&dst[i], &lit[0], len(lit)), so we push + // DI, R10 and AX as arguments. + MOVQ DI, 0(SP) + MOVQ R10, 8(SP) + MOVQ AX, 16(SP) + ADDQ AX, DI // Finish the "d +=" part of "d += emitLiteral(etc)". + MOVQ SI, 72(SP) + MOVQ DI, 80(SP) + MOVQ R15, 112(SP) + CALL runtime·memmove(SB) + MOVQ 56(SP), CX + MOVQ 64(SP), DX + MOVQ 72(SP), SI + MOVQ 80(SP), DI + MOVQ 88(SP), R9 + MOVQ 112(SP), R15 + JMP inner1 + +inlineEmitLiteralEnd: + // End inline of the emitLiteral call. + // ---------------------------------------- + +emitLiteralFastPath: + // !!! Emit the 1-byte encoding "uint8(len(lit)-1)<<2". + MOVB AX, BX + SUBB $1, BX + SHLB $2, BX + MOVB BX, (DI) + ADDQ $1, DI + + // !!! Implement the copy from lit to dst as a 16-byte load and store. + // (Encode's documentation says that dst and src must not overlap.) + // + // This always copies 16 bytes, instead of only len(lit) bytes, but that's + // OK. Subsequent iterations will fix up the overrun. + // + // Note that on amd64, it is legal and cheap to issue unaligned 8-byte or + // 16-byte loads and stores. This technique probably wouldn't be as + // effective on architectures that are fussier about alignment. + MOVOU 0(R10), X0 + MOVOU X0, 0(DI) + ADDQ AX, DI + +inner1: + // for { etc } + + // base := s + MOVQ SI, R12 + + // !!! offset := base - candidate + MOVQ R12, R11 + SUBQ R15, R11 + SUBQ DX, R11 + + // ---------------------------------------- + // Begin inline of the extendMatch call. + // + // s = extendMatch(src, candidate+4, s+4) + + // !!! R14 = &src[len(src)] + MOVQ src_len+32(FP), R14 + ADDQ DX, R14 + + // !!! R13 = &src[len(src) - 8] + MOVQ R14, R13 + SUBQ $8, R13 + + // !!! R15 = &src[candidate + 4] + ADDQ $4, R15 + ADDQ DX, R15 + + // !!! s += 4 + ADDQ $4, SI + +inlineExtendMatchCmp8: + // As long as we are 8 or more bytes before the end of src, we can load and + // compare 8 bytes at a time. If those 8 bytes are equal, repeat. + CMPQ SI, R13 + JA inlineExtendMatchCmp1 + MOVQ (R15), AX + MOVQ (SI), BX + CMPQ AX, BX + JNE inlineExtendMatchBSF + ADDQ $8, R15 + ADDQ $8, SI + JMP inlineExtendMatchCmp8 + +inlineExtendMatchBSF: + // If those 8 bytes were not equal, XOR the two 8 byte values, and return + // the index of the first byte that differs. The BSF instruction finds the + // least significant 1 bit, the amd64 architecture is little-endian, and + // the shift by 3 converts a bit index to a byte index. + XORQ AX, BX + BSFQ BX, BX + SHRQ $3, BX + ADDQ BX, SI + JMP inlineExtendMatchEnd + +inlineExtendMatchCmp1: + // In src's tail, compare 1 byte at a time. + CMPQ SI, R14 + JAE inlineExtendMatchEnd + MOVB (R15), AX + MOVB (SI), BX + CMPB AX, BX + JNE inlineExtendMatchEnd + ADDQ $1, R15 + ADDQ $1, SI + JMP inlineExtendMatchCmp1 + +inlineExtendMatchEnd: + // End inline of the extendMatch call. + // ---------------------------------------- + + // ---------------------------------------- + // Begin inline of the emitCopy call. + // + // d += emitCopy(dst[d:], base-candidate, s-base) + + // !!! length := s - base + MOVQ SI, AX + SUBQ R12, AX + +inlineEmitCopyLoop0: + // for length >= 68 { etc } + CMPL AX, $68 + JLT inlineEmitCopyStep1 + + // Emit a length 64 copy, encoded as 3 bytes. + MOVB $0xfe, 0(DI) + MOVW R11, 1(DI) + ADDQ $3, DI + SUBL $64, AX + JMP inlineEmitCopyLoop0 + +inlineEmitCopyStep1: + // if length > 64 { etc } + CMPL AX, $64 + JLE inlineEmitCopyStep2 + + // Emit a length 60 copy, encoded as 3 bytes. + MOVB $0xee, 0(DI) + MOVW R11, 1(DI) + ADDQ $3, DI + SUBL $60, AX + +inlineEmitCopyStep2: + // if length >= 12 || offset >= 2048 { goto inlineEmitCopyStep3 } + CMPL AX, $12 + JGE inlineEmitCopyStep3 + CMPL R11, $2048 + JGE inlineEmitCopyStep3 + + // Emit the remaining copy, encoded as 2 bytes. + MOVB R11, 1(DI) + SHRL $8, R11 + SHLB $5, R11 + SUBB $4, AX + SHLB $2, AX + ORB AX, R11 + ORB $1, R11 + MOVB R11, 0(DI) + ADDQ $2, DI + JMP inlineEmitCopyEnd + +inlineEmitCopyStep3: + // Emit the remaining copy, encoded as 3 bytes. + SUBL $1, AX + SHLB $2, AX + ORB $2, AX + MOVB AX, 0(DI) + MOVW R11, 1(DI) + ADDQ $3, DI + +inlineEmitCopyEnd: + // End inline of the emitCopy call. + // ---------------------------------------- + + // nextEmit = s + MOVQ SI, R10 + + // if s >= sLimit { goto emitRemainder } + MOVQ SI, AX + SUBQ DX, AX + CMPQ AX, R9 + JAE emitRemainder + + // As per the encode_other.go code: + // + // We could immediately etc. + + // x := load64(src, s-1) + MOVQ -1(SI), R14 + + // prevHash := hash(uint32(x>>0), shift) + MOVL R14, R11 + IMULL $0x1e35a7bd, R11 + SHRL CX, R11 + + // table[prevHash] = uint16(s-1) + MOVQ SI, AX + SUBQ DX, AX + SUBQ $1, AX + + // XXX: MOVW AX, table-32768(SP)(R11*2) + // XXX: 66 42 89 44 5c 78 mov %ax,0x78(%rsp,%r11,2) + BYTE $0x66 + BYTE $0x42 + BYTE $0x89 + BYTE $0x44 + BYTE $0x5c + BYTE $0x78 + + // currHash := hash(uint32(x>>8), shift) + SHRQ $8, R14 + MOVL R14, R11 + IMULL $0x1e35a7bd, R11 + SHRL CX, R11 + + // candidate = int(table[currHash]) + // XXX: MOVWQZX table-32768(SP)(R11*2), R15 + // XXX: 4e 0f b7 7c 5c 78 movzwq 0x78(%rsp,%r11,2),%r15 + BYTE $0x4e + BYTE $0x0f + BYTE $0xb7 + BYTE $0x7c + BYTE $0x5c + BYTE $0x78 + + // table[currHash] = uint16(s) + ADDQ $1, AX + + // XXX: MOVW AX, table-32768(SP)(R11*2) + // XXX: 66 42 89 44 5c 78 mov %ax,0x78(%rsp,%r11,2) + BYTE $0x66 + BYTE $0x42 + BYTE $0x89 + BYTE $0x44 + BYTE $0x5c + BYTE $0x78 + + // if uint32(x>>8) == load32(src, candidate) { continue } + MOVL (DX)(R15*1), BX + CMPL R14, BX + JEQ inner1 + + // nextHash = hash(uint32(x>>16), shift) + SHRQ $8, R14 + MOVL R14, R11 + IMULL $0x1e35a7bd, R11 + SHRL CX, R11 + + // s++ + ADDQ $1, SI + + // break out of the inner1 for loop, i.e. continue the outer loop. + JMP outer + +emitRemainder: + // if nextEmit < len(src) { etc } + MOVQ src_len+32(FP), AX + ADDQ DX, AX + CMPQ R10, AX + JEQ encodeBlockEnd + + // d += emitLiteral(dst[d:], src[nextEmit:]) + // + // Push args. + MOVQ DI, 0(SP) + MOVQ $0, 8(SP) // Unnecessary, as the callee ignores it, but conservative. + MOVQ $0, 16(SP) // Unnecessary, as the callee ignores it, but conservative. + MOVQ R10, 24(SP) + SUBQ R10, AX + MOVQ AX, 32(SP) + MOVQ AX, 40(SP) // Unnecessary, as the callee ignores it, but conservative. + + // Spill local variables (registers) onto the stack; call; unspill. + MOVQ DI, 80(SP) + CALL ·emitLiteral(SB) + MOVQ 80(SP), DI + + // Finish the "d +=" part of "d += emitLiteral(etc)". + ADDQ 48(SP), DI + +encodeBlockEnd: + MOVQ dst_base+0(FP), AX + SUBQ AX, DI + MOVQ DI, d+48(FP) + RET diff --git a/vendor/github.com/golang/snappy/encode_arm64.s b/vendor/github.com/golang/snappy/encode_arm64.s new file mode 100644 index 0000000..f8d54ad --- /dev/null +++ b/vendor/github.com/golang/snappy/encode_arm64.s @@ -0,0 +1,722 @@ +// Copyright 2020 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" + +// The asm code generally follows the pure Go code in encode_other.go, except +// where marked with a "!!!". + +// ---------------------------------------------------------------------------- + +// func emitLiteral(dst, lit []byte) int +// +// All local variables fit into registers. The register allocation: +// - R3 len(lit) +// - R4 n +// - R6 return value +// - R8 &dst[i] +// - R10 &lit[0] +// +// The 32 bytes of stack space is to call runtime·memmove. +// +// The unusual register allocation of local variables, such as R10 for the +// source pointer, matches the allocation used at the call site in encodeBlock, +// which makes it easier to manually inline this function. +TEXT ·emitLiteral(SB), NOSPLIT, $32-56 + MOVD dst_base+0(FP), R8 + MOVD lit_base+24(FP), R10 + MOVD lit_len+32(FP), R3 + MOVD R3, R6 + MOVW R3, R4 + SUBW $1, R4, R4 + + CMPW $60, R4 + BLT oneByte + CMPW $256, R4 + BLT twoBytes + +threeBytes: + MOVD $0xf4, R2 + MOVB R2, 0(R8) + MOVW R4, 1(R8) + ADD $3, R8, R8 + ADD $3, R6, R6 + B memmove + +twoBytes: + MOVD $0xf0, R2 + MOVB R2, 0(R8) + MOVB R4, 1(R8) + ADD $2, R8, R8 + ADD $2, R6, R6 + B memmove + +oneByte: + LSLW $2, R4, R4 + MOVB R4, 0(R8) + ADD $1, R8, R8 + ADD $1, R6, R6 + +memmove: + MOVD R6, ret+48(FP) + + // copy(dst[i:], lit) + // + // This means calling runtime·memmove(&dst[i], &lit[0], len(lit)), so we push + // R8, R10 and R3 as arguments. + MOVD R8, 8(RSP) + MOVD R10, 16(RSP) + MOVD R3, 24(RSP) + CALL runtime·memmove(SB) + RET + +// ---------------------------------------------------------------------------- + +// func emitCopy(dst []byte, offset, length int) int +// +// All local variables fit into registers. The register allocation: +// - R3 length +// - R7 &dst[0] +// - R8 &dst[i] +// - R11 offset +// +// The unusual register allocation of local variables, such as R11 for the +// offset, matches the allocation used at the call site in encodeBlock, which +// makes it easier to manually inline this function. +TEXT ·emitCopy(SB), NOSPLIT, $0-48 + MOVD dst_base+0(FP), R8 + MOVD R8, R7 + MOVD offset+24(FP), R11 + MOVD length+32(FP), R3 + +loop0: + // for length >= 68 { etc } + CMPW $68, R3 + BLT step1 + + // Emit a length 64 copy, encoded as 3 bytes. + MOVD $0xfe, R2 + MOVB R2, 0(R8) + MOVW R11, 1(R8) + ADD $3, R8, R8 + SUB $64, R3, R3 + B loop0 + +step1: + // if length > 64 { etc } + CMP $64, R3 + BLE step2 + + // Emit a length 60 copy, encoded as 3 bytes. + MOVD $0xee, R2 + MOVB R2, 0(R8) + MOVW R11, 1(R8) + ADD $3, R8, R8 + SUB $60, R3, R3 + +step2: + // if length >= 12 || offset >= 2048 { goto step3 } + CMP $12, R3 + BGE step3 + CMPW $2048, R11 + BGE step3 + + // Emit the remaining copy, encoded as 2 bytes. + MOVB R11, 1(R8) + LSRW $3, R11, R11 + AND $0xe0, R11, R11 + SUB $4, R3, R3 + LSLW $2, R3 + AND $0xff, R3, R3 + ORRW R3, R11, R11 + ORRW $1, R11, R11 + MOVB R11, 0(R8) + ADD $2, R8, R8 + + // Return the number of bytes written. + SUB R7, R8, R8 + MOVD R8, ret+40(FP) + RET + +step3: + // Emit the remaining copy, encoded as 3 bytes. + SUB $1, R3, R3 + AND $0xff, R3, R3 + LSLW $2, R3, R3 + ORRW $2, R3, R3 + MOVB R3, 0(R8) + MOVW R11, 1(R8) + ADD $3, R8, R8 + + // Return the number of bytes written. + SUB R7, R8, R8 + MOVD R8, ret+40(FP) + RET + +// ---------------------------------------------------------------------------- + +// func extendMatch(src []byte, i, j int) int +// +// All local variables fit into registers. The register allocation: +// - R6 &src[0] +// - R7 &src[j] +// - R13 &src[len(src) - 8] +// - R14 &src[len(src)] +// - R15 &src[i] +// +// The unusual register allocation of local variables, such as R15 for a source +// pointer, matches the allocation used at the call site in encodeBlock, which +// makes it easier to manually inline this function. +TEXT ·extendMatch(SB), NOSPLIT, $0-48 + MOVD src_base+0(FP), R6 + MOVD src_len+8(FP), R14 + MOVD i+24(FP), R15 + MOVD j+32(FP), R7 + ADD R6, R14, R14 + ADD R6, R15, R15 + ADD R6, R7, R7 + MOVD R14, R13 + SUB $8, R13, R13 + +cmp8: + // As long as we are 8 or more bytes before the end of src, we can load and + // compare 8 bytes at a time. If those 8 bytes are equal, repeat. + CMP R13, R7 + BHI cmp1 + MOVD (R15), R3 + MOVD (R7), R4 + CMP R4, R3 + BNE bsf + ADD $8, R15, R15 + ADD $8, R7, R7 + B cmp8 + +bsf: + // If those 8 bytes were not equal, XOR the two 8 byte values, and return + // the index of the first byte that differs. + // RBIT reverses the bit order, then CLZ counts the leading zeros, the + // combination of which finds the least significant bit which is set. + // The arm64 architecture is little-endian, and the shift by 3 converts + // a bit index to a byte index. + EOR R3, R4, R4 + RBIT R4, R4 + CLZ R4, R4 + ADD R4>>3, R7, R7 + + // Convert from &src[ret] to ret. + SUB R6, R7, R7 + MOVD R7, ret+40(FP) + RET + +cmp1: + // In src's tail, compare 1 byte at a time. + CMP R7, R14 + BLS extendMatchEnd + MOVB (R15), R3 + MOVB (R7), R4 + CMP R4, R3 + BNE extendMatchEnd + ADD $1, R15, R15 + ADD $1, R7, R7 + B cmp1 + +extendMatchEnd: + // Convert from &src[ret] to ret. + SUB R6, R7, R7 + MOVD R7, ret+40(FP) + RET + +// ---------------------------------------------------------------------------- + +// func encodeBlock(dst, src []byte) (d int) +// +// All local variables fit into registers, other than "var table". The register +// allocation: +// - R3 . . +// - R4 . . +// - R5 64 shift +// - R6 72 &src[0], tableSize +// - R7 80 &src[s] +// - R8 88 &dst[d] +// - R9 96 sLimit +// - R10 . &src[nextEmit] +// - R11 104 prevHash, currHash, nextHash, offset +// - R12 112 &src[base], skip +// - R13 . &src[nextS], &src[len(src) - 8] +// - R14 . len(src), bytesBetweenHashLookups, &src[len(src)], x +// - R15 120 candidate +// - R16 . hash constant, 0x1e35a7bd +// - R17 . &table +// - . 128 table +// +// The second column (64, 72, etc) is the stack offset to spill the registers +// when calling other functions. We could pack this slightly tighter, but it's +// simpler to have a dedicated spill map independent of the function called. +// +// "var table [maxTableSize]uint16" takes up 32768 bytes of stack space. An +// extra 64 bytes, to call other functions, and an extra 64 bytes, to spill +// local variables (registers) during calls gives 32768 + 64 + 64 = 32896. +TEXT ·encodeBlock(SB), 0, $32896-56 + MOVD dst_base+0(FP), R8 + MOVD src_base+24(FP), R7 + MOVD src_len+32(FP), R14 + + // shift, tableSize := uint32(32-8), 1<<8 + MOVD $24, R5 + MOVD $256, R6 + MOVW $0xa7bd, R16 + MOVKW $(0x1e35<<16), R16 + +calcShift: + // for ; tableSize < maxTableSize && tableSize < len(src); tableSize *= 2 { + // shift-- + // } + MOVD $16384, R2 + CMP R2, R6 + BGE varTable + CMP R14, R6 + BGE varTable + SUB $1, R5, R5 + LSL $1, R6, R6 + B calcShift + +varTable: + // var table [maxTableSize]uint16 + // + // In the asm code, unlike the Go code, we can zero-initialize only the + // first tableSize elements. Each uint16 element is 2 bytes and each + // iterations writes 64 bytes, so we can do only tableSize/32 writes + // instead of the 2048 writes that would zero-initialize all of table's + // 32768 bytes. This clear could overrun the first tableSize elements, but + // it won't overrun the allocated stack size. + ADD $128, RSP, R17 + MOVD R17, R4 + + // !!! R6 = &src[tableSize] + ADD R6<<1, R17, R6 + +memclr: + STP.P (ZR, ZR), 64(R4) + STP (ZR, ZR), -48(R4) + STP (ZR, ZR), -32(R4) + STP (ZR, ZR), -16(R4) + CMP R4, R6 + BHI memclr + + // !!! R6 = &src[0] + MOVD R7, R6 + + // sLimit := len(src) - inputMargin + MOVD R14, R9 + SUB $15, R9, R9 + + // !!! Pre-emptively spill R5, R6 and R9 to the stack. Their values don't + // change for the rest of the function. + MOVD R5, 64(RSP) + MOVD R6, 72(RSP) + MOVD R9, 96(RSP) + + // nextEmit := 0 + MOVD R6, R10 + + // s := 1 + ADD $1, R7, R7 + + // nextHash := hash(load32(src, s), shift) + MOVW 0(R7), R11 + MULW R16, R11, R11 + LSRW R5, R11, R11 + +outer: + // for { etc } + + // skip := 32 + MOVD $32, R12 + + // nextS := s + MOVD R7, R13 + + // candidate := 0 + MOVD $0, R15 + +inner0: + // for { etc } + + // s := nextS + MOVD R13, R7 + + // bytesBetweenHashLookups := skip >> 5 + MOVD R12, R14 + LSR $5, R14, R14 + + // nextS = s + bytesBetweenHashLookups + ADD R14, R13, R13 + + // skip += bytesBetweenHashLookups + ADD R14, R12, R12 + + // if nextS > sLimit { goto emitRemainder } + MOVD R13, R3 + SUB R6, R3, R3 + CMP R9, R3 + BHI emitRemainder + + // candidate = int(table[nextHash]) + MOVHU 0(R17)(R11<<1), R15 + + // table[nextHash] = uint16(s) + MOVD R7, R3 + SUB R6, R3, R3 + + MOVH R3, 0(R17)(R11<<1) + + // nextHash = hash(load32(src, nextS), shift) + MOVW 0(R13), R11 + MULW R16, R11 + LSRW R5, R11, R11 + + // if load32(src, s) != load32(src, candidate) { continue } break + MOVW 0(R7), R3 + MOVW (R6)(R15), R4 + CMPW R4, R3 + BNE inner0 + +fourByteMatch: + // As per the encode_other.go code: + // + // A 4-byte match has been found. We'll later see etc. + + // !!! Jump to a fast path for short (<= 16 byte) literals. See the comment + // on inputMargin in encode.go. + MOVD R7, R3 + SUB R10, R3, R3 + CMP $16, R3 + BLE emitLiteralFastPath + + // ---------------------------------------- + // Begin inline of the emitLiteral call. + // + // d += emitLiteral(dst[d:], src[nextEmit:s]) + + MOVW R3, R4 + SUBW $1, R4, R4 + + MOVW $60, R2 + CMPW R2, R4 + BLT inlineEmitLiteralOneByte + MOVW $256, R2 + CMPW R2, R4 + BLT inlineEmitLiteralTwoBytes + +inlineEmitLiteralThreeBytes: + MOVD $0xf4, R1 + MOVB R1, 0(R8) + MOVW R4, 1(R8) + ADD $3, R8, R8 + B inlineEmitLiteralMemmove + +inlineEmitLiteralTwoBytes: + MOVD $0xf0, R1 + MOVB R1, 0(R8) + MOVB R4, 1(R8) + ADD $2, R8, R8 + B inlineEmitLiteralMemmove + +inlineEmitLiteralOneByte: + LSLW $2, R4, R4 + MOVB R4, 0(R8) + ADD $1, R8, R8 + +inlineEmitLiteralMemmove: + // Spill local variables (registers) onto the stack; call; unspill. + // + // copy(dst[i:], lit) + // + // This means calling runtime·memmove(&dst[i], &lit[0], len(lit)), so we push + // R8, R10 and R3 as arguments. + MOVD R8, 8(RSP) + MOVD R10, 16(RSP) + MOVD R3, 24(RSP) + + // Finish the "d +=" part of "d += emitLiteral(etc)". + ADD R3, R8, R8 + MOVD R7, 80(RSP) + MOVD R8, 88(RSP) + MOVD R15, 120(RSP) + CALL runtime·memmove(SB) + MOVD 64(RSP), R5 + MOVD 72(RSP), R6 + MOVD 80(RSP), R7 + MOVD 88(RSP), R8 + MOVD 96(RSP), R9 + MOVD 120(RSP), R15 + ADD $128, RSP, R17 + MOVW $0xa7bd, R16 + MOVKW $(0x1e35<<16), R16 + B inner1 + +inlineEmitLiteralEnd: + // End inline of the emitLiteral call. + // ---------------------------------------- + +emitLiteralFastPath: + // !!! Emit the 1-byte encoding "uint8(len(lit)-1)<<2". + MOVB R3, R4 + SUBW $1, R4, R4 + AND $0xff, R4, R4 + LSLW $2, R4, R4 + MOVB R4, (R8) + ADD $1, R8, R8 + + // !!! Implement the copy from lit to dst as a 16-byte load and store. + // (Encode's documentation says that dst and src must not overlap.) + // + // This always copies 16 bytes, instead of only len(lit) bytes, but that's + // OK. Subsequent iterations will fix up the overrun. + // + // Note that on arm64, it is legal and cheap to issue unaligned 8-byte or + // 16-byte loads and stores. This technique probably wouldn't be as + // effective on architectures that are fussier about alignment. + LDP 0(R10), (R0, R1) + STP (R0, R1), 0(R8) + ADD R3, R8, R8 + +inner1: + // for { etc } + + // base := s + MOVD R7, R12 + + // !!! offset := base - candidate + MOVD R12, R11 + SUB R15, R11, R11 + SUB R6, R11, R11 + + // ---------------------------------------- + // Begin inline of the extendMatch call. + // + // s = extendMatch(src, candidate+4, s+4) + + // !!! R14 = &src[len(src)] + MOVD src_len+32(FP), R14 + ADD R6, R14, R14 + + // !!! R13 = &src[len(src) - 8] + MOVD R14, R13 + SUB $8, R13, R13 + + // !!! R15 = &src[candidate + 4] + ADD $4, R15, R15 + ADD R6, R15, R15 + + // !!! s += 4 + ADD $4, R7, R7 + +inlineExtendMatchCmp8: + // As long as we are 8 or more bytes before the end of src, we can load and + // compare 8 bytes at a time. If those 8 bytes are equal, repeat. + CMP R13, R7 + BHI inlineExtendMatchCmp1 + MOVD (R15), R3 + MOVD (R7), R4 + CMP R4, R3 + BNE inlineExtendMatchBSF + ADD $8, R15, R15 + ADD $8, R7, R7 + B inlineExtendMatchCmp8 + +inlineExtendMatchBSF: + // If those 8 bytes were not equal, XOR the two 8 byte values, and return + // the index of the first byte that differs. + // RBIT reverses the bit order, then CLZ counts the leading zeros, the + // combination of which finds the least significant bit which is set. + // The arm64 architecture is little-endian, and the shift by 3 converts + // a bit index to a byte index. + EOR R3, R4, R4 + RBIT R4, R4 + CLZ R4, R4 + ADD R4>>3, R7, R7 + B inlineExtendMatchEnd + +inlineExtendMatchCmp1: + // In src's tail, compare 1 byte at a time. + CMP R7, R14 + BLS inlineExtendMatchEnd + MOVB (R15), R3 + MOVB (R7), R4 + CMP R4, R3 + BNE inlineExtendMatchEnd + ADD $1, R15, R15 + ADD $1, R7, R7 + B inlineExtendMatchCmp1 + +inlineExtendMatchEnd: + // End inline of the extendMatch call. + // ---------------------------------------- + + // ---------------------------------------- + // Begin inline of the emitCopy call. + // + // d += emitCopy(dst[d:], base-candidate, s-base) + + // !!! length := s - base + MOVD R7, R3 + SUB R12, R3, R3 + +inlineEmitCopyLoop0: + // for length >= 68 { etc } + MOVW $68, R2 + CMPW R2, R3 + BLT inlineEmitCopyStep1 + + // Emit a length 64 copy, encoded as 3 bytes. + MOVD $0xfe, R1 + MOVB R1, 0(R8) + MOVW R11, 1(R8) + ADD $3, R8, R8 + SUBW $64, R3, R3 + B inlineEmitCopyLoop0 + +inlineEmitCopyStep1: + // if length > 64 { etc } + MOVW $64, R2 + CMPW R2, R3 + BLE inlineEmitCopyStep2 + + // Emit a length 60 copy, encoded as 3 bytes. + MOVD $0xee, R1 + MOVB R1, 0(R8) + MOVW R11, 1(R8) + ADD $3, R8, R8 + SUBW $60, R3, R3 + +inlineEmitCopyStep2: + // if length >= 12 || offset >= 2048 { goto inlineEmitCopyStep3 } + MOVW $12, R2 + CMPW R2, R3 + BGE inlineEmitCopyStep3 + MOVW $2048, R2 + CMPW R2, R11 + BGE inlineEmitCopyStep3 + + // Emit the remaining copy, encoded as 2 bytes. + MOVB R11, 1(R8) + LSRW $8, R11, R11 + LSLW $5, R11, R11 + SUBW $4, R3, R3 + AND $0xff, R3, R3 + LSLW $2, R3, R3 + ORRW R3, R11, R11 + ORRW $1, R11, R11 + MOVB R11, 0(R8) + ADD $2, R8, R8 + B inlineEmitCopyEnd + +inlineEmitCopyStep3: + // Emit the remaining copy, encoded as 3 bytes. + SUBW $1, R3, R3 + LSLW $2, R3, R3 + ORRW $2, R3, R3 + MOVB R3, 0(R8) + MOVW R11, 1(R8) + ADD $3, R8, R8 + +inlineEmitCopyEnd: + // End inline of the emitCopy call. + // ---------------------------------------- + + // nextEmit = s + MOVD R7, R10 + + // if s >= sLimit { goto emitRemainder } + MOVD R7, R3 + SUB R6, R3, R3 + CMP R3, R9 + BLS emitRemainder + + // As per the encode_other.go code: + // + // We could immediately etc. + + // x := load64(src, s-1) + MOVD -1(R7), R14 + + // prevHash := hash(uint32(x>>0), shift) + MOVW R14, R11 + MULW R16, R11, R11 + LSRW R5, R11, R11 + + // table[prevHash] = uint16(s-1) + MOVD R7, R3 + SUB R6, R3, R3 + SUB $1, R3, R3 + + MOVHU R3, 0(R17)(R11<<1) + + // currHash := hash(uint32(x>>8), shift) + LSR $8, R14, R14 + MOVW R14, R11 + MULW R16, R11, R11 + LSRW R5, R11, R11 + + // candidate = int(table[currHash]) + MOVHU 0(R17)(R11<<1), R15 + + // table[currHash] = uint16(s) + ADD $1, R3, R3 + MOVHU R3, 0(R17)(R11<<1) + + // if uint32(x>>8) == load32(src, candidate) { continue } + MOVW (R6)(R15), R4 + CMPW R4, R14 + BEQ inner1 + + // nextHash = hash(uint32(x>>16), shift) + LSR $8, R14, R14 + MOVW R14, R11 + MULW R16, R11, R11 + LSRW R5, R11, R11 + + // s++ + ADD $1, R7, R7 + + // break out of the inner1 for loop, i.e. continue the outer loop. + B outer + +emitRemainder: + // if nextEmit < len(src) { etc } + MOVD src_len+32(FP), R3 + ADD R6, R3, R3 + CMP R3, R10 + BEQ encodeBlockEnd + + // d += emitLiteral(dst[d:], src[nextEmit:]) + // + // Push args. + MOVD R8, 8(RSP) + MOVD $0, 16(RSP) // Unnecessary, as the callee ignores it, but conservative. + MOVD $0, 24(RSP) // Unnecessary, as the callee ignores it, but conservative. + MOVD R10, 32(RSP) + SUB R10, R3, R3 + MOVD R3, 40(RSP) + MOVD R3, 48(RSP) // Unnecessary, as the callee ignores it, but conservative. + + // Spill local variables (registers) onto the stack; call; unspill. + MOVD R8, 88(RSP) + CALL ·emitLiteral(SB) + MOVD 88(RSP), R8 + + // Finish the "d +=" part of "d += emitLiteral(etc)". + MOVD 56(RSP), R1 + ADD R1, R8, R8 + +encodeBlockEnd: + MOVD dst_base+0(FP), R3 + SUB R3, R8, R8 + MOVD R8, d+48(FP) + RET diff --git a/vendor/github.com/golang/snappy/encode_asm.go b/vendor/github.com/golang/snappy/encode_asm.go new file mode 100644 index 0000000..107c1e7 --- /dev/null +++ b/vendor/github.com/golang/snappy/encode_asm.go @@ -0,0 +1,30 @@ +// Copyright 2016 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine +// +build gc +// +build !noasm +// +build amd64 arm64 + +package snappy + +// emitLiteral has the same semantics as in encode_other.go. +// +//go:noescape +func emitLiteral(dst, lit []byte) int + +// emitCopy has the same semantics as in encode_other.go. +// +//go:noescape +func emitCopy(dst []byte, offset, length int) int + +// extendMatch has the same semantics as in encode_other.go. +// +//go:noescape +func extendMatch(src []byte, i, j int) int + +// encodeBlock has the same semantics as in encode_other.go. +// +//go:noescape +func encodeBlock(dst, src []byte) (d int) diff --git a/vendor/github.com/golang/snappy/encode_other.go b/vendor/github.com/golang/snappy/encode_other.go new file mode 100644 index 0000000..296d7f0 --- /dev/null +++ b/vendor/github.com/golang/snappy/encode_other.go @@ -0,0 +1,238 @@ +// Copyright 2016 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64,!arm64 appengine !gc noasm + +package snappy + +func load32(b []byte, i int) uint32 { + b = b[i : i+4 : len(b)] // Help the compiler eliminate bounds checks on the next line. + return uint32(b[0]) | uint32(b[1])<<8 | uint32(b[2])<<16 | uint32(b[3])<<24 +} + +func load64(b []byte, i int) uint64 { + b = b[i : i+8 : len(b)] // Help the compiler eliminate bounds checks on the next line. + return uint64(b[0]) | uint64(b[1])<<8 | uint64(b[2])<<16 | uint64(b[3])<<24 | + uint64(b[4])<<32 | uint64(b[5])<<40 | uint64(b[6])<<48 | uint64(b[7])<<56 +} + +// emitLiteral writes a literal chunk and returns the number of bytes written. +// +// It assumes that: +// dst is long enough to hold the encoded bytes +// 1 <= len(lit) && len(lit) <= 65536 +func emitLiteral(dst, lit []byte) int { + i, n := 0, uint(len(lit)-1) + switch { + case n < 60: + dst[0] = uint8(n)<<2 | tagLiteral + i = 1 + case n < 1<<8: + dst[0] = 60<<2 | tagLiteral + dst[1] = uint8(n) + i = 2 + default: + dst[0] = 61<<2 | tagLiteral + dst[1] = uint8(n) + dst[2] = uint8(n >> 8) + i = 3 + } + return i + copy(dst[i:], lit) +} + +// emitCopy writes a copy chunk and returns the number of bytes written. +// +// It assumes that: +// dst is long enough to hold the encoded bytes +// 1 <= offset && offset <= 65535 +// 4 <= length && length <= 65535 +func emitCopy(dst []byte, offset, length int) int { + i := 0 + // The maximum length for a single tagCopy1 or tagCopy2 op is 64 bytes. The + // threshold for this loop is a little higher (at 68 = 64 + 4), and the + // length emitted down below is is a little lower (at 60 = 64 - 4), because + // it's shorter to encode a length 67 copy as a length 60 tagCopy2 followed + // by a length 7 tagCopy1 (which encodes as 3+2 bytes) than to encode it as + // a length 64 tagCopy2 followed by a length 3 tagCopy2 (which encodes as + // 3+3 bytes). The magic 4 in the 64±4 is because the minimum length for a + // tagCopy1 op is 4 bytes, which is why a length 3 copy has to be an + // encodes-as-3-bytes tagCopy2 instead of an encodes-as-2-bytes tagCopy1. + for length >= 68 { + // Emit a length 64 copy, encoded as 3 bytes. + dst[i+0] = 63<<2 | tagCopy2 + dst[i+1] = uint8(offset) + dst[i+2] = uint8(offset >> 8) + i += 3 + length -= 64 + } + if length > 64 { + // Emit a length 60 copy, encoded as 3 bytes. + dst[i+0] = 59<<2 | tagCopy2 + dst[i+1] = uint8(offset) + dst[i+2] = uint8(offset >> 8) + i += 3 + length -= 60 + } + if length >= 12 || offset >= 2048 { + // Emit the remaining copy, encoded as 3 bytes. + dst[i+0] = uint8(length-1)<<2 | tagCopy2 + dst[i+1] = uint8(offset) + dst[i+2] = uint8(offset >> 8) + return i + 3 + } + // Emit the remaining copy, encoded as 2 bytes. + dst[i+0] = uint8(offset>>8)<<5 | uint8(length-4)<<2 | tagCopy1 + dst[i+1] = uint8(offset) + return i + 2 +} + +// extendMatch returns the largest k such that k <= len(src) and that +// src[i:i+k-j] and src[j:k] have the same contents. +// +// It assumes that: +// 0 <= i && i < j && j <= len(src) +func extendMatch(src []byte, i, j int) int { + for ; j < len(src) && src[i] == src[j]; i, j = i+1, j+1 { + } + return j +} + +func hash(u, shift uint32) uint32 { + return (u * 0x1e35a7bd) >> shift +} + +// encodeBlock encodes a non-empty src to a guaranteed-large-enough dst. It +// assumes that the varint-encoded length of the decompressed bytes has already +// been written. +// +// It also assumes that: +// len(dst) >= MaxEncodedLen(len(src)) && +// minNonLiteralBlockSize <= len(src) && len(src) <= maxBlockSize +func encodeBlock(dst, src []byte) (d int) { + // Initialize the hash table. Its size ranges from 1<<8 to 1<<14 inclusive. + // The table element type is uint16, as s < sLimit and sLimit < len(src) + // and len(src) <= maxBlockSize and maxBlockSize == 65536. + const ( + maxTableSize = 1 << 14 + // tableMask is redundant, but helps the compiler eliminate bounds + // checks. + tableMask = maxTableSize - 1 + ) + shift := uint32(32 - 8) + for tableSize := 1 << 8; tableSize < maxTableSize && tableSize < len(src); tableSize *= 2 { + shift-- + } + // In Go, all array elements are zero-initialized, so there is no advantage + // to a smaller tableSize per se. However, it matches the C++ algorithm, + // and in the asm versions of this code, we can get away with zeroing only + // the first tableSize elements. + var table [maxTableSize]uint16 + + // sLimit is when to stop looking for offset/length copies. The inputMargin + // lets us use a fast path for emitLiteral in the main loop, while we are + // looking for copies. + sLimit := len(src) - inputMargin + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := 0 + + // The encoded form must start with a literal, as there are no previous + // bytes to copy, so we start looking for hash matches at s == 1. + s := 1 + nextHash := hash(load32(src, s), shift) + + for { + // Copied from the C++ snappy implementation: + // + // Heuristic match skipping: If 32 bytes are scanned with no matches + // found, start looking only at every other byte. If 32 more bytes are + // scanned (or skipped), look at every third byte, etc.. When a match + // is found, immediately go back to looking at every byte. This is a + // small loss (~5% performance, ~0.1% density) for compressible data + // due to more bookkeeping, but for non-compressible data (such as + // JPEG) it's a huge win since the compressor quickly "realizes" the + // data is incompressible and doesn't bother looking for matches + // everywhere. + // + // The "skip" variable keeps track of how many bytes there are since + // the last match; dividing it by 32 (ie. right-shifting by five) gives + // the number of bytes to move ahead for each iteration. + skip := 32 + + nextS := s + candidate := 0 + for { + s = nextS + bytesBetweenHashLookups := skip >> 5 + nextS = s + bytesBetweenHashLookups + skip += bytesBetweenHashLookups + if nextS > sLimit { + goto emitRemainder + } + candidate = int(table[nextHash&tableMask]) + table[nextHash&tableMask] = uint16(s) + nextHash = hash(load32(src, nextS), shift) + if load32(src, s) == load32(src, candidate) { + break + } + } + + // A 4-byte match has been found. We'll later see if more than 4 bytes + // match. But, prior to the match, src[nextEmit:s] are unmatched. Emit + // them as literal bytes. + d += emitLiteral(dst[d:], src[nextEmit:s]) + + // Call emitCopy, and then see if another emitCopy could be our next + // move. Repeat until we find no match for the input immediately after + // what was consumed by the last emitCopy call. + // + // If we exit this loop normally then we need to call emitLiteral next, + // though we don't yet know how big the literal will be. We handle that + // by proceeding to the next iteration of the main loop. We also can + // exit this loop via goto if we get close to exhausting the input. + for { + // Invariant: we have a 4-byte match at s, and no need to emit any + // literal bytes prior to s. + base := s + + // Extend the 4-byte match as long as possible. + // + // This is an inlined version of: + // s = extendMatch(src, candidate+4, s+4) + s += 4 + for i := candidate + 4; s < len(src) && src[i] == src[s]; i, s = i+1, s+1 { + } + + d += emitCopy(dst[d:], base-candidate, s-base) + nextEmit = s + if s >= sLimit { + goto emitRemainder + } + + // We could immediately start working at s now, but to improve + // compression we first update the hash table at s-1 and at s. If + // another emitCopy is not our next move, also calculate nextHash + // at s+1. At least on GOARCH=amd64, these three hash calculations + // are faster as one load64 call (with some shifts) instead of + // three load32 calls. + x := load64(src, s-1) + prevHash := hash(uint32(x>>0), shift) + table[prevHash&tableMask] = uint16(s - 1) + currHash := hash(uint32(x>>8), shift) + candidate = int(table[currHash&tableMask]) + table[currHash&tableMask] = uint16(s) + if uint32(x>>8) != load32(src, candidate) { + nextHash = hash(uint32(x>>16), shift) + s++ + break + } + } + } + +emitRemainder: + if nextEmit < len(src) { + d += emitLiteral(dst[d:], src[nextEmit:]) + } + return d +} diff --git a/vendor/github.com/golang/snappy/snappy.go b/vendor/github.com/golang/snappy/snappy.go new file mode 100644 index 0000000..ece692e --- /dev/null +++ b/vendor/github.com/golang/snappy/snappy.go @@ -0,0 +1,98 @@ +// Copyright 2011 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package snappy implements the Snappy compression format. It aims for very +// high speeds and reasonable compression. +// +// There are actually two Snappy formats: block and stream. They are related, +// but different: trying to decompress block-compressed data as a Snappy stream +// will fail, and vice versa. The block format is the Decode and Encode +// functions and the stream format is the Reader and Writer types. +// +// The block format, the more common case, is used when the complete size (the +// number of bytes) of the original data is known upfront, at the time +// compression starts. The stream format, also known as the framing format, is +// for when that isn't always true. +// +// The canonical, C++ implementation is at https://github.com/google/snappy and +// it only implements the block format. +package snappy // import "github.com/golang/snappy" + +import ( + "hash/crc32" +) + +/* +Each encoded block begins with the varint-encoded length of the decoded data, +followed by a sequence of chunks. Chunks begin and end on byte boundaries. The +first byte of each chunk is broken into its 2 least and 6 most significant bits +called l and m: l ranges in [0, 4) and m ranges in [0, 64). l is the chunk tag. +Zero means a literal tag. All other values mean a copy tag. + +For literal tags: + - If m < 60, the next 1 + m bytes are literal bytes. + - Otherwise, let n be the little-endian unsigned integer denoted by the next + m - 59 bytes. The next 1 + n bytes after that are literal bytes. + +For copy tags, length bytes are copied from offset bytes ago, in the style of +Lempel-Ziv compression algorithms. In particular: + - For l == 1, the offset ranges in [0, 1<<11) and the length in [4, 12). + The length is 4 + the low 3 bits of m. The high 3 bits of m form bits 8-10 + of the offset. The next byte is bits 0-7 of the offset. + - For l == 2, the offset ranges in [0, 1<<16) and the length in [1, 65). + The length is 1 + m. The offset is the little-endian unsigned integer + denoted by the next 2 bytes. + - For l == 3, this tag is a legacy format that is no longer issued by most + encoders. Nonetheless, the offset ranges in [0, 1<<32) and the length in + [1, 65). The length is 1 + m. The offset is the little-endian unsigned + integer denoted by the next 4 bytes. +*/ +const ( + tagLiteral = 0x00 + tagCopy1 = 0x01 + tagCopy2 = 0x02 + tagCopy4 = 0x03 +) + +const ( + checksumSize = 4 + chunkHeaderSize = 4 + magicChunk = "\xff\x06\x00\x00" + magicBody + magicBody = "sNaPpY" + + // maxBlockSize is the maximum size of the input to encodeBlock. It is not + // part of the wire format per se, but some parts of the encoder assume + // that an offset fits into a uint16. + // + // Also, for the framing format (Writer type instead of Encode function), + // https://github.com/google/snappy/blob/master/framing_format.txt says + // that "the uncompressed data in a chunk must be no longer than 65536 + // bytes". + maxBlockSize = 65536 + + // maxEncodedLenOfMaxBlockSize equals MaxEncodedLen(maxBlockSize), but is + // hard coded to be a const instead of a variable, so that obufLen can also + // be a const. Their equivalence is confirmed by + // TestMaxEncodedLenOfMaxBlockSize. + maxEncodedLenOfMaxBlockSize = 76490 + + obufHeaderLen = len(magicChunk) + checksumSize + chunkHeaderSize + obufLen = obufHeaderLen + maxEncodedLenOfMaxBlockSize +) + +const ( + chunkTypeCompressedData = 0x00 + chunkTypeUncompressedData = 0x01 + chunkTypePadding = 0xfe + chunkTypeStreamIdentifier = 0xff +) + +var crcTable = crc32.MakeTable(crc32.Castagnoli) + +// crc implements the checksum specified in section 3 of +// https://github.com/google/snappy/blob/master/framing_format.txt +func crc(b []byte) uint32 { + c := crc32.Update(0, crcTable, b) + return uint32(c>>15|c<<17) + 0xa282ead8 +} diff --git a/vendor/github.com/google/uuid/.travis.yml b/vendor/github.com/google/uuid/.travis.yml new file mode 100644 index 0000000..d8156a6 --- /dev/null +++ b/vendor/github.com/google/uuid/.travis.yml @@ -0,0 +1,9 @@ +language: go + +go: + - 1.4.3 + - 1.5.3 + - tip + +script: + - go test -v ./... diff --git a/vendor/github.com/google/uuid/CONTRIBUTING.md b/vendor/github.com/google/uuid/CONTRIBUTING.md new file mode 100644 index 0000000..04fdf09 --- /dev/null +++ b/vendor/github.com/google/uuid/CONTRIBUTING.md @@ -0,0 +1,10 @@ +# How to contribute + +We definitely welcome patches and contribution to this project! + +### Legal requirements + +In order to protect both you and ourselves, you will need to sign the +[Contributor License Agreement](https://cla.developers.google.com/clas). + +You may have already signed it for other Google projects. diff --git a/vendor/github.com/google/uuid/CONTRIBUTORS b/vendor/github.com/google/uuid/CONTRIBUTORS new file mode 100644 index 0000000..b4bb97f --- /dev/null +++ b/vendor/github.com/google/uuid/CONTRIBUTORS @@ -0,0 +1,9 @@ +Paul Borman +bmatsuo +shawnps +theory +jboverfelt +dsymonds +cd1 +wallclockbuilder +dansouza diff --git a/vendor/github.com/google/uuid/LICENSE b/vendor/github.com/google/uuid/LICENSE new file mode 100644 index 0000000..5dc6826 --- /dev/null +++ b/vendor/github.com/google/uuid/LICENSE @@ -0,0 +1,27 @@ +Copyright (c) 2009,2014 Google Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/github.com/google/uuid/README.md b/vendor/github.com/google/uuid/README.md new file mode 100644 index 0000000..f765a46 --- /dev/null +++ b/vendor/github.com/google/uuid/README.md @@ -0,0 +1,19 @@ +# uuid ![build status](https://travis-ci.org/google/uuid.svg?branch=master) +The uuid package generates and inspects UUIDs based on +[RFC 4122](http://tools.ietf.org/html/rfc4122) +and DCE 1.1: Authentication and Security Services. + +This package is based on the github.com/pborman/uuid package (previously named +code.google.com/p/go-uuid). It differs from these earlier packages in that +a UUID is a 16 byte array rather than a byte slice. One loss due to this +change is the ability to represent an invalid UUID (vs a NIL UUID). + +###### Install +`go get github.com/google/uuid` + +###### Documentation +[![GoDoc](https://godoc.org/github.com/google/uuid?status.svg)](http://godoc.org/github.com/google/uuid) + +Full `go doc` style documentation for the package can be viewed online without +installing this package by using the GoDoc site here: +http://pkg.go.dev/github.com/google/uuid diff --git a/vendor/github.com/google/uuid/dce.go b/vendor/github.com/google/uuid/dce.go new file mode 100644 index 0000000..fa820b9 --- /dev/null +++ b/vendor/github.com/google/uuid/dce.go @@ -0,0 +1,80 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "encoding/binary" + "fmt" + "os" +) + +// A Domain represents a Version 2 domain +type Domain byte + +// Domain constants for DCE Security (Version 2) UUIDs. +const ( + Person = Domain(0) + Group = Domain(1) + Org = Domain(2) +) + +// NewDCESecurity returns a DCE Security (Version 2) UUID. +// +// The domain should be one of Person, Group or Org. +// On a POSIX system the id should be the users UID for the Person +// domain and the users GID for the Group. The meaning of id for +// the domain Org or on non-POSIX systems is site defined. +// +// For a given domain/id pair the same token may be returned for up to +// 7 minutes and 10 seconds. +func NewDCESecurity(domain Domain, id uint32) (UUID, error) { + uuid, err := NewUUID() + if err == nil { + uuid[6] = (uuid[6] & 0x0f) | 0x20 // Version 2 + uuid[9] = byte(domain) + binary.BigEndian.PutUint32(uuid[0:], id) + } + return uuid, err +} + +// NewDCEPerson returns a DCE Security (Version 2) UUID in the person +// domain with the id returned by os.Getuid. +// +// NewDCESecurity(Person, uint32(os.Getuid())) +func NewDCEPerson() (UUID, error) { + return NewDCESecurity(Person, uint32(os.Getuid())) +} + +// NewDCEGroup returns a DCE Security (Version 2) UUID in the group +// domain with the id returned by os.Getgid. +// +// NewDCESecurity(Group, uint32(os.Getgid())) +func NewDCEGroup() (UUID, error) { + return NewDCESecurity(Group, uint32(os.Getgid())) +} + +// Domain returns the domain for a Version 2 UUID. Domains are only defined +// for Version 2 UUIDs. +func (uuid UUID) Domain() Domain { + return Domain(uuid[9]) +} + +// ID returns the id for a Version 2 UUID. IDs are only defined for Version 2 +// UUIDs. +func (uuid UUID) ID() uint32 { + return binary.BigEndian.Uint32(uuid[0:4]) +} + +func (d Domain) String() string { + switch d { + case Person: + return "Person" + case Group: + return "Group" + case Org: + return "Org" + } + return fmt.Sprintf("Domain%d", int(d)) +} diff --git a/vendor/github.com/google/uuid/doc.go b/vendor/github.com/google/uuid/doc.go new file mode 100644 index 0000000..5b8a4b9 --- /dev/null +++ b/vendor/github.com/google/uuid/doc.go @@ -0,0 +1,12 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package uuid generates and inspects UUIDs. +// +// UUIDs are based on RFC 4122 and DCE 1.1: Authentication and Security +// Services. +// +// A UUID is a 16 byte (128 bit) array. UUIDs may be used as keys to +// maps or compared directly. +package uuid diff --git a/vendor/github.com/google/uuid/hash.go b/vendor/github.com/google/uuid/hash.go new file mode 100644 index 0000000..b404f4b --- /dev/null +++ b/vendor/github.com/google/uuid/hash.go @@ -0,0 +1,53 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "crypto/md5" + "crypto/sha1" + "hash" +) + +// Well known namespace IDs and UUIDs +var ( + NameSpaceDNS = Must(Parse("6ba7b810-9dad-11d1-80b4-00c04fd430c8")) + NameSpaceURL = Must(Parse("6ba7b811-9dad-11d1-80b4-00c04fd430c8")) + NameSpaceOID = Must(Parse("6ba7b812-9dad-11d1-80b4-00c04fd430c8")) + NameSpaceX500 = Must(Parse("6ba7b814-9dad-11d1-80b4-00c04fd430c8")) + Nil UUID // empty UUID, all zeros +) + +// NewHash returns a new UUID derived from the hash of space concatenated with +// data generated by h. The hash should be at least 16 byte in length. The +// first 16 bytes of the hash are used to form the UUID. The version of the +// UUID will be the lower 4 bits of version. NewHash is used to implement +// NewMD5 and NewSHA1. +func NewHash(h hash.Hash, space UUID, data []byte, version int) UUID { + h.Reset() + h.Write(space[:]) //nolint:errcheck + h.Write(data) //nolint:errcheck + s := h.Sum(nil) + var uuid UUID + copy(uuid[:], s) + uuid[6] = (uuid[6] & 0x0f) | uint8((version&0xf)<<4) + uuid[8] = (uuid[8] & 0x3f) | 0x80 // RFC 4122 variant + return uuid +} + +// NewMD5 returns a new MD5 (Version 3) UUID based on the +// supplied name space and data. It is the same as calling: +// +// NewHash(md5.New(), space, data, 3) +func NewMD5(space UUID, data []byte) UUID { + return NewHash(md5.New(), space, data, 3) +} + +// NewSHA1 returns a new SHA1 (Version 5) UUID based on the +// supplied name space and data. It is the same as calling: +// +// NewHash(sha1.New(), space, data, 5) +func NewSHA1(space UUID, data []byte) UUID { + return NewHash(sha1.New(), space, data, 5) +} diff --git a/vendor/github.com/google/uuid/marshal.go b/vendor/github.com/google/uuid/marshal.go new file mode 100644 index 0000000..14bd340 --- /dev/null +++ b/vendor/github.com/google/uuid/marshal.go @@ -0,0 +1,38 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import "fmt" + +// MarshalText implements encoding.TextMarshaler. +func (uuid UUID) MarshalText() ([]byte, error) { + var js [36]byte + encodeHex(js[:], uuid) + return js[:], nil +} + +// UnmarshalText implements encoding.TextUnmarshaler. +func (uuid *UUID) UnmarshalText(data []byte) error { + id, err := ParseBytes(data) + if err != nil { + return err + } + *uuid = id + return nil +} + +// MarshalBinary implements encoding.BinaryMarshaler. +func (uuid UUID) MarshalBinary() ([]byte, error) { + return uuid[:], nil +} + +// UnmarshalBinary implements encoding.BinaryUnmarshaler. +func (uuid *UUID) UnmarshalBinary(data []byte) error { + if len(data) != 16 { + return fmt.Errorf("invalid UUID (got %d bytes)", len(data)) + } + copy(uuid[:], data) + return nil +} diff --git a/vendor/github.com/google/uuid/node.go b/vendor/github.com/google/uuid/node.go new file mode 100644 index 0000000..d651a2b --- /dev/null +++ b/vendor/github.com/google/uuid/node.go @@ -0,0 +1,90 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "sync" +) + +var ( + nodeMu sync.Mutex + ifname string // name of interface being used + nodeID [6]byte // hardware for version 1 UUIDs + zeroID [6]byte // nodeID with only 0's +) + +// NodeInterface returns the name of the interface from which the NodeID was +// derived. The interface "user" is returned if the NodeID was set by +// SetNodeID. +func NodeInterface() string { + defer nodeMu.Unlock() + nodeMu.Lock() + return ifname +} + +// SetNodeInterface selects the hardware address to be used for Version 1 UUIDs. +// If name is "" then the first usable interface found will be used or a random +// Node ID will be generated. If a named interface cannot be found then false +// is returned. +// +// SetNodeInterface never fails when name is "". +func SetNodeInterface(name string) bool { + defer nodeMu.Unlock() + nodeMu.Lock() + return setNodeInterface(name) +} + +func setNodeInterface(name string) bool { + iname, addr := getHardwareInterface(name) // null implementation for js + if iname != "" && addr != nil { + ifname = iname + copy(nodeID[:], addr) + return true + } + + // We found no interfaces with a valid hardware address. If name + // does not specify a specific interface generate a random Node ID + // (section 4.1.6) + if name == "" { + ifname = "random" + randomBits(nodeID[:]) + return true + } + return false +} + +// NodeID returns a slice of a copy of the current Node ID, setting the Node ID +// if not already set. +func NodeID() []byte { + defer nodeMu.Unlock() + nodeMu.Lock() + if nodeID == zeroID { + setNodeInterface("") + } + nid := nodeID + return nid[:] +} + +// SetNodeID sets the Node ID to be used for Version 1 UUIDs. The first 6 bytes +// of id are used. If id is less than 6 bytes then false is returned and the +// Node ID is not set. +func SetNodeID(id []byte) bool { + if len(id) < 6 { + return false + } + defer nodeMu.Unlock() + nodeMu.Lock() + copy(nodeID[:], id) + ifname = "user" + return true +} + +// NodeID returns the 6 byte node id encoded in uuid. It returns nil if uuid is +// not valid. The NodeID is only well defined for version 1 and 2 UUIDs. +func (uuid UUID) NodeID() []byte { + var node [6]byte + copy(node[:], uuid[10:]) + return node[:] +} diff --git a/vendor/github.com/google/uuid/node_js.go b/vendor/github.com/google/uuid/node_js.go new file mode 100644 index 0000000..24b78ed --- /dev/null +++ b/vendor/github.com/google/uuid/node_js.go @@ -0,0 +1,12 @@ +// Copyright 2017 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build js + +package uuid + +// getHardwareInterface returns nil values for the JS version of the code. +// This remvoves the "net" dependency, because it is not used in the browser. +// Using the "net" library inflates the size of the transpiled JS code by 673k bytes. +func getHardwareInterface(name string) (string, []byte) { return "", nil } diff --git a/vendor/github.com/google/uuid/node_net.go b/vendor/github.com/google/uuid/node_net.go new file mode 100644 index 0000000..0cbbcdd --- /dev/null +++ b/vendor/github.com/google/uuid/node_net.go @@ -0,0 +1,33 @@ +// Copyright 2017 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !js + +package uuid + +import "net" + +var interfaces []net.Interface // cached list of interfaces + +// getHardwareInterface returns the name and hardware address of interface name. +// If name is "" then the name and hardware address of one of the system's +// interfaces is returned. If no interfaces are found (name does not exist or +// there are no interfaces) then "", nil is returned. +// +// Only addresses of at least 6 bytes are returned. +func getHardwareInterface(name string) (string, []byte) { + if interfaces == nil { + var err error + interfaces, err = net.Interfaces() + if err != nil { + return "", nil + } + } + for _, ifs := range interfaces { + if len(ifs.HardwareAddr) >= 6 && (name == "" || name == ifs.Name) { + return ifs.Name, ifs.HardwareAddr + } + } + return "", nil +} diff --git a/vendor/github.com/google/uuid/null.go b/vendor/github.com/google/uuid/null.go new file mode 100644 index 0000000..d7fcbf2 --- /dev/null +++ b/vendor/github.com/google/uuid/null.go @@ -0,0 +1,118 @@ +// Copyright 2021 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "bytes" + "database/sql/driver" + "encoding/json" + "fmt" +) + +var jsonNull = []byte("null") + +// NullUUID represents a UUID that may be null. +// NullUUID implements the SQL driver.Scanner interface so +// it can be used as a scan destination: +// +// var u uuid.NullUUID +// err := db.QueryRow("SELECT name FROM foo WHERE id=?", id).Scan(&u) +// ... +// if u.Valid { +// // use u.UUID +// } else { +// // NULL value +// } +// +type NullUUID struct { + UUID UUID + Valid bool // Valid is true if UUID is not NULL +} + +// Scan implements the SQL driver.Scanner interface. +func (nu *NullUUID) Scan(value interface{}) error { + if value == nil { + nu.UUID, nu.Valid = Nil, false + return nil + } + + err := nu.UUID.Scan(value) + if err != nil { + nu.Valid = false + return err + } + + nu.Valid = true + return nil +} + +// Value implements the driver Valuer interface. +func (nu NullUUID) Value() (driver.Value, error) { + if !nu.Valid { + return nil, nil + } + // Delegate to UUID Value function + return nu.UUID.Value() +} + +// MarshalBinary implements encoding.BinaryMarshaler. +func (nu NullUUID) MarshalBinary() ([]byte, error) { + if nu.Valid { + return nu.UUID[:], nil + } + + return []byte(nil), nil +} + +// UnmarshalBinary implements encoding.BinaryUnmarshaler. +func (nu *NullUUID) UnmarshalBinary(data []byte) error { + if len(data) != 16 { + return fmt.Errorf("invalid UUID (got %d bytes)", len(data)) + } + copy(nu.UUID[:], data) + nu.Valid = true + return nil +} + +// MarshalText implements encoding.TextMarshaler. +func (nu NullUUID) MarshalText() ([]byte, error) { + if nu.Valid { + return nu.UUID.MarshalText() + } + + return jsonNull, nil +} + +// UnmarshalText implements encoding.TextUnmarshaler. +func (nu *NullUUID) UnmarshalText(data []byte) error { + id, err := ParseBytes(data) + if err != nil { + nu.Valid = false + return err + } + nu.UUID = id + nu.Valid = true + return nil +} + +// MarshalJSON implements json.Marshaler. +func (nu NullUUID) MarshalJSON() ([]byte, error) { + if nu.Valid { + return json.Marshal(nu.UUID) + } + + return jsonNull, nil +} + +// UnmarshalJSON implements json.Unmarshaler. +func (nu *NullUUID) UnmarshalJSON(data []byte) error { + if bytes.Equal(data, jsonNull) { + *nu = NullUUID{} + return nil // valid null UUID + } + err := json.Unmarshal(data, &nu.UUID) + nu.Valid = err == nil + return err +} diff --git a/vendor/github.com/google/uuid/sql.go b/vendor/github.com/google/uuid/sql.go new file mode 100644 index 0000000..2e02ec0 --- /dev/null +++ b/vendor/github.com/google/uuid/sql.go @@ -0,0 +1,59 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "database/sql/driver" + "fmt" +) + +// Scan implements sql.Scanner so UUIDs can be read from databases transparently. +// Currently, database types that map to string and []byte are supported. Please +// consult database-specific driver documentation for matching types. +func (uuid *UUID) Scan(src interface{}) error { + switch src := src.(type) { + case nil: + return nil + + case string: + // if an empty UUID comes from a table, we return a null UUID + if src == "" { + return nil + } + + // see Parse for required string format + u, err := Parse(src) + if err != nil { + return fmt.Errorf("Scan: %v", err) + } + + *uuid = u + + case []byte: + // if an empty UUID comes from a table, we return a null UUID + if len(src) == 0 { + return nil + } + + // assumes a simple slice of bytes if 16 bytes + // otherwise attempts to parse + if len(src) != 16 { + return uuid.Scan(string(src)) + } + copy((*uuid)[:], src) + + default: + return fmt.Errorf("Scan: unable to scan type %T into UUID", src) + } + + return nil +} + +// Value implements sql.Valuer so that UUIDs can be written to databases +// transparently. Currently, UUIDs map to strings. Please consult +// database-specific driver documentation for matching types. +func (uuid UUID) Value() (driver.Value, error) { + return uuid.String(), nil +} diff --git a/vendor/github.com/google/uuid/time.go b/vendor/github.com/google/uuid/time.go new file mode 100644 index 0000000..e6ef06c --- /dev/null +++ b/vendor/github.com/google/uuid/time.go @@ -0,0 +1,123 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "encoding/binary" + "sync" + "time" +) + +// A Time represents a time as the number of 100's of nanoseconds since 15 Oct +// 1582. +type Time int64 + +const ( + lillian = 2299160 // Julian day of 15 Oct 1582 + unix = 2440587 // Julian day of 1 Jan 1970 + epoch = unix - lillian // Days between epochs + g1582 = epoch * 86400 // seconds between epochs + g1582ns100 = g1582 * 10000000 // 100s of a nanoseconds between epochs +) + +var ( + timeMu sync.Mutex + lasttime uint64 // last time we returned + clockSeq uint16 // clock sequence for this run + + timeNow = time.Now // for testing +) + +// UnixTime converts t the number of seconds and nanoseconds using the Unix +// epoch of 1 Jan 1970. +func (t Time) UnixTime() (sec, nsec int64) { + sec = int64(t - g1582ns100) + nsec = (sec % 10000000) * 100 + sec /= 10000000 + return sec, nsec +} + +// GetTime returns the current Time (100s of nanoseconds since 15 Oct 1582) and +// clock sequence as well as adjusting the clock sequence as needed. An error +// is returned if the current time cannot be determined. +func GetTime() (Time, uint16, error) { + defer timeMu.Unlock() + timeMu.Lock() + return getTime() +} + +func getTime() (Time, uint16, error) { + t := timeNow() + + // If we don't have a clock sequence already, set one. + if clockSeq == 0 { + setClockSequence(-1) + } + now := uint64(t.UnixNano()/100) + g1582ns100 + + // If time has gone backwards with this clock sequence then we + // increment the clock sequence + if now <= lasttime { + clockSeq = ((clockSeq + 1) & 0x3fff) | 0x8000 + } + lasttime = now + return Time(now), clockSeq, nil +} + +// ClockSequence returns the current clock sequence, generating one if not +// already set. The clock sequence is only used for Version 1 UUIDs. +// +// The uuid package does not use global static storage for the clock sequence or +// the last time a UUID was generated. Unless SetClockSequence is used, a new +// random clock sequence is generated the first time a clock sequence is +// requested by ClockSequence, GetTime, or NewUUID. (section 4.2.1.1) +func ClockSequence() int { + defer timeMu.Unlock() + timeMu.Lock() + return clockSequence() +} + +func clockSequence() int { + if clockSeq == 0 { + setClockSequence(-1) + } + return int(clockSeq & 0x3fff) +} + +// SetClockSequence sets the clock sequence to the lower 14 bits of seq. Setting to +// -1 causes a new sequence to be generated. +func SetClockSequence(seq int) { + defer timeMu.Unlock() + timeMu.Lock() + setClockSequence(seq) +} + +func setClockSequence(seq int) { + if seq == -1 { + var b [2]byte + randomBits(b[:]) // clock sequence + seq = int(b[0])<<8 | int(b[1]) + } + oldSeq := clockSeq + clockSeq = uint16(seq&0x3fff) | 0x8000 // Set our variant + if oldSeq != clockSeq { + lasttime = 0 + } +} + +// Time returns the time in 100s of nanoseconds since 15 Oct 1582 encoded in +// uuid. The time is only defined for version 1 and 2 UUIDs. +func (uuid UUID) Time() Time { + time := int64(binary.BigEndian.Uint32(uuid[0:4])) + time |= int64(binary.BigEndian.Uint16(uuid[4:6])) << 32 + time |= int64(binary.BigEndian.Uint16(uuid[6:8])&0xfff) << 48 + return Time(time) +} + +// ClockSequence returns the clock sequence encoded in uuid. +// The clock sequence is only well defined for version 1 and 2 UUIDs. +func (uuid UUID) ClockSequence() int { + return int(binary.BigEndian.Uint16(uuid[8:10])) & 0x3fff +} diff --git a/vendor/github.com/google/uuid/util.go b/vendor/github.com/google/uuid/util.go new file mode 100644 index 0000000..5ea6c73 --- /dev/null +++ b/vendor/github.com/google/uuid/util.go @@ -0,0 +1,43 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "io" +) + +// randomBits completely fills slice b with random data. +func randomBits(b []byte) { + if _, err := io.ReadFull(rander, b); err != nil { + panic(err.Error()) // rand should never fail + } +} + +// xvalues returns the value of a byte as a hexadecimal digit or 255. +var xvalues = [256]byte{ + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 255, 255, 255, 255, 255, 255, + 255, 10, 11, 12, 13, 14, 15, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 10, 11, 12, 13, 14, 15, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, + 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, +} + +// xtob converts hex characters x1 and x2 into a byte. +func xtob(x1, x2 byte) (byte, bool) { + b1 := xvalues[x1] + b2 := xvalues[x2] + return (b1 << 4) | b2, b1 != 255 && b2 != 255 +} diff --git a/vendor/github.com/google/uuid/uuid.go b/vendor/github.com/google/uuid/uuid.go new file mode 100644 index 0000000..a57207a --- /dev/null +++ b/vendor/github.com/google/uuid/uuid.go @@ -0,0 +1,294 @@ +// Copyright 2018 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "bytes" + "crypto/rand" + "encoding/hex" + "errors" + "fmt" + "io" + "strings" + "sync" +) + +// A UUID is a 128 bit (16 byte) Universal Unique IDentifier as defined in RFC +// 4122. +type UUID [16]byte + +// A Version represents a UUID's version. +type Version byte + +// A Variant represents a UUID's variant. +type Variant byte + +// Constants returned by Variant. +const ( + Invalid = Variant(iota) // Invalid UUID + RFC4122 // The variant specified in RFC4122 + Reserved // Reserved, NCS backward compatibility. + Microsoft // Reserved, Microsoft Corporation backward compatibility. + Future // Reserved for future definition. +) + +const randPoolSize = 16 * 16 + +var ( + rander = rand.Reader // random function + poolEnabled = false + poolMu sync.Mutex + poolPos = randPoolSize // protected with poolMu + pool [randPoolSize]byte // protected with poolMu +) + +type invalidLengthError struct{ len int } + +func (err invalidLengthError) Error() string { + return fmt.Sprintf("invalid UUID length: %d", err.len) +} + +// IsInvalidLengthError is matcher function for custom error invalidLengthError +func IsInvalidLengthError(err error) bool { + _, ok := err.(invalidLengthError) + return ok +} + +// Parse decodes s into a UUID or returns an error. Both the standard UUID +// forms of xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx and +// urn:uuid:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx are decoded as well as the +// Microsoft encoding {xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx} and the raw hex +// encoding: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx. +func Parse(s string) (UUID, error) { + var uuid UUID + switch len(s) { + // xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx + case 36: + + // urn:uuid:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx + case 36 + 9: + if strings.ToLower(s[:9]) != "urn:uuid:" { + return uuid, fmt.Errorf("invalid urn prefix: %q", s[:9]) + } + s = s[9:] + + // {xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx} + case 36 + 2: + s = s[1:] + + // xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx + case 32: + var ok bool + for i := range uuid { + uuid[i], ok = xtob(s[i*2], s[i*2+1]) + if !ok { + return uuid, errors.New("invalid UUID format") + } + } + return uuid, nil + default: + return uuid, invalidLengthError{len(s)} + } + // s is now at least 36 bytes long + // it must be of the form xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx + if s[8] != '-' || s[13] != '-' || s[18] != '-' || s[23] != '-' { + return uuid, errors.New("invalid UUID format") + } + for i, x := range [16]int{ + 0, 2, 4, 6, + 9, 11, + 14, 16, + 19, 21, + 24, 26, 28, 30, 32, 34} { + v, ok := xtob(s[x], s[x+1]) + if !ok { + return uuid, errors.New("invalid UUID format") + } + uuid[i] = v + } + return uuid, nil +} + +// ParseBytes is like Parse, except it parses a byte slice instead of a string. +func ParseBytes(b []byte) (UUID, error) { + var uuid UUID + switch len(b) { + case 36: // xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx + case 36 + 9: // urn:uuid:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx + if !bytes.Equal(bytes.ToLower(b[:9]), []byte("urn:uuid:")) { + return uuid, fmt.Errorf("invalid urn prefix: %q", b[:9]) + } + b = b[9:] + case 36 + 2: // {xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx} + b = b[1:] + case 32: // xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx + var ok bool + for i := 0; i < 32; i += 2 { + uuid[i/2], ok = xtob(b[i], b[i+1]) + if !ok { + return uuid, errors.New("invalid UUID format") + } + } + return uuid, nil + default: + return uuid, invalidLengthError{len(b)} + } + // s is now at least 36 bytes long + // it must be of the form xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx + if b[8] != '-' || b[13] != '-' || b[18] != '-' || b[23] != '-' { + return uuid, errors.New("invalid UUID format") + } + for i, x := range [16]int{ + 0, 2, 4, 6, + 9, 11, + 14, 16, + 19, 21, + 24, 26, 28, 30, 32, 34} { + v, ok := xtob(b[x], b[x+1]) + if !ok { + return uuid, errors.New("invalid UUID format") + } + uuid[i] = v + } + return uuid, nil +} + +// MustParse is like Parse but panics if the string cannot be parsed. +// It simplifies safe initialization of global variables holding compiled UUIDs. +func MustParse(s string) UUID { + uuid, err := Parse(s) + if err != nil { + panic(`uuid: Parse(` + s + `): ` + err.Error()) + } + return uuid +} + +// FromBytes creates a new UUID from a byte slice. Returns an error if the slice +// does not have a length of 16. The bytes are copied from the slice. +func FromBytes(b []byte) (uuid UUID, err error) { + err = uuid.UnmarshalBinary(b) + return uuid, err +} + +// Must returns uuid if err is nil and panics otherwise. +func Must(uuid UUID, err error) UUID { + if err != nil { + panic(err) + } + return uuid +} + +// String returns the string form of uuid, xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx +// , or "" if uuid is invalid. +func (uuid UUID) String() string { + var buf [36]byte + encodeHex(buf[:], uuid) + return string(buf[:]) +} + +// URN returns the RFC 2141 URN form of uuid, +// urn:uuid:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx, or "" if uuid is invalid. +func (uuid UUID) URN() string { + var buf [36 + 9]byte + copy(buf[:], "urn:uuid:") + encodeHex(buf[9:], uuid) + return string(buf[:]) +} + +func encodeHex(dst []byte, uuid UUID) { + hex.Encode(dst, uuid[:4]) + dst[8] = '-' + hex.Encode(dst[9:13], uuid[4:6]) + dst[13] = '-' + hex.Encode(dst[14:18], uuid[6:8]) + dst[18] = '-' + hex.Encode(dst[19:23], uuid[8:10]) + dst[23] = '-' + hex.Encode(dst[24:], uuid[10:]) +} + +// Variant returns the variant encoded in uuid. +func (uuid UUID) Variant() Variant { + switch { + case (uuid[8] & 0xc0) == 0x80: + return RFC4122 + case (uuid[8] & 0xe0) == 0xc0: + return Microsoft + case (uuid[8] & 0xe0) == 0xe0: + return Future + default: + return Reserved + } +} + +// Version returns the version of uuid. +func (uuid UUID) Version() Version { + return Version(uuid[6] >> 4) +} + +func (v Version) String() string { + if v > 15 { + return fmt.Sprintf("BAD_VERSION_%d", v) + } + return fmt.Sprintf("VERSION_%d", v) +} + +func (v Variant) String() string { + switch v { + case RFC4122: + return "RFC4122" + case Reserved: + return "Reserved" + case Microsoft: + return "Microsoft" + case Future: + return "Future" + case Invalid: + return "Invalid" + } + return fmt.Sprintf("BadVariant%d", int(v)) +} + +// SetRand sets the random number generator to r, which implements io.Reader. +// If r.Read returns an error when the package requests random data then +// a panic will be issued. +// +// Calling SetRand with nil sets the random number generator to the default +// generator. +func SetRand(r io.Reader) { + if r == nil { + rander = rand.Reader + return + } + rander = r +} + +// EnableRandPool enables internal randomness pool used for Random +// (Version 4) UUID generation. The pool contains random bytes read from +// the random number generator on demand in batches. Enabling the pool +// may improve the UUID generation throughput significantly. +// +// Since the pool is stored on the Go heap, this feature may be a bad fit +// for security sensitive applications. +// +// Both EnableRandPool and DisableRandPool are not thread-safe and should +// only be called when there is no possibility that New or any other +// UUID Version 4 generation function will be called concurrently. +func EnableRandPool() { + poolEnabled = true +} + +// DisableRandPool disables the randomness pool if it was previously +// enabled with EnableRandPool. +// +// Both EnableRandPool and DisableRandPool are not thread-safe and should +// only be called when there is no possibility that New or any other +// UUID Version 4 generation function will be called concurrently. +func DisableRandPool() { + poolEnabled = false + defer poolMu.Unlock() + poolMu.Lock() + poolPos = randPoolSize +} diff --git a/vendor/github.com/google/uuid/version1.go b/vendor/github.com/google/uuid/version1.go new file mode 100644 index 0000000..4631096 --- /dev/null +++ b/vendor/github.com/google/uuid/version1.go @@ -0,0 +1,44 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import ( + "encoding/binary" +) + +// NewUUID returns a Version 1 UUID based on the current NodeID and clock +// sequence, and the current time. If the NodeID has not been set by SetNodeID +// or SetNodeInterface then it will be set automatically. If the NodeID cannot +// be set NewUUID returns nil. If clock sequence has not been set by +// SetClockSequence then it will be set automatically. If GetTime fails to +// return the current NewUUID returns nil and an error. +// +// In most cases, New should be used. +func NewUUID() (UUID, error) { + var uuid UUID + now, seq, err := GetTime() + if err != nil { + return uuid, err + } + + timeLow := uint32(now & 0xffffffff) + timeMid := uint16((now >> 32) & 0xffff) + timeHi := uint16((now >> 48) & 0x0fff) + timeHi |= 0x1000 // Version 1 + + binary.BigEndian.PutUint32(uuid[0:], timeLow) + binary.BigEndian.PutUint16(uuid[4:], timeMid) + binary.BigEndian.PutUint16(uuid[6:], timeHi) + binary.BigEndian.PutUint16(uuid[8:], seq) + + nodeMu.Lock() + if nodeID == zeroID { + setNodeInterface("") + } + copy(uuid[10:], nodeID[:]) + nodeMu.Unlock() + + return uuid, nil +} diff --git a/vendor/github.com/google/uuid/version4.go b/vendor/github.com/google/uuid/version4.go new file mode 100644 index 0000000..7697802 --- /dev/null +++ b/vendor/github.com/google/uuid/version4.go @@ -0,0 +1,76 @@ +// Copyright 2016 Google Inc. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package uuid + +import "io" + +// New creates a new random UUID or panics. New is equivalent to +// the expression +// +// uuid.Must(uuid.NewRandom()) +func New() UUID { + return Must(NewRandom()) +} + +// NewString creates a new random UUID and returns it as a string or panics. +// NewString is equivalent to the expression +// +// uuid.New().String() +func NewString() string { + return Must(NewRandom()).String() +} + +// NewRandom returns a Random (Version 4) UUID. +// +// The strength of the UUIDs is based on the strength of the crypto/rand +// package. +// +// Uses the randomness pool if it was enabled with EnableRandPool. +// +// A note about uniqueness derived from the UUID Wikipedia entry: +// +// Randomly generated UUIDs have 122 random bits. One's annual risk of being +// hit by a meteorite is estimated to be one chance in 17 billion, that +// means the probability is about 0.00000000006 (6 × 10−11), +// equivalent to the odds of creating a few tens of trillions of UUIDs in a +// year and having one duplicate. +func NewRandom() (UUID, error) { + if !poolEnabled { + return NewRandomFromReader(rander) + } + return newRandomFromPool() +} + +// NewRandomFromReader returns a UUID based on bytes read from a given io.Reader. +func NewRandomFromReader(r io.Reader) (UUID, error) { + var uuid UUID + _, err := io.ReadFull(r, uuid[:]) + if err != nil { + return Nil, err + } + uuid[6] = (uuid[6] & 0x0f) | 0x40 // Version 4 + uuid[8] = (uuid[8] & 0x3f) | 0x80 // Variant is 10 + return uuid, nil +} + +func newRandomFromPool() (UUID, error) { + var uuid UUID + poolMu.Lock() + if poolPos == randPoolSize { + _, err := io.ReadFull(rander, pool[:]) + if err != nil { + poolMu.Unlock() + return Nil, err + } + poolPos = 0 + } + copy(uuid[:], pool[poolPos:(poolPos+16)]) + poolPos += 16 + poolMu.Unlock() + + uuid[6] = (uuid[6] & 0x0f) | 0x40 // Version 4 + uuid[8] = (uuid[8] & 0x3f) | 0x80 // Variant is 10 + return uuid, nil +} diff --git a/vendor/github.com/klauspost/compress/.gitattributes b/vendor/github.com/klauspost/compress/.gitattributes new file mode 100644 index 0000000..4024335 --- /dev/null +++ b/vendor/github.com/klauspost/compress/.gitattributes @@ -0,0 +1,2 @@ +* -text +*.bin -text -diff diff --git a/vendor/github.com/klauspost/compress/.gitignore b/vendor/github.com/klauspost/compress/.gitignore new file mode 100644 index 0000000..b35f844 --- /dev/null +++ b/vendor/github.com/klauspost/compress/.gitignore @@ -0,0 +1,25 @@ +# Compiled Object files, Static and Dynamic libs (Shared Objects) +*.o +*.a +*.so + +# Folders +_obj +_test + +# Architecture specific extensions/prefixes +*.[568vq] +[568vq].out + +*.cgo1.go +*.cgo2.c +_cgo_defun.c +_cgo_gotypes.go +_cgo_export.* + +_testmain.go + +*.exe +*.test +*.prof +/s2/cmd/_s2sx/sfx-exe diff --git a/vendor/github.com/klauspost/compress/.goreleaser.yml b/vendor/github.com/klauspost/compress/.goreleaser.yml new file mode 100644 index 0000000..0af08e6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/.goreleaser.yml @@ -0,0 +1,141 @@ +# This is an example goreleaser.yaml file with some sane defaults. +# Make sure to check the documentation at http://goreleaser.com +before: + hooks: + - ./gen.sh + - go install mvdan.cc/garble@latest + +builds: + - + id: "s2c" + binary: s2c + main: ./s2/cmd/s2c/main.go + flags: + - -trimpath + env: + - CGO_ENABLED=0 + goos: + - aix + - linux + - freebsd + - netbsd + - windows + - darwin + goarch: + - 386 + - amd64 + - arm + - arm64 + - ppc64 + - ppc64le + - mips64 + - mips64le + goarm: + - 7 + gobinary: garble + - + id: "s2d" + binary: s2d + main: ./s2/cmd/s2d/main.go + flags: + - -trimpath + env: + - CGO_ENABLED=0 + goos: + - aix + - linux + - freebsd + - netbsd + - windows + - darwin + goarch: + - 386 + - amd64 + - arm + - arm64 + - ppc64 + - ppc64le + - mips64 + - mips64le + goarm: + - 7 + gobinary: garble + - + id: "s2sx" + binary: s2sx + main: ./s2/cmd/_s2sx/main.go + flags: + - -modfile=s2sx.mod + - -trimpath + env: + - CGO_ENABLED=0 + goos: + - aix + - linux + - freebsd + - netbsd + - windows + - darwin + goarch: + - 386 + - amd64 + - arm + - arm64 + - ppc64 + - ppc64le + - mips64 + - mips64le + goarm: + - 7 + gobinary: garble + +archives: + - + id: s2-binaries + name_template: "s2-{{ .Os }}_{{ .Arch }}_{{ .Version }}" + replacements: + aix: AIX + darwin: OSX + linux: Linux + windows: Windows + 386: i386 + amd64: x86_64 + freebsd: FreeBSD + netbsd: NetBSD + format_overrides: + - goos: windows + format: zip + files: + - unpack/* + - s2/LICENSE + - s2/README.md +checksum: + name_template: 'checksums.txt' +snapshot: + name_template: "{{ .Tag }}-next" +changelog: + sort: asc + filters: + exclude: + - '^doc:' + - '^docs:' + - '^test:' + - '^tests:' + - '^Update\sREADME.md' + +nfpms: + - + file_name_template: "s2_package_{{ .Version }}_{{ .Os }}_{{ .Arch }}" + vendor: Klaus Post + homepage: https://github.com/klauspost/compress + maintainer: Klaus Post + description: S2 Compression Tool + license: BSD 3-Clause + formats: + - deb + - rpm + replacements: + darwin: Darwin + linux: Linux + freebsd: FreeBSD + amd64: x86_64 diff --git a/vendor/github.com/klauspost/compress/LICENSE b/vendor/github.com/klauspost/compress/LICENSE new file mode 100644 index 0000000..87d5574 --- /dev/null +++ b/vendor/github.com/klauspost/compress/LICENSE @@ -0,0 +1,304 @@ +Copyright (c) 2012 The Go Authors. All rights reserved. +Copyright (c) 2019 Klaus Post. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +------------------ + +Files: gzhttp/* + + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright 2016-2017 The New York Times Company + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. + +------------------ + +Files: s2/cmd/internal/readahead/* + +The MIT License (MIT) + +Copyright (c) 2015 Klaus Post + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. + +--------------------- +Files: snappy/* +Files: internal/snapref/* + +Copyright (c) 2011 The Snappy-Go Authors. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +----------------- + +Files: s2/cmd/internal/filepathx/* + +Copyright 2016 The filepathx Authors + +Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. diff --git a/vendor/github.com/klauspost/compress/README.md b/vendor/github.com/klauspost/compress/README.md new file mode 100644 index 0000000..0e2dc11 --- /dev/null +++ b/vendor/github.com/klauspost/compress/README.md @@ -0,0 +1,489 @@ +# compress + +This package provides various compression algorithms. + +* [zstandard](https://github.com/klauspost/compress/tree/master/zstd#zstd) compression and decompression in pure Go. +* [S2](https://github.com/klauspost/compress/tree/master/s2#s2-compression) is a high performance replacement for Snappy. +* Optimized [deflate](https://godoc.org/github.com/klauspost/compress/flate) packages which can be used as a dropin replacement for [gzip](https://godoc.org/github.com/klauspost/compress/gzip), [zip](https://godoc.org/github.com/klauspost/compress/zip) and [zlib](https://godoc.org/github.com/klauspost/compress/zlib). +* [snappy](https://github.com/klauspost/compress/tree/master/snappy) is a drop-in replacement for `github.com/golang/snappy` offering better compression and concurrent streams. +* [huff0](https://github.com/klauspost/compress/tree/master/huff0) and [FSE](https://github.com/klauspost/compress/tree/master/fse) implementations for raw entropy encoding. +* [gzhttp](https://github.com/klauspost/compress/tree/master/gzhttp) Provides client and server wrappers for handling gzipped requests efficiently. +* [pgzip](https://github.com/klauspost/pgzip) is a separate package that provides a very fast parallel gzip implementation. +* [fuzz package](https://github.com/klauspost/compress-fuzz) for fuzz testing all compressors/decompressors here. + +[![Go Reference](https://pkg.go.dev/badge/klauspost/compress.svg)](https://pkg.go.dev/github.com/klauspost/compress?tab=subdirectories) +[![Go](https://github.com/klauspost/compress/actions/workflows/go.yml/badge.svg)](https://github.com/klauspost/compress/actions/workflows/go.yml) +[![Sourcegraph Badge](https://sourcegraph.com/github.com/klauspost/compress/-/badge.svg)](https://sourcegraph.com/github.com/klauspost/compress?badge) + +# changelog + +* Mar 3, 2022 (v1.15.0) + * zstd: Refactor decoder by @klauspost in [#498](https://github.com/klauspost/compress/pull/498) + * zstd: Add stream encoding without goroutines by @klauspost in [#505](https://github.com/klauspost/compress/pull/505) + * huff0: Prevent single blocks exceeding 16 bits by @klauspost in[#507](https://github.com/klauspost/compress/pull/507) + * flate: Inline literal emission by @klauspost in [#509](https://github.com/klauspost/compress/pull/509) + * gzhttp: Add zstd to transport by @klauspost in [#400](https://github.com/klauspost/compress/pull/400) + * gzhttp: Make content-type optional by @klauspost in [#510](https://github.com/klauspost/compress/pull/510) + +
+ See Details +Both compression and decompression now supports "synchronous" stream operations. This means that whenever "concurrency" is set to 1, they will operate without spawning goroutines. + +Stream decompression is now faster on asynchronous, since the goroutine allocation much more effectively splits the workload. On typical streams this will typically use 2 cores fully for decompression. When a stream has finished decoding no goroutines will be left over, so decoders can now safely be pooled and still be garbage collected. + +While the release has been extensively tested, it is recommended to testing when upgrading. +
+ +* Feb 22, 2022 (v1.14.4) + * flate: Fix rare huffman only (-2) corruption. [#503](https://github.com/klauspost/compress/pull/503) + * zip: Update deprecated CreateHeaderRaw to correctly call CreateRaw by @saracen in [#502](https://github.com/klauspost/compress/pull/502) + * zip: don't read data descriptor early by @saracen in [#501](https://github.com/klauspost/compress/pull/501) #501 + * huff0: Use static decompression buffer up to 30% faster by @klauspost in [#499](https://github.com/klauspost/compress/pull/499) [#500](https://github.com/klauspost/compress/pull/500) + +* Feb 17, 2022 (v1.14.3) + * flate: Improve fastest levels compression speed ~10% more throughput. [#482](https://github.com/klauspost/compress/pull/482) [#489](https://github.com/klauspost/compress/pull/489) [#490](https://github.com/klauspost/compress/pull/490) [#491](https://github.com/klauspost/compress/pull/491) [#494](https://github.com/klauspost/compress/pull/494) [#478](https://github.com/klauspost/compress/pull/478) + * flate: Faster decompression speed, ~5-10%. [#483](https://github.com/klauspost/compress/pull/483) + * s2: Faster compression with Go v1.18 and amd64 microarch level 3+. [#484](https://github.com/klauspost/compress/pull/484) [#486](https://github.com/klauspost/compress/pull/486) + +* Jan 25, 2022 (v1.14.2) + * zstd: improve header decoder by @dsnet [#476](https://github.com/klauspost/compress/pull/476) + * zstd: Add bigger default blocks [#469](https://github.com/klauspost/compress/pull/469) + * zstd: Remove unused decompression buffer [#470](https://github.com/klauspost/compress/pull/470) + * zstd: Fix logically dead code by @ningmingxiao [#472](https://github.com/klauspost/compress/pull/472) + * flate: Improve level 7-9 [#471](https://github.com/klauspost/compress/pull/471) [#473](https://github.com/klauspost/compress/pull/473) + * zstd: Add noasm tag for xxhash [#475](https://github.com/klauspost/compress/pull/475) + +* Jan 11, 2022 (v1.14.1) + * s2: Add stream index in [#462](https://github.com/klauspost/compress/pull/462) + * flate: Speed and efficiency improvements in [#439](https://github.com/klauspost/compress/pull/439) [#461](https://github.com/klauspost/compress/pull/461) [#455](https://github.com/klauspost/compress/pull/455) [#452](https://github.com/klauspost/compress/pull/452) [#458](https://github.com/klauspost/compress/pull/458) + * zstd: Performance improvement in [#420]( https://github.com/klauspost/compress/pull/420) [#456](https://github.com/klauspost/compress/pull/456) [#437](https://github.com/klauspost/compress/pull/437) [#467](https://github.com/klauspost/compress/pull/467) [#468](https://github.com/klauspost/compress/pull/468) + * zstd: add arm64 xxhash assembly in [#464](https://github.com/klauspost/compress/pull/464) + * Add garbled for binaries for s2 in [#445](https://github.com/klauspost/compress/pull/445) + +* Aug 30, 2021 (v1.13.5) + * gz/zlib/flate: Alias stdlib errors [#425](https://github.com/klauspost/compress/pull/425) + * s2: Add block support to commandline tools [#413](https://github.com/klauspost/compress/pull/413) + * zstd: pooledZipWriter should return Writers to the same pool [#426](https://github.com/klauspost/compress/pull/426) + * Removed golang/snappy as external dependency for tests [#421](https://github.com/klauspost/compress/pull/421) + +* Aug 12, 2021 (v1.13.4) + * Add [snappy replacement package](https://github.com/klauspost/compress/tree/master/snappy). + * zstd: Fix incorrect encoding in "best" mode [#415](https://github.com/klauspost/compress/pull/415) + +* Aug 3, 2021 (v1.13.3) + * zstd: Improve Best compression [#404](https://github.com/klauspost/compress/pull/404) + * zstd: Fix WriteTo error forwarding [#411](https://github.com/klauspost/compress/pull/411) + * gzhttp: Return http.HandlerFunc instead of http.Handler. Unlikely breaking change. [#406](https://github.com/klauspost/compress/pull/406) + * s2sx: Fix max size error [#399](https://github.com/klauspost/compress/pull/399) + * zstd: Add optional stream content size on reset [#401](https://github.com/klauspost/compress/pull/401) + * zstd: use SpeedBestCompression for level >= 10 [#410](https://github.com/klauspost/compress/pull/410) + +* Jun 14, 2021 (v1.13.1) + * s2: Add full Snappy output support [#396](https://github.com/klauspost/compress/pull/396) + * zstd: Add configurable [Decoder window](https://pkg.go.dev/github.com/klauspost/compress/zstd#WithDecoderMaxWindow) size [#394](https://github.com/klauspost/compress/pull/394) + * gzhttp: Add header to skip compression [#389](https://github.com/klauspost/compress/pull/389) + * s2: Improve speed with bigger output margin [#395](https://github.com/klauspost/compress/pull/395) + +* Jun 3, 2021 (v1.13.0) + * Added [gzhttp](https://github.com/klauspost/compress/tree/master/gzhttp#gzip-handler) which allows wrapping HTTP servers and clients with GZIP compressors. + * zstd: Detect short invalid signatures [#382](https://github.com/klauspost/compress/pull/382) + * zstd: Spawn decoder goroutine only if needed. [#380](https://github.com/klauspost/compress/pull/380) + +
+ See changes to v1.12.x + +* May 25, 2021 (v1.12.3) + * deflate: Better/faster Huffman encoding [#374](https://github.com/klauspost/compress/pull/374) + * deflate: Allocate less for history. [#375](https://github.com/klauspost/compress/pull/375) + * zstd: Forward read errors [#373](https://github.com/klauspost/compress/pull/373) + +* Apr 27, 2021 (v1.12.2) + * zstd: Improve better/best compression [#360](https://github.com/klauspost/compress/pull/360) [#364](https://github.com/klauspost/compress/pull/364) [#365](https://github.com/klauspost/compress/pull/365) + * zstd: Add helpers to compress/decompress zstd inside zip files [#363](https://github.com/klauspost/compress/pull/363) + * deflate: Improve level 5+6 compression [#367](https://github.com/klauspost/compress/pull/367) + * s2: Improve better/best compression [#358](https://github.com/klauspost/compress/pull/358) [#359](https://github.com/klauspost/compress/pull/358) + * s2: Load after checking src limit on amd64. [#362](https://github.com/klauspost/compress/pull/362) + * s2sx: Limit max executable size [#368](https://github.com/klauspost/compress/pull/368) + +* Apr 14, 2021 (v1.12.1) + * snappy package removed. Upstream added as dependency. + * s2: Better compression in "best" mode [#353](https://github.com/klauspost/compress/pull/353) + * s2sx: Add stdin input and detect pre-compressed from signature [#352](https://github.com/klauspost/compress/pull/352) + * s2c/s2d: Add http as possible input [#348](https://github.com/klauspost/compress/pull/348) + * s2c/s2d/s2sx: Always truncate when writing files [#352](https://github.com/klauspost/compress/pull/352) + * zstd: Reduce memory usage further when using [WithLowerEncoderMem](https://pkg.go.dev/github.com/klauspost/compress/zstd#WithLowerEncoderMem) [#346](https://github.com/klauspost/compress/pull/346) + * s2: Fix potential problem with amd64 assembly and profilers [#349](https://github.com/klauspost/compress/pull/349) +
+ +
+ See changes to v1.11.x + +* Mar 26, 2021 (v1.11.13) + * zstd: Big speedup on small dictionary encodes [#344](https://github.com/klauspost/compress/pull/344) [#345](https://github.com/klauspost/compress/pull/345) + * zstd: Add [WithLowerEncoderMem](https://pkg.go.dev/github.com/klauspost/compress/zstd#WithLowerEncoderMem) encoder option [#336](https://github.com/klauspost/compress/pull/336) + * deflate: Improve entropy compression [#338](https://github.com/klauspost/compress/pull/338) + * s2: Clean up and minor performance improvement in best [#341](https://github.com/klauspost/compress/pull/341) + +* Mar 5, 2021 (v1.11.12) + * s2: Add `s2sx` binary that creates [self extracting archives](https://github.com/klauspost/compress/tree/master/s2#s2sx-self-extracting-archives). + * s2: Speed up decompression on non-assembly platforms [#328](https://github.com/klauspost/compress/pull/328) + +* Mar 1, 2021 (v1.11.9) + * s2: Add ARM64 decompression assembly. Around 2x output speed. [#324](https://github.com/klauspost/compress/pull/324) + * s2: Improve "better" speed and efficiency. [#325](https://github.com/klauspost/compress/pull/325) + * s2: Fix binaries. + +* Feb 25, 2021 (v1.11.8) + * s2: Fixed occational out-of-bounds write on amd64. Upgrade recommended. + * s2: Add AMD64 assembly for better mode. 25-50% faster. [#315](https://github.com/klauspost/compress/pull/315) + * s2: Less upfront decoder allocation. [#322](https://github.com/klauspost/compress/pull/322) + * zstd: Faster "compression" of incompressible data. [#314](https://github.com/klauspost/compress/pull/314) + * zip: Fix zip64 headers. [#313](https://github.com/klauspost/compress/pull/313) + +* Jan 14, 2021 (v1.11.7) + * Use Bytes() interface to get bytes across packages. [#309](https://github.com/klauspost/compress/pull/309) + * s2: Add 'best' compression option. [#310](https://github.com/klauspost/compress/pull/310) + * s2: Add ReaderMaxBlockSize, changes `s2.NewReader` signature to include varargs. [#311](https://github.com/klauspost/compress/pull/311) + * s2: Fix crash on small better buffers. [#308](https://github.com/klauspost/compress/pull/308) + * s2: Clean up decoder. [#312](https://github.com/klauspost/compress/pull/312) + +* Jan 7, 2021 (v1.11.6) + * zstd: Make decoder allocations smaller [#306](https://github.com/klauspost/compress/pull/306) + * zstd: Free Decoder resources when Reset is called with a nil io.Reader [#305](https://github.com/klauspost/compress/pull/305) + +* Dec 20, 2020 (v1.11.4) + * zstd: Add Best compression mode [#304](https://github.com/klauspost/compress/pull/304) + * Add header decoder [#299](https://github.com/klauspost/compress/pull/299) + * s2: Add uncompressed stream option [#297](https://github.com/klauspost/compress/pull/297) + * Simplify/speed up small blocks with known max size. [#300](https://github.com/klauspost/compress/pull/300) + * zstd: Always reset literal dict encoder [#303](https://github.com/klauspost/compress/pull/303) + +* Nov 15, 2020 (v1.11.3) + * inflate: 10-15% faster decompression [#293](https://github.com/klauspost/compress/pull/293) + * zstd: Tweak DecodeAll default allocation [#295](https://github.com/klauspost/compress/pull/295) + +* Oct 11, 2020 (v1.11.2) + * s2: Fix out of bounds read in "better" block compression [#291](https://github.com/klauspost/compress/pull/291) + +* Oct 1, 2020 (v1.11.1) + * zstd: Set allLitEntropy true in default configuration [#286](https://github.com/klauspost/compress/pull/286) + +* Sept 8, 2020 (v1.11.0) + * zstd: Add experimental compression [dictionaries](https://github.com/klauspost/compress/tree/master/zstd#dictionaries) [#281](https://github.com/klauspost/compress/pull/281) + * zstd: Fix mixed Write and ReadFrom calls [#282](https://github.com/klauspost/compress/pull/282) + * inflate/gz: Limit variable shifts, ~5% faster decompression [#274](https://github.com/klauspost/compress/pull/274) +
+ +
+ See changes to v1.10.x + +* July 8, 2020 (v1.10.11) + * zstd: Fix extra block when compressing with ReadFrom. [#278](https://github.com/klauspost/compress/pull/278) + * huff0: Also populate compression table when reading decoding table. [#275](https://github.com/klauspost/compress/pull/275) + +* June 23, 2020 (v1.10.10) + * zstd: Skip entropy compression in fastest mode when no matches. [#270](https://github.com/klauspost/compress/pull/270) + +* June 16, 2020 (v1.10.9): + * zstd: API change for specifying dictionaries. See [#268](https://github.com/klauspost/compress/pull/268) + * zip: update CreateHeaderRaw to handle zip64 fields. [#266](https://github.com/klauspost/compress/pull/266) + * Fuzzit tests removed. The service has been purchased and is no longer available. + +* June 5, 2020 (v1.10.8): + * 1.15x faster zstd block decompression. [#265](https://github.com/klauspost/compress/pull/265) + +* June 1, 2020 (v1.10.7): + * Added zstd decompression [dictionary support](https://github.com/klauspost/compress/tree/master/zstd#dictionaries) + * Increase zstd decompression speed up to 1.19x. [#259](https://github.com/klauspost/compress/pull/259) + * Remove internal reset call in zstd compression and reduce allocations. [#263](https://github.com/klauspost/compress/pull/263) + +* May 21, 2020: (v1.10.6) + * zstd: Reduce allocations while decoding. [#258](https://github.com/klauspost/compress/pull/258), [#252](https://github.com/klauspost/compress/pull/252) + * zstd: Stricter decompression checks. + +* April 12, 2020: (v1.10.5) + * s2-commands: Flush output when receiving SIGINT. [#239](https://github.com/klauspost/compress/pull/239) + +* Apr 8, 2020: (v1.10.4) + * zstd: Minor/special case optimizations. [#251](https://github.com/klauspost/compress/pull/251), [#250](https://github.com/klauspost/compress/pull/250), [#249](https://github.com/klauspost/compress/pull/249), [#247](https://github.com/klauspost/compress/pull/247) +* Mar 11, 2020: (v1.10.3) + * s2: Use S2 encoder in pure Go mode for Snappy output as well. [#245](https://github.com/klauspost/compress/pull/245) + * s2: Fix pure Go block encoder. [#244](https://github.com/klauspost/compress/pull/244) + * zstd: Added "better compression" mode. [#240](https://github.com/klauspost/compress/pull/240) + * zstd: Improve speed of fastest compression mode by 5-10% [#241](https://github.com/klauspost/compress/pull/241) + * zstd: Skip creating encoders when not needed. [#238](https://github.com/klauspost/compress/pull/238) + +* Feb 27, 2020: (v1.10.2) + * Close to 50% speedup in inflate (gzip/zip decompression). [#236](https://github.com/klauspost/compress/pull/236) [#234](https://github.com/klauspost/compress/pull/234) [#232](https://github.com/klauspost/compress/pull/232) + * Reduce deflate level 1-6 memory usage up to 59%. [#227](https://github.com/klauspost/compress/pull/227) + +* Feb 18, 2020: (v1.10.1) + * Fix zstd crash when resetting multiple times without sending data. [#226](https://github.com/klauspost/compress/pull/226) + * deflate: Fix dictionary use on level 1-6. [#224](https://github.com/klauspost/compress/pull/224) + * Remove deflate writer reference when closing. [#224](https://github.com/klauspost/compress/pull/224) + +* Feb 4, 2020: (v1.10.0) + * Add optional dictionary to [stateless deflate](https://pkg.go.dev/github.com/klauspost/compress/flate?tab=doc#StatelessDeflate). Breaking change, send `nil` for previous behaviour. [#216](https://github.com/klauspost/compress/pull/216) + * Fix buffer overflow on repeated small block deflate. [#218](https://github.com/klauspost/compress/pull/218) + * Allow copying content from an existing ZIP file without decompressing+compressing. [#214](https://github.com/klauspost/compress/pull/214) + * Added [S2](https://github.com/klauspost/compress/tree/master/s2#s2-compression) AMD64 assembler and various optimizations. Stream speed >10GB/s. [#186](https://github.com/klauspost/compress/pull/186) + +
+ +
+ See changes prior to v1.10.0 + +* Jan 20,2020 (v1.9.8) Optimize gzip/deflate with better size estimates and faster table generation. [#207](https://github.com/klauspost/compress/pull/207) by [luyu6056](https://github.com/luyu6056), [#206](https://github.com/klauspost/compress/pull/206). +* Jan 11, 2020: S2 Encode/Decode will use provided buffer if capacity is big enough. [#204](https://github.com/klauspost/compress/pull/204) +* Jan 5, 2020: (v1.9.7) Fix another zstd regression in v1.9.5 - v1.9.6 removed. +* Jan 4, 2020: (v1.9.6) Regression in v1.9.5 fixed causing corrupt zstd encodes in rare cases. +* Jan 4, 2020: Faster IO in [s2c + s2d commandline tools](https://github.com/klauspost/compress/tree/master/s2#commandline-tools) compression/decompression. [#192](https://github.com/klauspost/compress/pull/192) +* Dec 29, 2019: Removed v1.9.5 since fuzz tests showed a compatibility problem with the reference zstandard decoder. +* Dec 29, 2019: (v1.9.5) zstd: 10-20% faster block compression. [#199](https://github.com/klauspost/compress/pull/199) +* Dec 29, 2019: [zip](https://godoc.org/github.com/klauspost/compress/zip) package updated with latest Go features +* Dec 29, 2019: zstd: Single segment flag condintions tweaked. [#197](https://github.com/klauspost/compress/pull/197) +* Dec 18, 2019: s2: Faster compression when ReadFrom is used. [#198](https://github.com/klauspost/compress/pull/198) +* Dec 10, 2019: s2: Fix repeat length output when just above at 16MB limit. +* Dec 10, 2019: zstd: Add function to get decoder as io.ReadCloser. [#191](https://github.com/klauspost/compress/pull/191) +* Dec 3, 2019: (v1.9.4) S2: limit max repeat length. [#188](https://github.com/klauspost/compress/pull/188) +* Dec 3, 2019: Add [WithNoEntropyCompression](https://godoc.org/github.com/klauspost/compress/zstd#WithNoEntropyCompression) to zstd [#187](https://github.com/klauspost/compress/pull/187) +* Dec 3, 2019: Reduce memory use for tests. Check for leaked goroutines. +* Nov 28, 2019 (v1.9.3) Less allocations in stateless deflate. +* Nov 28, 2019: 5-20% Faster huff0 decode. Impacts zstd as well. [#184](https://github.com/klauspost/compress/pull/184) +* Nov 12, 2019 (v1.9.2) Added [Stateless Compression](#stateless-compression) for gzip/deflate. +* Nov 12, 2019: Fixed zstd decompression of large single blocks. [#180](https://github.com/klauspost/compress/pull/180) +* Nov 11, 2019: Set default [s2c](https://github.com/klauspost/compress/tree/master/s2#commandline-tools) block size to 4MB. +* Nov 11, 2019: Reduce inflate memory use by 1KB. +* Nov 10, 2019: Less allocations in deflate bit writer. +* Nov 10, 2019: Fix inconsistent error returned by zstd decoder. +* Oct 28, 2019 (v1.9.1) ztsd: Fix crash when compressing blocks. [#174](https://github.com/klauspost/compress/pull/174) +* Oct 24, 2019 (v1.9.0) zstd: Fix rare data corruption [#173](https://github.com/klauspost/compress/pull/173) +* Oct 24, 2019 zstd: Fix huff0 out of buffer write [#171](https://github.com/klauspost/compress/pull/171) and always return errors [#172](https://github.com/klauspost/compress/pull/172) +* Oct 10, 2019: Big deflate rewrite, 30-40% faster with better compression [#105](https://github.com/klauspost/compress/pull/105) + +
+ +
+ See changes prior to v1.9.0 + +* Oct 10, 2019: (v1.8.6) zstd: Allow partial reads to get flushed data. [#169](https://github.com/klauspost/compress/pull/169) +* Oct 3, 2019: Fix inconsistent results on broken zstd streams. +* Sep 25, 2019: Added `-rm` (remove source files) and `-q` (no output except errors) to `s2c` and `s2d` [commands](https://github.com/klauspost/compress/tree/master/s2#commandline-tools) +* Sep 16, 2019: (v1.8.4) Add `s2c` and `s2d` [commandline tools](https://github.com/klauspost/compress/tree/master/s2#commandline-tools). +* Sep 10, 2019: (v1.8.3) Fix s2 decoder [Skip](https://godoc.org/github.com/klauspost/compress/s2#Reader.Skip). +* Sep 7, 2019: zstd: Added [WithWindowSize](https://godoc.org/github.com/klauspost/compress/zstd#WithWindowSize), contributed by [ianwilkes](https://github.com/ianwilkes). +* Sep 5, 2019: (v1.8.2) Add [WithZeroFrames](https://godoc.org/github.com/klauspost/compress/zstd#WithZeroFrames) which adds full zero payload block encoding option. +* Sep 5, 2019: Lazy initialization of zstandard predefined en/decoder tables. +* Aug 26, 2019: (v1.8.1) S2: 1-2% compression increase in "better" compression mode. +* Aug 26, 2019: zstd: Check maximum size of Huffman 1X compressed literals while decoding. +* Aug 24, 2019: (v1.8.0) Added [S2 compression](https://github.com/klauspost/compress/tree/master/s2#s2-compression), a high performance replacement for Snappy. +* Aug 21, 2019: (v1.7.6) Fixed minor issues found by fuzzer. One could lead to zstd not decompressing. +* Aug 18, 2019: Add [fuzzit](https://fuzzit.dev/) continuous fuzzing. +* Aug 14, 2019: zstd: Skip incompressible data 2x faster. [#147](https://github.com/klauspost/compress/pull/147) +* Aug 4, 2019 (v1.7.5): Better literal compression. [#146](https://github.com/klauspost/compress/pull/146) +* Aug 4, 2019: Faster zstd compression. [#143](https://github.com/klauspost/compress/pull/143) [#144](https://github.com/klauspost/compress/pull/144) +* Aug 4, 2019: Faster zstd decompression. [#145](https://github.com/klauspost/compress/pull/145) [#143](https://github.com/klauspost/compress/pull/143) [#142](https://github.com/klauspost/compress/pull/142) +* July 15, 2019 (v1.7.4): Fix double EOF block in rare cases on zstd encoder. +* July 15, 2019 (v1.7.3): Minor speedup/compression increase in default zstd encoder. +* July 14, 2019: zstd decoder: Fix decompression error on multiple uses with mixed content. +* July 7, 2019 (v1.7.2): Snappy update, zstd decoder potential race fix. +* June 17, 2019: zstd decompression bugfix. +* June 17, 2019: fix 32 bit builds. +* June 17, 2019: Easier use in modules (less dependencies). +* June 9, 2019: New stronger "default" [zstd](https://github.com/klauspost/compress/tree/master/zstd#zstd) compression mode. Matches zstd default compression ratio. +* June 5, 2019: 20-40% throughput in [zstandard](https://github.com/klauspost/compress/tree/master/zstd#zstd) compression and better compression. +* June 5, 2019: deflate/gzip compression: Reduce memory usage of lower compression levels. +* June 2, 2019: Added [zstandard](https://github.com/klauspost/compress/tree/master/zstd#zstd) compression! +* May 25, 2019: deflate/gzip: 10% faster bit writer, mostly visible in lower levels. +* Apr 22, 2019: [zstd](https://github.com/klauspost/compress/tree/master/zstd#zstd) decompression added. +* Aug 1, 2018: Added [huff0 README](https://github.com/klauspost/compress/tree/master/huff0#huff0-entropy-compression). +* Jul 8, 2018: Added [Performance Update 2018](#performance-update-2018) below. +* Jun 23, 2018: Merged [Go 1.11 inflate optimizations](https://go-review.googlesource.com/c/go/+/102235). Go 1.9 is now required. Backwards compatible version tagged with [v1.3.0](https://github.com/klauspost/compress/releases/tag/v1.3.0). +* Apr 2, 2018: Added [huff0](https://godoc.org/github.com/klauspost/compress/huff0) en/decoder. Experimental for now, API may change. +* Mar 4, 2018: Added [FSE Entropy](https://godoc.org/github.com/klauspost/compress/fse) en/decoder. Experimental for now, API may change. +* Nov 3, 2017: Add compression [Estimate](https://godoc.org/github.com/klauspost/compress#Estimate) function. +* May 28, 2017: Reduce allocations when resetting decoder. +* Apr 02, 2017: Change back to official crc32, since changes were merged in Go 1.7. +* Jan 14, 2017: Reduce stack pressure due to array copies. See [Issue #18625](https://github.com/golang/go/issues/18625). +* Oct 25, 2016: Level 2-4 have been rewritten and now offers significantly better performance than before. +* Oct 20, 2016: Port zlib changes from Go 1.7 to fix zlib writer issue. Please update. +* Oct 16, 2016: Go 1.7 changes merged. Apples to apples this package is a few percent faster, but has a significantly better balance between speed and compression per level. +* Mar 24, 2016: Always attempt Huffman encoding on level 4-7. This improves base 64 encoded data compression. +* Mar 24, 2016: Small speedup for level 1-3. +* Feb 19, 2016: Faster bit writer, level -2 is 15% faster, level 1 is 4% faster. +* Feb 19, 2016: Handle small payloads faster in level 1-3. +* Feb 19, 2016: Added faster level 2 + 3 compression modes. +* Feb 19, 2016: [Rebalanced compression levels](https://blog.klauspost.com/rebalancing-deflate-compression-levels/), so there is a more even progresssion in terms of compression. New default level is 5. +* Feb 14, 2016: Snappy: Merge upstream changes. +* Feb 14, 2016: Snappy: Fix aggressive skipping. +* Feb 14, 2016: Snappy: Update benchmark. +* Feb 13, 2016: Deflate: Fixed assembler problem that could lead to sub-optimal compression. +* Feb 12, 2016: Snappy: Added AMD64 SSE 4.2 optimizations to matching, which makes easy to compress material run faster. Typical speedup is around 25%. +* Feb 9, 2016: Added Snappy package fork. This version is 5-7% faster, much more on hard to compress content. +* Jan 30, 2016: Optimize level 1 to 3 by not considering static dictionary or storing uncompressed. ~4-5% speedup. +* Jan 16, 2016: Optimization on deflate level 1,2,3 compression. +* Jan 8 2016: Merge [CL 18317](https://go-review.googlesource.com/#/c/18317): fix reading, writing of zip64 archives. +* Dec 8 2015: Make level 1 and -2 deterministic even if write size differs. +* Dec 8 2015: Split encoding functions, so hashing and matching can potentially be inlined. 1-3% faster on AMD64. 5% faster on other platforms. +* Dec 8 2015: Fixed rare [one byte out-of bounds read](https://github.com/klauspost/compress/issues/20). Please update! +* Nov 23 2015: Optimization on token writer. ~2-4% faster. Contributed by [@dsnet](https://github.com/dsnet). +* Nov 20 2015: Small optimization to bit writer on 64 bit systems. +* Nov 17 2015: Fixed out-of-bound errors if the underlying Writer returned an error. See [#15](https://github.com/klauspost/compress/issues/15). +* Nov 12 2015: Added [io.WriterTo](https://golang.org/pkg/io/#WriterTo) support to gzip/inflate. +* Nov 11 2015: Merged [CL 16669](https://go-review.googlesource.com/#/c/16669/4): archive/zip: enable overriding (de)compressors per file +* Oct 15 2015: Added skipping on uncompressible data. Random data speed up >5x. + +
+ +# deflate usage + +The packages are drop-in replacements for standard libraries. Simply replace the import path to use them: + +| old import | new import | Documentation +|--------------------|-----------------------------------------|--------------------| +| `compress/gzip` | `github.com/klauspost/compress/gzip` | [gzip](https://pkg.go.dev/github.com/klauspost/compress/gzip?tab=doc) +| `compress/zlib` | `github.com/klauspost/compress/zlib` | [zlib](https://pkg.go.dev/github.com/klauspost/compress/zlib?tab=doc) +| `archive/zip` | `github.com/klauspost/compress/zip` | [zip](https://pkg.go.dev/github.com/klauspost/compress/zip?tab=doc) +| `compress/flate` | `github.com/klauspost/compress/flate` | [flate](https://pkg.go.dev/github.com/klauspost/compress/flate?tab=doc) + +* Optimized [deflate](https://godoc.org/github.com/klauspost/compress/flate) packages which can be used as a dropin replacement for [gzip](https://godoc.org/github.com/klauspost/compress/gzip), [zip](https://godoc.org/github.com/klauspost/compress/zip) and [zlib](https://godoc.org/github.com/klauspost/compress/zlib). + +You may also be interested in [pgzip](https://github.com/klauspost/pgzip), which is a drop in replacement for gzip, which support multithreaded compression on big files and the optimized [crc32](https://github.com/klauspost/crc32) package used by these packages. + +The packages contains the same as the standard library, so you can use the godoc for that: [gzip](http://golang.org/pkg/compress/gzip/), [zip](http://golang.org/pkg/archive/zip/), [zlib](http://golang.org/pkg/compress/zlib/), [flate](http://golang.org/pkg/compress/flate/). + +Currently there is only minor speedup on decompression (mostly CRC32 calculation). + +Memory usage is typically 1MB for a Writer. stdlib is in the same range. +If you expect to have a lot of concurrently allocated Writers consider using +the stateless compress described below. + +For compression performance, see: [this spreadsheet](https://docs.google.com/spreadsheets/d/1nuNE2nPfuINCZJRMt6wFWhKpToF95I47XjSsc-1rbPQ/edit?usp=sharing). + +# Stateless compression + +This package offers stateless compression as a special option for gzip/deflate. +It will do compression but without maintaining any state between Write calls. + +This means there will be no memory kept between Write calls, but compression and speed will be suboptimal. + +This is only relevant in cases where you expect to run many thousands of compressors concurrently, +but with very little activity. This is *not* intended for regular web servers serving individual requests. + +Because of this, the size of actual Write calls will affect output size. + +In gzip, specify level `-3` / `gzip.StatelessCompression` to enable. + +For direct deflate use, NewStatelessWriter and StatelessDeflate are available. See [documentation](https://godoc.org/github.com/klauspost/compress/flate#NewStatelessWriter) + +A `bufio.Writer` can of course be used to control write sizes. For example, to use a 4KB buffer: + +``` + // replace 'ioutil.Discard' with your output. + gzw, err := gzip.NewWriterLevel(ioutil.Discard, gzip.StatelessCompression) + if err != nil { + return err + } + defer gzw.Close() + + w := bufio.NewWriterSize(gzw, 4096) + defer w.Flush() + + // Write to 'w' +``` + +This will only use up to 4KB in memory when the writer is idle. + +Compression is almost always worse than the fastest compression level +and each write will allocate (a little) memory. + +# Performance Update 2018 + +It has been a while since we have been looking at the speed of this package compared to the standard library, so I thought I would re-do my tests and give some overall recommendations based on the current state. All benchmarks have been performed with Go 1.10 on my Desktop Intel(R) Core(TM) i7-2600 CPU @3.40GHz. Since I last ran the tests, I have gotten more RAM, which means tests with big files are no longer limited by my SSD. + +The raw results are in my [updated spreadsheet](https://docs.google.com/spreadsheets/d/1nuNE2nPfuINCZJRMt6wFWhKpToF95I47XjSsc-1rbPQ/edit?usp=sharing). Due to cgo changes and upstream updates i could not get the cgo version of gzip to compile. Instead I included the [zstd](https://github.com/datadog/zstd) cgo implementation. If I get cgo gzip to work again, I might replace the results in the sheet. + +The columns to take note of are: *MB/s* - the throughput. *Reduction* - the data size reduction in percent of the original. *Rel Speed* relative speed compared to the standard library at the same level. *Smaller* - how many percent smaller is the compressed output compared to stdlib. Negative means the output was bigger. *Loss* means the loss (or gain) in compression as a percentage difference of the input. + +The `gzstd` (standard library gzip) and `gzkp` (this package gzip) only uses one CPU core. [`pgzip`](https://github.com/klauspost/pgzip), [`bgzf`](https://github.com/biogo/hts/tree/master/bgzf) uses all 4 cores. [`zstd`](https://github.com/DataDog/zstd) uses one core, and is a beast (but not Go, yet). + + +## Overall differences. + +There appears to be a roughly 5-10% speed advantage over the standard library when comparing at similar compression levels. + +The biggest difference you will see is the result of [re-balancing](https://blog.klauspost.com/rebalancing-deflate-compression-levels/) the compression levels. I wanted by library to give a smoother transition between the compression levels than the standard library. + +This package attempts to provide a more smooth transition, where "1" is taking a lot of shortcuts, "5" is the reasonable trade-off and "9" is the "give me the best compression", and the values in between gives something reasonable in between. The standard library has big differences in levels 1-4, but levels 5-9 having no significant gains - often spending a lot more time than can be justified by the achieved compression. + +There are links to all the test data in the [spreadsheet](https://docs.google.com/spreadsheets/d/1nuNE2nPfuINCZJRMt6wFWhKpToF95I47XjSsc-1rbPQ/edit?usp=sharing) in the top left field on each tab. + +## Web Content + +This test set aims to emulate typical use in a web server. The test-set is 4GB data in 53k files, and is a mixture of (mostly) HTML, JS, CSS. + +Since level 1 and 9 are close to being the same code, they are quite close. But looking at the levels in-between the differences are quite big. + +Looking at level 6, this package is 88% faster, but will output about 6% more data. For a web server, this means you can serve 88% more data, but have to pay for 6% more bandwidth. You can draw your own conclusions on what would be the most expensive for your case. + +## Object files + +This test is for typical data files stored on a server. In this case it is a collection of Go precompiled objects. They are very compressible. + +The picture is similar to the web content, but with small differences since this is very compressible. Levels 2-3 offer good speed, but is sacrificing quite a bit of compression. + +The standard library seems suboptimal on level 3 and 4 - offering both worse compression and speed than level 6 & 7 of this package respectively. + +## Highly Compressible File + +This is a JSON file with very high redundancy. The reduction starts at 95% on level 1, so in real life terms we are dealing with something like a highly redundant stream of data, etc. + +It is definitely visible that we are dealing with specialized content here, so the results are very scattered. This package does not do very well at levels 1-4, but picks up significantly at level 5 and levels 7 and 8 offering great speed for the achieved compression. + +So if you know you content is extremely compressible you might want to go slightly higher than the defaults. The standard library has a huge gap between levels 3 and 4 in terms of speed (2.75x slowdown), so it offers little "middle ground". + +## Medium-High Compressible + +This is a pretty common test corpus: [enwik9](http://mattmahoney.net/dc/textdata.html). It contains the first 10^9 bytes of the English Wikipedia dump on Mar. 3, 2006. This is a very good test of typical text based compression and more data heavy streams. + +We see a similar picture here as in "Web Content". On equal levels some compression is sacrificed for more speed. Level 5 seems to be the best trade-off between speed and size, beating stdlib level 3 in both. + +## Medium Compressible + +I will combine two test sets, one [10GB file set](http://mattmahoney.net/dc/10gb.html) and a VM disk image (~8GB). Both contain different data types and represent a typical backup scenario. + +The most notable thing is how quickly the standard library drops to very low compression speeds around level 5-6 without any big gains in compression. Since this type of data is fairly common, this does not seem like good behavior. + + +## Un-compressible Content + +This is mainly a test of how good the algorithms are at detecting un-compressible input. The standard library only offers this feature with very conservative settings at level 1. Obviously there is no reason for the algorithms to try to compress input that cannot be compressed. The only downside is that it might skip some compressible data on false detections. + + +## Huffman only compression + +This compression library adds a special compression level, named `HuffmanOnly`, which allows near linear time compression. This is done by completely disabling matching of previous data, and only reduce the number of bits to represent each character. + +This means that often used characters, like 'e' and ' ' (space) in text use the fewest bits to represent, and rare characters like '¤' takes more bits to represent. For more information see [wikipedia](https://en.wikipedia.org/wiki/Huffman_coding) or this nice [video](https://youtu.be/ZdooBTdW5bM). + +Since this type of compression has much less variance, the compression speed is mostly unaffected by the input data, and is usually more than *180MB/s* for a single core. + +The downside is that the compression ratio is usually considerably worse than even the fastest conventional compression. The compression ratio can never be better than 8:1 (12.5%). + +The linear time compression can be used as a "better than nothing" mode, where you cannot risk the encoder to slow down on some content. For comparison, the size of the "Twain" text is *233460 bytes* (+29% vs. level 1) and encode speed is 144MB/s (4.5x level 1). So in this case you trade a 30% size increase for a 4 times speedup. + +For more information see my blog post on [Fast Linear Time Compression](http://blog.klauspost.com/constant-time-gzipzip-compression/). + +This is implemented on Go 1.7 as "Huffman Only" mode, though not exposed for gzip. + +# Other packages + +Here are other packages of good quality and pure Go (no cgo wrappers or autoconverted code): + +* [github.com/pierrec/lz4](https://github.com/pierrec/lz4) - strong multithreaded LZ4 compression. +* [github.com/cosnicolaou/pbzip2](https://github.com/cosnicolaou/pbzip2) - multithreaded bzip2 decompression. +* [github.com/dsnet/compress](https://github.com/dsnet/compress) - brotli decompression, bzip2 writer. + +# license + +This code is licensed under the same conditions as the original Go code. See LICENSE file. diff --git a/vendor/github.com/klauspost/compress/compressible.go b/vendor/github.com/klauspost/compress/compressible.go new file mode 100644 index 0000000..ea5a692 --- /dev/null +++ b/vendor/github.com/klauspost/compress/compressible.go @@ -0,0 +1,85 @@ +package compress + +import "math" + +// Estimate returns a normalized compressibility estimate of block b. +// Values close to zero are likely uncompressible. +// Values above 0.1 are likely to be compressible. +// Values above 0.5 are very compressible. +// Very small lengths will return 0. +func Estimate(b []byte) float64 { + if len(b) < 16 { + return 0 + } + + // Correctly predicted order 1 + hits := 0 + lastMatch := false + var o1 [256]byte + var hist [256]int + c1 := byte(0) + for _, c := range b { + if c == o1[c1] { + // We only count a hit if there was two correct predictions in a row. + if lastMatch { + hits++ + } + lastMatch = true + } else { + lastMatch = false + } + o1[c1] = c + c1 = c + hist[c]++ + } + + // Use x^0.6 to give better spread + prediction := math.Pow(float64(hits)/float64(len(b)), 0.6) + + // Calculate histogram distribution + variance := float64(0) + avg := float64(len(b)) / 256 + + for _, v := range hist { + Δ := float64(v) - avg + variance += Δ * Δ + } + + stddev := math.Sqrt(float64(variance)) / float64(len(b)) + exp := math.Sqrt(1 / float64(len(b))) + + // Subtract expected stddev + stddev -= exp + if stddev < 0 { + stddev = 0 + } + stddev *= 1 + exp + + // Use x^0.4 to give better spread + entropy := math.Pow(stddev, 0.4) + + // 50/50 weight between prediction and histogram distribution + return math.Pow((prediction+entropy)/2, 0.9) +} + +// ShannonEntropyBits returns the number of bits minimum required to represent +// an entropy encoding of the input bytes. +// https://en.wiktionary.org/wiki/Shannon_entropy +func ShannonEntropyBits(b []byte) int { + if len(b) == 0 { + return 0 + } + var hist [256]int + for _, c := range b { + hist[c]++ + } + shannon := float64(0) + invTotal := 1.0 / float64(len(b)) + for _, v := range hist[:] { + if v > 0 { + n := float64(v) + shannon += math.Ceil(-math.Log2(n*invTotal) * n) + } + } + return int(math.Ceil(shannon)) +} diff --git a/vendor/github.com/klauspost/compress/fse/README.md b/vendor/github.com/klauspost/compress/fse/README.md new file mode 100644 index 0000000..ea7324d --- /dev/null +++ b/vendor/github.com/klauspost/compress/fse/README.md @@ -0,0 +1,79 @@ +# Finite State Entropy + +This package provides Finite State Entropy encoding and decoding. + +Finite State Entropy (also referenced as [tANS](https://en.wikipedia.org/wiki/Asymmetric_numeral_systems#tANS)) +encoding provides a fast near-optimal symbol encoding/decoding +for byte blocks as implemented in [zstandard](https://github.com/facebook/zstd). + +This can be used for compressing input with a lot of similar input values to the smallest number of bytes. +This does not perform any multi-byte [dictionary coding](https://en.wikipedia.org/wiki/Dictionary_coder) as LZ coders, +but it can be used as a secondary step to compressors (like Snappy) that does not do entropy encoding. + +* [Godoc documentation](https://godoc.org/github.com/klauspost/compress/fse) + +## News + + * Feb 2018: First implementation released. Consider this beta software for now. + +# Usage + +This package provides a low level interface that allows to compress single independent blocks. + +Each block is separate, and there is no built in integrity checks. +This means that the caller should keep track of block sizes and also do checksums if needed. + +Compressing a block is done via the [`Compress`](https://godoc.org/github.com/klauspost/compress/fse#Compress) function. +You must provide input and will receive the output and maybe an error. + +These error values can be returned: + +| Error | Description | +|---------------------|-----------------------------------------------------------------------------| +| `` | Everything ok, output is returned | +| `ErrIncompressible` | Returned when input is judged to be too hard to compress | +| `ErrUseRLE` | Returned from the compressor when the input is a single byte value repeated | +| `(error)` | An internal error occurred. | + +As can be seen above there are errors that will be returned even under normal operation so it is important to handle these. + +To reduce allocations you can provide a [`Scratch`](https://godoc.org/github.com/klauspost/compress/fse#Scratch) object +that can be re-used for successive calls. Both compression and decompression accepts a `Scratch` object, and the same +object can be used for both. + +Be aware, that when re-using a `Scratch` object that the *output* buffer is also re-used, so if you are still using this +you must set the `Out` field in the scratch to nil. The same buffer is used for compression and decompression output. + +Decompressing is done by calling the [`Decompress`](https://godoc.org/github.com/klauspost/compress/fse#Decompress) function. +You must provide the output from the compression stage, at exactly the size you got back. If you receive an error back +your input was likely corrupted. + +It is important to note that a successful decoding does *not* mean your output matches your original input. +There are no integrity checks, so relying on errors from the decompressor does not assure your data is valid. + +For more detailed usage, see examples in the [godoc documentation](https://godoc.org/github.com/klauspost/compress/fse#pkg-examples). + +# Performance + +A lot of factors are affecting speed. Block sizes and compressibility of the material are primary factors. +All compression functions are currently only running on the calling goroutine so only one core will be used per block. + +The compressor is significantly faster if symbols are kept as small as possible. The highest byte value of the input +is used to reduce some of the processing, so if all your input is above byte value 64 for instance, it may be +beneficial to transpose all your input values down by 64. + +With moderate block sizes around 64k speed are typically 200MB/s per core for compression and +around 300MB/s decompression speed. + +The same hardware typically does Huffman (deflate) encoding at 125MB/s and decompression at 100MB/s. + +# Plans + +At one point, more internals will be exposed to facilitate more "expert" usage of the components. + +A streaming interface is also likely to be implemented. Likely compatible with [FSE stream format](https://github.com/Cyan4973/FiniteStateEntropy/blob/dev/programs/fileio.c#L261). + +# Contributing + +Contributions are always welcome. Be aware that adding public functions will require good justification and breaking +changes will likely not be accepted. If in doubt open an issue before writing the PR. \ No newline at end of file diff --git a/vendor/github.com/klauspost/compress/fse/bitreader.go b/vendor/github.com/klauspost/compress/fse/bitreader.go new file mode 100644 index 0000000..f65eb39 --- /dev/null +++ b/vendor/github.com/klauspost/compress/fse/bitreader.go @@ -0,0 +1,122 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package fse + +import ( + "encoding/binary" + "errors" + "io" +) + +// bitReader reads a bitstream in reverse. +// The last set bit indicates the start of the stream and is used +// for aligning the input. +type bitReader struct { + in []byte + off uint // next byte to read is at in[off - 1] + value uint64 + bitsRead uint8 +} + +// init initializes and resets the bit reader. +func (b *bitReader) init(in []byte) error { + if len(in) < 1 { + return errors.New("corrupt stream: too short") + } + b.in = in + b.off = uint(len(in)) + // The highest bit of the last byte indicates where to start + v := in[len(in)-1] + if v == 0 { + return errors.New("corrupt stream, did not find end of stream") + } + b.bitsRead = 64 + b.value = 0 + if len(in) >= 8 { + b.fillFastStart() + } else { + b.fill() + b.fill() + } + b.bitsRead += 8 - uint8(highBits(uint32(v))) + return nil +} + +// getBits will return n bits. n can be 0. +func (b *bitReader) getBits(n uint8) uint16 { + if n == 0 || b.bitsRead >= 64 { + return 0 + } + return b.getBitsFast(n) +} + +// getBitsFast requires that at least one bit is requested every time. +// There are no checks if the buffer is filled. +func (b *bitReader) getBitsFast(n uint8) uint16 { + const regMask = 64 - 1 + v := uint16((b.value << (b.bitsRead & regMask)) >> ((regMask + 1 - n) & regMask)) + b.bitsRead += n + return v +} + +// fillFast() will make sure at least 32 bits are available. +// There must be at least 4 bytes available. +func (b *bitReader) fillFast() { + if b.bitsRead < 32 { + return + } + // 2 bounds checks. + v := b.in[b.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value = (b.value << 32) | uint64(low) + b.bitsRead -= 32 + b.off -= 4 +} + +// fill() will make sure at least 32 bits are available. +func (b *bitReader) fill() { + if b.bitsRead < 32 { + return + } + if b.off > 4 { + v := b.in[b.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value = (b.value << 32) | uint64(low) + b.bitsRead -= 32 + b.off -= 4 + return + } + for b.off > 0 { + b.value = (b.value << 8) | uint64(b.in[b.off-1]) + b.bitsRead -= 8 + b.off-- + } +} + +// fillFastStart() assumes the bitreader is empty and there is at least 8 bytes to read. +func (b *bitReader) fillFastStart() { + // Do single re-slice to avoid bounds checks. + b.value = binary.LittleEndian.Uint64(b.in[b.off-8:]) + b.bitsRead = 0 + b.off -= 8 +} + +// finished returns true if all bits have been read from the bit stream. +func (b *bitReader) finished() bool { + return b.bitsRead >= 64 && b.off == 0 +} + +// close the bitstream and returns an error if out-of-buffer reads occurred. +func (b *bitReader) close() error { + // Release reference. + b.in = nil + if b.bitsRead > 64 { + return io.ErrUnexpectedEOF + } + return nil +} diff --git a/vendor/github.com/klauspost/compress/fse/bitwriter.go b/vendor/github.com/klauspost/compress/fse/bitwriter.go new file mode 100644 index 0000000..43e4636 --- /dev/null +++ b/vendor/github.com/klauspost/compress/fse/bitwriter.go @@ -0,0 +1,168 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package fse + +import "fmt" + +// bitWriter will write bits. +// First bit will be LSB of the first byte of output. +type bitWriter struct { + bitContainer uint64 + nBits uint8 + out []byte +} + +// bitMask16 is bitmasks. Has extra to avoid bounds check. +var bitMask16 = [32]uint16{ + 0, 1, 3, 7, 0xF, 0x1F, + 0x3F, 0x7F, 0xFF, 0x1FF, 0x3FF, 0x7FF, + 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0xFFFF, + 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, + 0xFFFF, 0xFFFF} /* up to 16 bits */ + +// addBits16NC will add up to 16 bits. +// It will not check if there is space for them, +// so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits16NC(value uint16, bits uint8) { + b.bitContainer |= uint64(value&bitMask16[bits&31]) << (b.nBits & 63) + b.nBits += bits +} + +// addBits16Clean will add up to 16 bits. value may not contain more set bits than indicated. +// It will not check if there is space for them, so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits16Clean(value uint16, bits uint8) { + b.bitContainer |= uint64(value) << (b.nBits & 63) + b.nBits += bits +} + +// addBits16ZeroNC will add up to 16 bits. +// It will not check if there is space for them, +// so the caller must ensure that it has flushed recently. +// This is fastest if bits can be zero. +func (b *bitWriter) addBits16ZeroNC(value uint16, bits uint8) { + if bits == 0 { + return + } + value <<= (16 - bits) & 15 + value >>= (16 - bits) & 15 + b.bitContainer |= uint64(value) << (b.nBits & 63) + b.nBits += bits +} + +// flush will flush all pending full bytes. +// There will be at least 56 bits available for writing when this has been called. +// Using flush32 is faster, but leaves less space for writing. +func (b *bitWriter) flush() { + v := b.nBits >> 3 + switch v { + case 0: + case 1: + b.out = append(b.out, + byte(b.bitContainer), + ) + case 2: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + ) + case 3: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + ) + case 4: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + ) + case 5: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + ) + case 6: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + ) + case 7: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + byte(b.bitContainer>>48), + ) + case 8: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + byte(b.bitContainer>>48), + byte(b.bitContainer>>56), + ) + default: + panic(fmt.Errorf("bits (%d) > 64", b.nBits)) + } + b.bitContainer >>= v << 3 + b.nBits &= 7 +} + +// flush32 will flush out, so there are at least 32 bits available for writing. +func (b *bitWriter) flush32() { + if b.nBits < 32 { + return + } + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24)) + b.nBits -= 32 + b.bitContainer >>= 32 +} + +// flushAlign will flush remaining full bytes and align to next byte boundary. +func (b *bitWriter) flushAlign() { + nbBytes := (b.nBits + 7) >> 3 + for i := uint8(0); i < nbBytes; i++ { + b.out = append(b.out, byte(b.bitContainer>>(i*8))) + } + b.nBits = 0 + b.bitContainer = 0 +} + +// close will write the alignment bit and write the final byte(s) +// to the output. +func (b *bitWriter) close() error { + // End mark + b.addBits16Clean(1, 1) + // flush until next byte. + b.flushAlign() + return nil +} + +// reset and continue writing by appending to out. +func (b *bitWriter) reset(out []byte) { + b.bitContainer = 0 + b.nBits = 0 + b.out = out +} diff --git a/vendor/github.com/klauspost/compress/fse/bytereader.go b/vendor/github.com/klauspost/compress/fse/bytereader.go new file mode 100644 index 0000000..abade2d --- /dev/null +++ b/vendor/github.com/klauspost/compress/fse/bytereader.go @@ -0,0 +1,47 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package fse + +// byteReader provides a byte reader that reads +// little endian values from a byte stream. +// The input stream is manually advanced. +// The reader performs no bounds checks. +type byteReader struct { + b []byte + off int +} + +// init will initialize the reader and set the input. +func (b *byteReader) init(in []byte) { + b.b = in + b.off = 0 +} + +// advance the stream b n bytes. +func (b *byteReader) advance(n uint) { + b.off += int(n) +} + +// Uint32 returns a little endian uint32 starting at current offset. +func (b byteReader) Uint32() uint32 { + b2 := b.b[b.off:] + b2 = b2[:4] + v3 := uint32(b2[3]) + v2 := uint32(b2[2]) + v1 := uint32(b2[1]) + v0 := uint32(b2[0]) + return v0 | (v1 << 8) | (v2 << 16) | (v3 << 24) +} + +// unread returns the unread portion of the input. +func (b byteReader) unread() []byte { + return b.b[b.off:] +} + +// remain will return the number of bytes remaining. +func (b byteReader) remain() int { + return len(b.b) - b.off +} diff --git a/vendor/github.com/klauspost/compress/fse/compress.go b/vendor/github.com/klauspost/compress/fse/compress.go new file mode 100644 index 0000000..6f34191 --- /dev/null +++ b/vendor/github.com/klauspost/compress/fse/compress.go @@ -0,0 +1,683 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package fse + +import ( + "errors" + "fmt" +) + +// Compress the input bytes. Input must be < 2GB. +// Provide a Scratch buffer to avoid memory allocations. +// Note that the output is also kept in the scratch buffer. +// If input is too hard to compress, ErrIncompressible is returned. +// If input is a single byte value repeated ErrUseRLE is returned. +func Compress(in []byte, s *Scratch) ([]byte, error) { + if len(in) <= 1 { + return nil, ErrIncompressible + } + if len(in) > (2<<30)-1 { + return nil, errors.New("input too big, must be < 2GB") + } + s, err := s.prepare(in) + if err != nil { + return nil, err + } + + // Create histogram, if none was provided. + maxCount := s.maxCount + if maxCount == 0 { + maxCount = s.countSimple(in) + } + // Reset for next run. + s.clearCount = true + s.maxCount = 0 + if maxCount == len(in) { + // One symbol, use RLE + return nil, ErrUseRLE + } + if maxCount == 1 || maxCount < (len(in)>>7) { + // Each symbol present maximum once or too well distributed. + return nil, ErrIncompressible + } + s.optimalTableLog() + err = s.normalizeCount() + if err != nil { + return nil, err + } + err = s.writeCount() + if err != nil { + return nil, err + } + + if false { + err = s.validateNorm() + if err != nil { + return nil, err + } + } + + err = s.buildCTable() + if err != nil { + return nil, err + } + err = s.compress(in) + if err != nil { + return nil, err + } + s.Out = s.bw.out + // Check if we compressed. + if len(s.Out) >= len(in) { + return nil, ErrIncompressible + } + return s.Out, nil +} + +// cState contains the compression state of a stream. +type cState struct { + bw *bitWriter + stateTable []uint16 + state uint16 +} + +// init will initialize the compression state to the first symbol of the stream. +func (c *cState) init(bw *bitWriter, ct *cTable, tableLog uint8, first symbolTransform) { + c.bw = bw + c.stateTable = ct.stateTable + + nbBitsOut := (first.deltaNbBits + (1 << 15)) >> 16 + im := int32((nbBitsOut << 16) - first.deltaNbBits) + lu := (im >> nbBitsOut) + first.deltaFindState + c.state = c.stateTable[lu] +} + +// encode the output symbol provided and write it to the bitstream. +func (c *cState) encode(symbolTT symbolTransform) { + nbBitsOut := (uint32(c.state) + symbolTT.deltaNbBits) >> 16 + dstState := int32(c.state>>(nbBitsOut&15)) + symbolTT.deltaFindState + c.bw.addBits16NC(c.state, uint8(nbBitsOut)) + c.state = c.stateTable[dstState] +} + +// encode the output symbol provided and write it to the bitstream. +func (c *cState) encodeZero(symbolTT symbolTransform) { + nbBitsOut := (uint32(c.state) + symbolTT.deltaNbBits) >> 16 + dstState := int32(c.state>>(nbBitsOut&15)) + symbolTT.deltaFindState + c.bw.addBits16ZeroNC(c.state, uint8(nbBitsOut)) + c.state = c.stateTable[dstState] +} + +// flush will write the tablelog to the output and flush the remaining full bytes. +func (c *cState) flush(tableLog uint8) { + c.bw.flush32() + c.bw.addBits16NC(c.state, tableLog) + c.bw.flush() +} + +// compress is the main compression loop that will encode the input from the last byte to the first. +func (s *Scratch) compress(src []byte) error { + if len(src) <= 2 { + return errors.New("compress: src too small") + } + tt := s.ct.symbolTT[:256] + s.bw.reset(s.Out) + + // Our two states each encodes every second byte. + // Last byte encoded (first byte decoded) will always be encoded by c1. + var c1, c2 cState + + // Encode so remaining size is divisible by 4. + ip := len(src) + if ip&1 == 1 { + c1.init(&s.bw, &s.ct, s.actualTableLog, tt[src[ip-1]]) + c2.init(&s.bw, &s.ct, s.actualTableLog, tt[src[ip-2]]) + c1.encodeZero(tt[src[ip-3]]) + ip -= 3 + } else { + c2.init(&s.bw, &s.ct, s.actualTableLog, tt[src[ip-1]]) + c1.init(&s.bw, &s.ct, s.actualTableLog, tt[src[ip-2]]) + ip -= 2 + } + if ip&2 != 0 { + c2.encodeZero(tt[src[ip-1]]) + c1.encodeZero(tt[src[ip-2]]) + ip -= 2 + } + + // Main compression loop. + switch { + case !s.zeroBits && s.actualTableLog <= 8: + // We can encode 4 symbols without requiring a flush. + // We do not need to check if any output is 0 bits. + for ip >= 4 { + s.bw.flush32() + v3, v2, v1, v0 := src[ip-4], src[ip-3], src[ip-2], src[ip-1] + c2.encode(tt[v0]) + c1.encode(tt[v1]) + c2.encode(tt[v2]) + c1.encode(tt[v3]) + ip -= 4 + } + case !s.zeroBits: + // We do not need to check if any output is 0 bits. + for ip >= 4 { + s.bw.flush32() + v3, v2, v1, v0 := src[ip-4], src[ip-3], src[ip-2], src[ip-1] + c2.encode(tt[v0]) + c1.encode(tt[v1]) + s.bw.flush32() + c2.encode(tt[v2]) + c1.encode(tt[v3]) + ip -= 4 + } + case s.actualTableLog <= 8: + // We can encode 4 symbols without requiring a flush + for ip >= 4 { + s.bw.flush32() + v3, v2, v1, v0 := src[ip-4], src[ip-3], src[ip-2], src[ip-1] + c2.encodeZero(tt[v0]) + c1.encodeZero(tt[v1]) + c2.encodeZero(tt[v2]) + c1.encodeZero(tt[v3]) + ip -= 4 + } + default: + for ip >= 4 { + s.bw.flush32() + v3, v2, v1, v0 := src[ip-4], src[ip-3], src[ip-2], src[ip-1] + c2.encodeZero(tt[v0]) + c1.encodeZero(tt[v1]) + s.bw.flush32() + c2.encodeZero(tt[v2]) + c1.encodeZero(tt[v3]) + ip -= 4 + } + } + + // Flush final state. + // Used to initialize state when decoding. + c2.flush(s.actualTableLog) + c1.flush(s.actualTableLog) + + return s.bw.close() +} + +// writeCount will write the normalized histogram count to header. +// This is read back by readNCount. +func (s *Scratch) writeCount() error { + var ( + tableLog = s.actualTableLog + tableSize = 1 << tableLog + previous0 bool + charnum uint16 + + maxHeaderSize = ((int(s.symbolLen) * int(tableLog)) >> 3) + 3 + + // Write Table Size + bitStream = uint32(tableLog - minTablelog) + bitCount = uint(4) + remaining = int16(tableSize + 1) /* +1 for extra accuracy */ + threshold = int16(tableSize) + nbBits = uint(tableLog + 1) + ) + if cap(s.Out) < maxHeaderSize { + s.Out = make([]byte, 0, s.br.remain()+maxHeaderSize) + } + outP := uint(0) + out := s.Out[:maxHeaderSize] + + // stops at 1 + for remaining > 1 { + if previous0 { + start := charnum + for s.norm[charnum] == 0 { + charnum++ + } + for charnum >= start+24 { + start += 24 + bitStream += uint32(0xFFFF) << bitCount + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += 2 + bitStream >>= 16 + } + for charnum >= start+3 { + start += 3 + bitStream += 3 << bitCount + bitCount += 2 + } + bitStream += uint32(charnum-start) << bitCount + bitCount += 2 + if bitCount > 16 { + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += 2 + bitStream >>= 16 + bitCount -= 16 + } + } + + count := s.norm[charnum] + charnum++ + max := (2*threshold - 1) - remaining + if count < 0 { + remaining += count + } else { + remaining -= count + } + count++ // +1 for extra accuracy + if count >= threshold { + count += max // [0..max[ [max..threshold[ (...) [threshold+max 2*threshold[ + } + bitStream += uint32(count) << bitCount + bitCount += nbBits + if count < max { + bitCount-- + } + + previous0 = count == 1 + if remaining < 1 { + return errors.New("internal error: remaining<1") + } + for remaining < threshold { + nbBits-- + threshold >>= 1 + } + + if bitCount > 16 { + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += 2 + bitStream >>= 16 + bitCount -= 16 + } + } + + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += (bitCount + 7) / 8 + + if charnum > s.symbolLen { + return errors.New("internal error: charnum > s.symbolLen") + } + s.Out = out[:outP] + return nil +} + +// symbolTransform contains the state transform for a symbol. +type symbolTransform struct { + deltaFindState int32 + deltaNbBits uint32 +} + +// String prints values as a human readable string. +func (s symbolTransform) String() string { + return fmt.Sprintf("dnbits: %08x, fs:%d", s.deltaNbBits, s.deltaFindState) +} + +// cTable contains tables used for compression. +type cTable struct { + tableSymbol []byte + stateTable []uint16 + symbolTT []symbolTransform +} + +// allocCtable will allocate tables needed for compression. +// If existing tables a re big enough, they are simply re-used. +func (s *Scratch) allocCtable() { + tableSize := 1 << s.actualTableLog + // get tableSymbol that is big enough. + if cap(s.ct.tableSymbol) < tableSize { + s.ct.tableSymbol = make([]byte, tableSize) + } + s.ct.tableSymbol = s.ct.tableSymbol[:tableSize] + + ctSize := tableSize + if cap(s.ct.stateTable) < ctSize { + s.ct.stateTable = make([]uint16, ctSize) + } + s.ct.stateTable = s.ct.stateTable[:ctSize] + + if cap(s.ct.symbolTT) < 256 { + s.ct.symbolTT = make([]symbolTransform, 256) + } + s.ct.symbolTT = s.ct.symbolTT[:256] +} + +// buildCTable will populate the compression table so it is ready to be used. +func (s *Scratch) buildCTable() error { + tableSize := uint32(1 << s.actualTableLog) + highThreshold := tableSize - 1 + var cumul [maxSymbolValue + 2]int16 + + s.allocCtable() + tableSymbol := s.ct.tableSymbol[:tableSize] + // symbol start positions + { + cumul[0] = 0 + for ui, v := range s.norm[:s.symbolLen-1] { + u := byte(ui) // one less than reference + if v == -1 { + // Low proba symbol + cumul[u+1] = cumul[u] + 1 + tableSymbol[highThreshold] = u + highThreshold-- + } else { + cumul[u+1] = cumul[u] + v + } + } + // Encode last symbol separately to avoid overflowing u + u := int(s.symbolLen - 1) + v := s.norm[s.symbolLen-1] + if v == -1 { + // Low proba symbol + cumul[u+1] = cumul[u] + 1 + tableSymbol[highThreshold] = byte(u) + highThreshold-- + } else { + cumul[u+1] = cumul[u] + v + } + if uint32(cumul[s.symbolLen]) != tableSize { + return fmt.Errorf("internal error: expected cumul[s.symbolLen] (%d) == tableSize (%d)", cumul[s.symbolLen], tableSize) + } + cumul[s.symbolLen] = int16(tableSize) + 1 + } + // Spread symbols + s.zeroBits = false + { + step := tableStep(tableSize) + tableMask := tableSize - 1 + var position uint32 + // if any symbol > largeLimit, we may have 0 bits output. + largeLimit := int16(1 << (s.actualTableLog - 1)) + for ui, v := range s.norm[:s.symbolLen] { + symbol := byte(ui) + if v > largeLimit { + s.zeroBits = true + } + for nbOccurrences := int16(0); nbOccurrences < v; nbOccurrences++ { + tableSymbol[position] = symbol + position = (position + step) & tableMask + for position > highThreshold { + position = (position + step) & tableMask + } /* Low proba area */ + } + } + + // Check if we have gone through all positions + if position != 0 { + return errors.New("position!=0") + } + } + + // Build table + table := s.ct.stateTable + { + tsi := int(tableSize) + for u, v := range tableSymbol { + // TableU16 : sorted by symbol order; gives next state value + table[cumul[v]] = uint16(tsi + u) + cumul[v]++ + } + } + + // Build Symbol Transformation Table + { + total := int16(0) + symbolTT := s.ct.symbolTT[:s.symbolLen] + tableLog := s.actualTableLog + tl := (uint32(tableLog) << 16) - (1 << tableLog) + for i, v := range s.norm[:s.symbolLen] { + switch v { + case 0: + case -1, 1: + symbolTT[i].deltaNbBits = tl + symbolTT[i].deltaFindState = int32(total - 1) + total++ + default: + maxBitsOut := uint32(tableLog) - highBits(uint32(v-1)) + minStatePlus := uint32(v) << maxBitsOut + symbolTT[i].deltaNbBits = (maxBitsOut << 16) - minStatePlus + symbolTT[i].deltaFindState = int32(total - v) + total += v + } + } + if total != int16(tableSize) { + return fmt.Errorf("total mismatch %d (got) != %d (want)", total, tableSize) + } + } + return nil +} + +// countSimple will create a simple histogram in s.count. +// Returns the biggest count. +// Does not update s.clearCount. +func (s *Scratch) countSimple(in []byte) (max int) { + for _, v := range in { + s.count[v]++ + } + m := uint32(0) + for i, v := range s.count[:] { + if v > m { + m = v + } + if v > 0 { + s.symbolLen = uint16(i) + 1 + } + } + return int(m) +} + +// minTableLog provides the minimum logSize to safely represent a distribution. +func (s *Scratch) minTableLog() uint8 { + minBitsSrc := highBits(uint32(s.br.remain()-1)) + 1 + minBitsSymbols := highBits(uint32(s.symbolLen-1)) + 2 + if minBitsSrc < minBitsSymbols { + return uint8(minBitsSrc) + } + return uint8(minBitsSymbols) +} + +// optimalTableLog calculates and sets the optimal tableLog in s.actualTableLog +func (s *Scratch) optimalTableLog() { + tableLog := s.TableLog + minBits := s.minTableLog() + maxBitsSrc := uint8(highBits(uint32(s.br.remain()-1))) - 2 + if maxBitsSrc < tableLog { + // Accuracy can be reduced + tableLog = maxBitsSrc + } + if minBits > tableLog { + tableLog = minBits + } + // Need a minimum to safely represent all symbol values + if tableLog < minTablelog { + tableLog = minTablelog + } + if tableLog > maxTableLog { + tableLog = maxTableLog + } + s.actualTableLog = tableLog +} + +var rtbTable = [...]uint32{0, 473195, 504333, 520860, 550000, 700000, 750000, 830000} + +// normalizeCount will normalize the count of the symbols so +// the total is equal to the table size. +func (s *Scratch) normalizeCount() error { + var ( + tableLog = s.actualTableLog + scale = 62 - uint64(tableLog) + step = (1 << 62) / uint64(s.br.remain()) + vStep = uint64(1) << (scale - 20) + stillToDistribute = int16(1 << tableLog) + largest int + largestP int16 + lowThreshold = (uint32)(s.br.remain() >> tableLog) + ) + + for i, cnt := range s.count[:s.symbolLen] { + // already handled + // if (count[s] == s.length) return 0; /* rle special case */ + + if cnt == 0 { + s.norm[i] = 0 + continue + } + if cnt <= lowThreshold { + s.norm[i] = -1 + stillToDistribute-- + } else { + proba := (int16)((uint64(cnt) * step) >> scale) + if proba < 8 { + restToBeat := vStep * uint64(rtbTable[proba]) + v := uint64(cnt)*step - (uint64(proba) << scale) + if v > restToBeat { + proba++ + } + } + if proba > largestP { + largestP = proba + largest = i + } + s.norm[i] = proba + stillToDistribute -= proba + } + } + + if -stillToDistribute >= (s.norm[largest] >> 1) { + // corner case, need another normalization method + return s.normalizeCount2() + } + s.norm[largest] += stillToDistribute + return nil +} + +// Secondary normalization method. +// To be used when primary method fails. +func (s *Scratch) normalizeCount2() error { + const notYetAssigned = -2 + var ( + distributed uint32 + total = uint32(s.br.remain()) + tableLog = s.actualTableLog + lowThreshold = total >> tableLog + lowOne = (total * 3) >> (tableLog + 1) + ) + for i, cnt := range s.count[:s.symbolLen] { + if cnt == 0 { + s.norm[i] = 0 + continue + } + if cnt <= lowThreshold { + s.norm[i] = -1 + distributed++ + total -= cnt + continue + } + if cnt <= lowOne { + s.norm[i] = 1 + distributed++ + total -= cnt + continue + } + s.norm[i] = notYetAssigned + } + toDistribute := (1 << tableLog) - distributed + + if (total / toDistribute) > lowOne { + // risk of rounding to zero + lowOne = (total * 3) / (toDistribute * 2) + for i, cnt := range s.count[:s.symbolLen] { + if (s.norm[i] == notYetAssigned) && (cnt <= lowOne) { + s.norm[i] = 1 + distributed++ + total -= cnt + continue + } + } + toDistribute = (1 << tableLog) - distributed + } + if distributed == uint32(s.symbolLen)+1 { + // all values are pretty poor; + // probably incompressible data (should have already been detected); + // find max, then give all remaining points to max + var maxV int + var maxC uint32 + for i, cnt := range s.count[:s.symbolLen] { + if cnt > maxC { + maxV = i + maxC = cnt + } + } + s.norm[maxV] += int16(toDistribute) + return nil + } + + if total == 0 { + // all of the symbols were low enough for the lowOne or lowThreshold + for i := uint32(0); toDistribute > 0; i = (i + 1) % (uint32(s.symbolLen)) { + if s.norm[i] > 0 { + toDistribute-- + s.norm[i]++ + } + } + return nil + } + + var ( + vStepLog = 62 - uint64(tableLog) + mid = uint64((1 << (vStepLog - 1)) - 1) + rStep = (((1 << vStepLog) * uint64(toDistribute)) + mid) / uint64(total) // scale on remaining + tmpTotal = mid + ) + for i, cnt := range s.count[:s.symbolLen] { + if s.norm[i] == notYetAssigned { + var ( + end = tmpTotal + uint64(cnt)*rStep + sStart = uint32(tmpTotal >> vStepLog) + sEnd = uint32(end >> vStepLog) + weight = sEnd - sStart + ) + if weight < 1 { + return errors.New("weight < 1") + } + s.norm[i] = int16(weight) + tmpTotal = end + } + } + return nil +} + +// validateNorm validates the normalized histogram table. +func (s *Scratch) validateNorm() (err error) { + var total int + for _, v := range s.norm[:s.symbolLen] { + if v >= 0 { + total += int(v) + } else { + total -= int(v) + } + } + defer func() { + if err == nil { + return + } + fmt.Printf("selected TableLog: %d, Symbol length: %d\n", s.actualTableLog, s.symbolLen) + for i, v := range s.norm[:s.symbolLen] { + fmt.Printf("%3d: %5d -> %4d \n", i, s.count[i], v) + } + }() + if total != (1 << s.actualTableLog) { + return fmt.Errorf("warning: Total == %d != %d", total, 1< tablelogAbsoluteMax { + return errors.New("tableLog too large") + } + bitStream >>= 4 + bitCount := uint(4) + + s.actualTableLog = uint8(nbBits) + remaining := int32((1 << nbBits) + 1) + threshold := int32(1 << nbBits) + gotTotal := int32(0) + nbBits++ + + for remaining > 1 { + if previous0 { + n0 := charnum + for (bitStream & 0xFFFF) == 0xFFFF { + n0 += 24 + if b.off < iend-5 { + b.advance(2) + bitStream = b.Uint32() >> bitCount + } else { + bitStream >>= 16 + bitCount += 16 + } + } + for (bitStream & 3) == 3 { + n0 += 3 + bitStream >>= 2 + bitCount += 2 + } + n0 += uint16(bitStream & 3) + bitCount += 2 + if n0 > maxSymbolValue { + return errors.New("maxSymbolValue too small") + } + for charnum < n0 { + s.norm[charnum&0xff] = 0 + charnum++ + } + + if b.off <= iend-7 || b.off+int(bitCount>>3) <= iend-4 { + b.advance(bitCount >> 3) + bitCount &= 7 + bitStream = b.Uint32() >> bitCount + } else { + bitStream >>= 2 + } + } + + max := (2*(threshold) - 1) - (remaining) + var count int32 + + if (int32(bitStream) & (threshold - 1)) < max { + count = int32(bitStream) & (threshold - 1) + bitCount += nbBits - 1 + } else { + count = int32(bitStream) & (2*threshold - 1) + if count >= threshold { + count -= max + } + bitCount += nbBits + } + + count-- // extra accuracy + if count < 0 { + // -1 means +1 + remaining += count + gotTotal -= count + } else { + remaining -= count + gotTotal += count + } + s.norm[charnum&0xff] = int16(count) + charnum++ + previous0 = count == 0 + for remaining < threshold { + nbBits-- + threshold >>= 1 + } + if b.off <= iend-7 || b.off+int(bitCount>>3) <= iend-4 { + b.advance(bitCount >> 3) + bitCount &= 7 + } else { + bitCount -= (uint)(8 * (len(b.b) - 4 - b.off)) + b.off = len(b.b) - 4 + } + bitStream = b.Uint32() >> (bitCount & 31) + } + s.symbolLen = charnum + + if s.symbolLen <= 1 { + return fmt.Errorf("symbolLen (%d) too small", s.symbolLen) + } + if s.symbolLen > maxSymbolValue+1 { + return fmt.Errorf("symbolLen (%d) too big", s.symbolLen) + } + if remaining != 1 { + return fmt.Errorf("corruption detected (remaining %d != 1)", remaining) + } + if bitCount > 32 { + return fmt.Errorf("corruption detected (bitCount %d > 32)", bitCount) + } + if gotTotal != 1<> 3) + return nil +} + +// decSymbol contains information about a state entry, +// Including the state offset base, the output symbol and +// the number of bits to read for the low part of the destination state. +type decSymbol struct { + newState uint16 + symbol uint8 + nbBits uint8 +} + +// allocDtable will allocate decoding tables if they are not big enough. +func (s *Scratch) allocDtable() { + tableSize := 1 << s.actualTableLog + if cap(s.decTable) < tableSize { + s.decTable = make([]decSymbol, tableSize) + } + s.decTable = s.decTable[:tableSize] + + if cap(s.ct.tableSymbol) < 256 { + s.ct.tableSymbol = make([]byte, 256) + } + s.ct.tableSymbol = s.ct.tableSymbol[:256] + + if cap(s.ct.stateTable) < 256 { + s.ct.stateTable = make([]uint16, 256) + } + s.ct.stateTable = s.ct.stateTable[:256] +} + +// buildDtable will build the decoding table. +func (s *Scratch) buildDtable() error { + tableSize := uint32(1 << s.actualTableLog) + highThreshold := tableSize - 1 + s.allocDtable() + symbolNext := s.ct.stateTable[:256] + + // Init, lay down lowprob symbols + s.zeroBits = false + { + largeLimit := int16(1 << (s.actualTableLog - 1)) + for i, v := range s.norm[:s.symbolLen] { + if v == -1 { + s.decTable[highThreshold].symbol = uint8(i) + highThreshold-- + symbolNext[i] = 1 + } else { + if v >= largeLimit { + s.zeroBits = true + } + symbolNext[i] = uint16(v) + } + } + } + // Spread symbols + { + tableMask := tableSize - 1 + step := tableStep(tableSize) + position := uint32(0) + for ss, v := range s.norm[:s.symbolLen] { + for i := 0; i < int(v); i++ { + s.decTable[position].symbol = uint8(ss) + position = (position + step) & tableMask + for position > highThreshold { + // lowprob area + position = (position + step) & tableMask + } + } + } + if position != 0 { + // position must reach all cells once, otherwise normalizedCounter is incorrect + return errors.New("corrupted input (position != 0)") + } + } + + // Build Decoding table + { + tableSize := uint16(1 << s.actualTableLog) + for u, v := range s.decTable { + symbol := v.symbol + nextState := symbolNext[symbol] + symbolNext[symbol] = nextState + 1 + nBits := s.actualTableLog - byte(highBits(uint32(nextState))) + s.decTable[u].nbBits = nBits + newState := (nextState << nBits) - tableSize + if newState >= tableSize { + return fmt.Errorf("newState (%d) outside table size (%d)", newState, tableSize) + } + if newState == uint16(u) && nBits == 0 { + // Seems weird that this is possible with nbits > 0. + return fmt.Errorf("newState (%d) == oldState (%d) and no bits", newState, u) + } + s.decTable[u].newState = newState + } + } + return nil +} + +// decompress will decompress the bitstream. +// If the buffer is over-read an error is returned. +func (s *Scratch) decompress() error { + br := &s.bits + br.init(s.br.unread()) + + var s1, s2 decoder + // Initialize and decode first state and symbol. + s1.init(br, s.decTable, s.actualTableLog) + s2.init(br, s.decTable, s.actualTableLog) + + // Use temp table to avoid bound checks/append penalty. + var tmp = s.ct.tableSymbol[:256] + var off uint8 + + // Main part + if !s.zeroBits { + for br.off >= 8 { + br.fillFast() + tmp[off+0] = s1.nextFast() + tmp[off+1] = s2.nextFast() + br.fillFast() + tmp[off+2] = s1.nextFast() + tmp[off+3] = s2.nextFast() + off += 4 + // When off is 0, we have overflowed and should write. + if off == 0 { + s.Out = append(s.Out, tmp...) + if len(s.Out) >= s.DecompressLimit { + return fmt.Errorf("output size (%d) > DecompressLimit (%d)", len(s.Out), s.DecompressLimit) + } + } + } + } else { + for br.off >= 8 { + br.fillFast() + tmp[off+0] = s1.next() + tmp[off+1] = s2.next() + br.fillFast() + tmp[off+2] = s1.next() + tmp[off+3] = s2.next() + off += 4 + if off == 0 { + s.Out = append(s.Out, tmp...) + // When off is 0, we have overflowed and should write. + if len(s.Out) >= s.DecompressLimit { + return fmt.Errorf("output size (%d) > DecompressLimit (%d)", len(s.Out), s.DecompressLimit) + } + } + } + } + s.Out = append(s.Out, tmp[:off]...) + + // Final bits, a bit more expensive check + for { + if s1.finished() { + s.Out = append(s.Out, s1.final(), s2.final()) + break + } + br.fill() + s.Out = append(s.Out, s1.next()) + if s2.finished() { + s.Out = append(s.Out, s2.final(), s1.final()) + break + } + s.Out = append(s.Out, s2.next()) + if len(s.Out) >= s.DecompressLimit { + return fmt.Errorf("output size (%d) > DecompressLimit (%d)", len(s.Out), s.DecompressLimit) + } + } + return br.close() +} + +// decoder keeps track of the current state and updates it from the bitstream. +type decoder struct { + state uint16 + br *bitReader + dt []decSymbol +} + +// init will initialize the decoder and read the first state from the stream. +func (d *decoder) init(in *bitReader, dt []decSymbol, tableLog uint8) { + d.dt = dt + d.br = in + d.state = in.getBits(tableLog) +} + +// next returns the next symbol and sets the next state. +// At least tablelog bits must be available in the bit reader. +func (d *decoder) next() uint8 { + n := &d.dt[d.state] + lowBits := d.br.getBits(n.nbBits) + d.state = n.newState + lowBits + return n.symbol +} + +// finished returns true if all bits have been read from the bitstream +// and the next state would require reading bits from the input. +func (d *decoder) finished() bool { + return d.br.finished() && d.dt[d.state].nbBits > 0 +} + +// final returns the current state symbol without decoding the next. +func (d *decoder) final() uint8 { + return d.dt[d.state].symbol +} + +// nextFast returns the next symbol and sets the next state. +// This can only be used if no symbols are 0 bits. +// At least tablelog bits must be available in the bit reader. +func (d *decoder) nextFast() uint8 { + n := d.dt[d.state] + lowBits := d.br.getBitsFast(n.nbBits) + d.state = n.newState + lowBits + return n.symbol +} diff --git a/vendor/github.com/klauspost/compress/fse/fse.go b/vendor/github.com/klauspost/compress/fse/fse.go new file mode 100644 index 0000000..535cbad --- /dev/null +++ b/vendor/github.com/klauspost/compress/fse/fse.go @@ -0,0 +1,144 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +// Package fse provides Finite State Entropy encoding and decoding. +// +// Finite State Entropy encoding provides a fast near-optimal symbol encoding/decoding +// for byte blocks as implemented in zstd. +// +// See https://github.com/klauspost/compress/tree/master/fse for more information. +package fse + +import ( + "errors" + "fmt" + "math/bits" +) + +const ( + /*!MEMORY_USAGE : + * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.) + * Increasing memory usage improves compression ratio + * Reduced memory usage can improve speed, due to cache effect + * Recommended max value is 14, for 16KB, which nicely fits into Intel x86 L1 cache */ + maxMemoryUsage = 14 + defaultMemoryUsage = 13 + + maxTableLog = maxMemoryUsage - 2 + maxTablesize = 1 << maxTableLog + defaultTablelog = defaultMemoryUsage - 2 + minTablelog = 5 + maxSymbolValue = 255 +) + +var ( + // ErrIncompressible is returned when input is judged to be too hard to compress. + ErrIncompressible = errors.New("input is not compressible") + + // ErrUseRLE is returned from the compressor when the input is a single byte value repeated. + ErrUseRLE = errors.New("input is single value repeated") +) + +// Scratch provides temporary storage for compression and decompression. +type Scratch struct { + // Private + count [maxSymbolValue + 1]uint32 + norm [maxSymbolValue + 1]int16 + br byteReader + bits bitReader + bw bitWriter + ct cTable // Compression tables. + decTable []decSymbol // Decompression table. + maxCount int // count of the most probable symbol + + // Per block parameters. + // These can be used to override compression parameters of the block. + // Do not touch, unless you know what you are doing. + + // Out is output buffer. + // If the scratch is re-used before the caller is done processing the output, + // set this field to nil. + // Otherwise the output buffer will be re-used for next Compression/Decompression step + // and allocation will be avoided. + Out []byte + + // DecompressLimit limits the maximum decoded size acceptable. + // If > 0 decompression will stop when approximately this many bytes + // has been decoded. + // If 0, maximum size will be 2GB. + DecompressLimit int + + symbolLen uint16 // Length of active part of the symbol table. + actualTableLog uint8 // Selected tablelog. + zeroBits bool // no bits has prob > 50%. + clearCount bool // clear count + + // MaxSymbolValue will override the maximum symbol value of the next block. + MaxSymbolValue uint8 + + // TableLog will attempt to override the tablelog for the next block. + TableLog uint8 +} + +// Histogram allows to populate the histogram and skip that step in the compression, +// It otherwise allows to inspect the histogram when compression is done. +// To indicate that you have populated the histogram call HistogramFinished +// with the value of the highest populated symbol, as well as the number of entries +// in the most populated entry. These are accepted at face value. +// The returned slice will always be length 256. +func (s *Scratch) Histogram() []uint32 { + return s.count[:] +} + +// HistogramFinished can be called to indicate that the histogram has been populated. +// maxSymbol is the index of the highest set symbol of the next data segment. +// maxCount is the number of entries in the most populated entry. +// These are accepted at face value. +func (s *Scratch) HistogramFinished(maxSymbol uint8, maxCount int) { + s.maxCount = maxCount + s.symbolLen = uint16(maxSymbol) + 1 + s.clearCount = maxCount != 0 +} + +// prepare will prepare and allocate scratch tables used for both compression and decompression. +func (s *Scratch) prepare(in []byte) (*Scratch, error) { + if s == nil { + s = &Scratch{} + } + if s.MaxSymbolValue == 0 { + s.MaxSymbolValue = 255 + } + if s.TableLog == 0 { + s.TableLog = defaultTablelog + } + if s.TableLog > maxTableLog { + return nil, fmt.Errorf("tableLog (%d) > maxTableLog (%d)", s.TableLog, maxTableLog) + } + if cap(s.Out) == 0 { + s.Out = make([]byte, 0, len(in)) + } + if s.clearCount && s.maxCount == 0 { + for i := range s.count { + s.count[i] = 0 + } + s.clearCount = false + } + s.br.init(in) + if s.DecompressLimit == 0 { + // Max size 2GB. + s.DecompressLimit = (2 << 30) - 1 + } + + return s, nil +} + +// tableStep returns the next table index. +func tableStep(tableSize uint32) uint32 { + return (tableSize >> 1) + (tableSize >> 3) + 3 +} + +func highBits(val uint32) (n uint32) { + return uint32(bits.Len32(val) - 1) +} diff --git a/vendor/github.com/klauspost/compress/gen.sh b/vendor/github.com/klauspost/compress/gen.sh new file mode 100644 index 0000000..aff9422 --- /dev/null +++ b/vendor/github.com/klauspost/compress/gen.sh @@ -0,0 +1,4 @@ +#!/bin/sh + +cd s2/cmd/_s2sx/ || exit 1 +go generate . diff --git a/vendor/github.com/klauspost/compress/huff0/.gitignore b/vendor/github.com/klauspost/compress/huff0/.gitignore new file mode 100644 index 0000000..b3d2629 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/.gitignore @@ -0,0 +1 @@ +/huff0-fuzz.zip diff --git a/vendor/github.com/klauspost/compress/huff0/README.md b/vendor/github.com/klauspost/compress/huff0/README.md new file mode 100644 index 0000000..8b6e5c6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/README.md @@ -0,0 +1,89 @@ +# Huff0 entropy compression + +This package provides Huff0 encoding and decoding as used in zstd. + +[Huff0](https://github.com/Cyan4973/FiniteStateEntropy#new-generation-entropy-coders), +a Huffman codec designed for modern CPU, featuring OoO (Out of Order) operations on multiple ALU +(Arithmetic Logic Unit), achieving extremely fast compression and decompression speeds. + +This can be used for compressing input with a lot of similar input values to the smallest number of bytes. +This does not perform any multi-byte [dictionary coding](https://en.wikipedia.org/wiki/Dictionary_coder) as LZ coders, +but it can be used as a secondary step to compressors (like Snappy) that does not do entropy encoding. + +* [Godoc documentation](https://godoc.org/github.com/klauspost/compress/huff0) + +## News + +This is used as part of the [zstandard](https://github.com/klauspost/compress/tree/master/zstd#zstd) compression and decompression package. + +This ensures that most functionality is well tested. + +# Usage + +This package provides a low level interface that allows to compress single independent blocks. + +Each block is separate, and there is no built in integrity checks. +This means that the caller should keep track of block sizes and also do checksums if needed. + +Compressing a block is done via the [`Compress1X`](https://godoc.org/github.com/klauspost/compress/huff0#Compress1X) and +[`Compress4X`](https://godoc.org/github.com/klauspost/compress/huff0#Compress4X) functions. +You must provide input and will receive the output and maybe an error. + +These error values can be returned: + +| Error | Description | +|---------------------|-----------------------------------------------------------------------------| +| `` | Everything ok, output is returned | +| `ErrIncompressible` | Returned when input is judged to be too hard to compress | +| `ErrUseRLE` | Returned from the compressor when the input is a single byte value repeated | +| `ErrTooBig` | Returned if the input block exceeds the maximum allowed size (128 Kib) | +| `(error)` | An internal error occurred. | + + +As can be seen above some of there are errors that will be returned even under normal operation so it is important to handle these. + +To reduce allocations you can provide a [`Scratch`](https://godoc.org/github.com/klauspost/compress/huff0#Scratch) object +that can be re-used for successive calls. Both compression and decompression accepts a `Scratch` object, and the same +object can be used for both. + +Be aware, that when re-using a `Scratch` object that the *output* buffer is also re-used, so if you are still using this +you must set the `Out` field in the scratch to nil. The same buffer is used for compression and decompression output. + +The `Scratch` object will retain state that allows to re-use previous tables for encoding and decoding. + +## Tables and re-use + +Huff0 allows for reusing tables from the previous block to save space if that is expected to give better/faster results. + +The Scratch object allows you to set a [`ReusePolicy`](https://godoc.org/github.com/klauspost/compress/huff0#ReusePolicy) +that controls this behaviour. See the documentation for details. This can be altered between each block. + +Do however note that this information is *not* stored in the output block and it is up to the users of the package to +record whether [`ReadTable`](https://godoc.org/github.com/klauspost/compress/huff0#ReadTable) should be called, +based on the boolean reported back from the CompressXX call. + +If you want to store the table separate from the data, you can access them as `OutData` and `OutTable` on the +[`Scratch`](https://godoc.org/github.com/klauspost/compress/huff0#Scratch) object. + +## Decompressing + +The first part of decoding is to initialize the decoding table through [`ReadTable`](https://godoc.org/github.com/klauspost/compress/huff0#ReadTable). +This will initialize the decoding tables. +You can supply the complete block to `ReadTable` and it will return the data part of the block +which can be given to the decompressor. + +Decompressing is done by calling the [`Decompress1X`](https://godoc.org/github.com/klauspost/compress/huff0#Scratch.Decompress1X) +or [`Decompress4X`](https://godoc.org/github.com/klauspost/compress/huff0#Scratch.Decompress4X) function. + +For concurrently decompressing content with a fixed table a stateless [`Decoder`](https://godoc.org/github.com/klauspost/compress/huff0#Decoder) can be requested which will remain correct as long as the scratch is unchanged. The capacity of the provided slice indicates the expected output size. + +You must provide the output from the compression stage, at exactly the size you got back. If you receive an error back +your input was likely corrupted. + +It is important to note that a successful decoding does *not* mean your output matches your original input. +There are no integrity checks, so relying on errors from the decompressor does not assure your data is valid. + +# Contributing + +Contributions are always welcome. Be aware that adding public functions will require good justification and breaking +changes will likely not be accepted. If in doubt open an issue before writing the PR. diff --git a/vendor/github.com/klauspost/compress/huff0/autogen.go b/vendor/github.com/klauspost/compress/huff0/autogen.go new file mode 100644 index 0000000..ff2c69d --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/autogen.go @@ -0,0 +1,5 @@ +package huff0 + +//go:generate go run generate.go +//go:generate asmfmt -w decompress_amd64.s +//go:generate asmfmt -w decompress_8b_amd64.s diff --git a/vendor/github.com/klauspost/compress/huff0/bitreader.go b/vendor/github.com/klauspost/compress/huff0/bitreader.go new file mode 100644 index 0000000..451160e --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/bitreader.go @@ -0,0 +1,243 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package huff0 + +import ( + "encoding/binary" + "errors" + "fmt" + "io" +) + +// bitReader reads a bitstream in reverse. +// The last set bit indicates the start of the stream and is used +// for aligning the input. +type bitReaderBytes struct { + in []byte + off uint // next byte to read is at in[off - 1] + value uint64 + bitsRead uint8 +} + +// init initializes and resets the bit reader. +func (b *bitReaderBytes) init(in []byte) error { + if len(in) < 1 { + return errors.New("corrupt stream: too short") + } + b.in = in + b.off = uint(len(in)) + // The highest bit of the last byte indicates where to start + v := in[len(in)-1] + if v == 0 { + return errors.New("corrupt stream, did not find end of stream") + } + b.bitsRead = 64 + b.value = 0 + if len(in) >= 8 { + b.fillFastStart() + } else { + b.fill() + b.fill() + } + b.advance(8 - uint8(highBit32(uint32(v)))) + return nil +} + +// peekBitsFast requires that at least one bit is requested every time. +// There are no checks if the buffer is filled. +func (b *bitReaderBytes) peekByteFast() uint8 { + got := uint8(b.value >> 56) + return got +} + +func (b *bitReaderBytes) advance(n uint8) { + b.bitsRead += n + b.value <<= n & 63 +} + +// fillFast() will make sure at least 32 bits are available. +// There must be at least 4 bytes available. +func (b *bitReaderBytes) fillFast() { + if b.bitsRead < 32 { + return + } + + // 2 bounds checks. + v := b.in[b.off-4 : b.off] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value |= uint64(low) << (b.bitsRead - 32) + b.bitsRead -= 32 + b.off -= 4 +} + +// fillFastStart() assumes the bitReaderBytes is empty and there is at least 8 bytes to read. +func (b *bitReaderBytes) fillFastStart() { + // Do single re-slice to avoid bounds checks. + b.value = binary.LittleEndian.Uint64(b.in[b.off-8:]) + b.bitsRead = 0 + b.off -= 8 +} + +// fill() will make sure at least 32 bits are available. +func (b *bitReaderBytes) fill() { + if b.bitsRead < 32 { + return + } + if b.off > 4 { + v := b.in[b.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value |= uint64(low) << (b.bitsRead - 32) + b.bitsRead -= 32 + b.off -= 4 + return + } + for b.off > 0 { + b.value |= uint64(b.in[b.off-1]) << (b.bitsRead - 8) + b.bitsRead -= 8 + b.off-- + } +} + +// finished returns true if all bits have been read from the bit stream. +func (b *bitReaderBytes) finished() bool { + return b.off == 0 && b.bitsRead >= 64 +} + +func (b *bitReaderBytes) remaining() uint { + return b.off*8 + uint(64-b.bitsRead) +} + +// close the bitstream and returns an error if out-of-buffer reads occurred. +func (b *bitReaderBytes) close() error { + // Release reference. + b.in = nil + if b.remaining() > 0 { + return fmt.Errorf("corrupt input: %d bits remain on stream", b.remaining()) + } + if b.bitsRead > 64 { + return io.ErrUnexpectedEOF + } + return nil +} + +// bitReaderShifted reads a bitstream in reverse. +// The last set bit indicates the start of the stream and is used +// for aligning the input. +type bitReaderShifted struct { + in []byte + off uint // next byte to read is at in[off - 1] + value uint64 + bitsRead uint8 +} + +// init initializes and resets the bit reader. +func (b *bitReaderShifted) init(in []byte) error { + if len(in) < 1 { + return errors.New("corrupt stream: too short") + } + b.in = in + b.off = uint(len(in)) + // The highest bit of the last byte indicates where to start + v := in[len(in)-1] + if v == 0 { + return errors.New("corrupt stream, did not find end of stream") + } + b.bitsRead = 64 + b.value = 0 + if len(in) >= 8 { + b.fillFastStart() + } else { + b.fill() + b.fill() + } + b.advance(8 - uint8(highBit32(uint32(v)))) + return nil +} + +// peekBitsFast requires that at least one bit is requested every time. +// There are no checks if the buffer is filled. +func (b *bitReaderShifted) peekBitsFast(n uint8) uint16 { + return uint16(b.value >> ((64 - n) & 63)) +} + +// peekTopBits(n) is equvialent to peekBitFast(64 - n) +func (b *bitReaderShifted) peekTopBits(n uint8) uint16 { + return uint16(b.value >> n) +} + +func (b *bitReaderShifted) advance(n uint8) { + b.bitsRead += n + b.value <<= n & 63 +} + +// fillFast() will make sure at least 32 bits are available. +// There must be at least 4 bytes available. +func (b *bitReaderShifted) fillFast() { + if b.bitsRead < 32 { + return + } + + // 2 bounds checks. + v := b.in[b.off-4 : b.off] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value |= uint64(low) << ((b.bitsRead - 32) & 63) + b.bitsRead -= 32 + b.off -= 4 +} + +// fillFastStart() assumes the bitReaderShifted is empty and there is at least 8 bytes to read. +func (b *bitReaderShifted) fillFastStart() { + // Do single re-slice to avoid bounds checks. + b.value = binary.LittleEndian.Uint64(b.in[b.off-8:]) + b.bitsRead = 0 + b.off -= 8 +} + +// fill() will make sure at least 32 bits are available. +func (b *bitReaderShifted) fill() { + if b.bitsRead < 32 { + return + } + if b.off > 4 { + v := b.in[b.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value |= uint64(low) << ((b.bitsRead - 32) & 63) + b.bitsRead -= 32 + b.off -= 4 + return + } + for b.off > 0 { + b.value |= uint64(b.in[b.off-1]) << ((b.bitsRead - 8) & 63) + b.bitsRead -= 8 + b.off-- + } +} + +// finished returns true if all bits have been read from the bit stream. +func (b *bitReaderShifted) finished() bool { + return b.off == 0 && b.bitsRead >= 64 +} + +func (b *bitReaderShifted) remaining() uint { + return b.off*8 + uint(64-b.bitsRead) +} + +// close the bitstream and returns an error if out-of-buffer reads occurred. +func (b *bitReaderShifted) close() error { + // Release reference. + b.in = nil + if b.remaining() > 0 { + return fmt.Errorf("corrupt input: %d bits remain on stream", b.remaining()) + } + if b.bitsRead > 64 { + return io.ErrUnexpectedEOF + } + return nil +} diff --git a/vendor/github.com/klauspost/compress/huff0/bitwriter.go b/vendor/github.com/klauspost/compress/huff0/bitwriter.go new file mode 100644 index 0000000..6bce4e8 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/bitwriter.go @@ -0,0 +1,210 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package huff0 + +import "fmt" + +// bitWriter will write bits. +// First bit will be LSB of the first byte of output. +type bitWriter struct { + bitContainer uint64 + nBits uint8 + out []byte +} + +// bitMask16 is bitmasks. Has extra to avoid bounds check. +var bitMask16 = [32]uint16{ + 0, 1, 3, 7, 0xF, 0x1F, + 0x3F, 0x7F, 0xFF, 0x1FF, 0x3FF, 0x7FF, + 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0xFFFF, + 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, + 0xFFFF, 0xFFFF} /* up to 16 bits */ + +// addBits16NC will add up to 16 bits. +// It will not check if there is space for them, +// so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits16NC(value uint16, bits uint8) { + b.bitContainer |= uint64(value&bitMask16[bits&31]) << (b.nBits & 63) + b.nBits += bits +} + +// addBits16Clean will add up to 16 bits. value may not contain more set bits than indicated. +// It will not check if there is space for them, so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits16Clean(value uint16, bits uint8) { + b.bitContainer |= uint64(value) << (b.nBits & 63) + b.nBits += bits +} + +// encSymbol will add up to 16 bits. value may not contain more set bits than indicated. +// It will not check if there is space for them, so the caller must ensure that it has flushed recently. +func (b *bitWriter) encSymbol(ct cTable, symbol byte) { + enc := ct[symbol] + b.bitContainer |= uint64(enc.val) << (b.nBits & 63) + if false { + if enc.nBits == 0 { + panic("nbits 0") + } + } + b.nBits += enc.nBits +} + +// encTwoSymbols will add up to 32 bits. value may not contain more set bits than indicated. +// It will not check if there is space for them, so the caller must ensure that it has flushed recently. +func (b *bitWriter) encTwoSymbols(ct cTable, av, bv byte) { + encA := ct[av] + encB := ct[bv] + sh := b.nBits & 63 + combined := uint64(encA.val) | (uint64(encB.val) << (encA.nBits & 63)) + b.bitContainer |= combined << sh + if false { + if encA.nBits == 0 { + panic("nbitsA 0") + } + if encB.nBits == 0 { + panic("nbitsB 0") + } + } + b.nBits += encA.nBits + encB.nBits +} + +// addBits16ZeroNC will add up to 16 bits. +// It will not check if there is space for them, +// so the caller must ensure that it has flushed recently. +// This is fastest if bits can be zero. +func (b *bitWriter) addBits16ZeroNC(value uint16, bits uint8) { + if bits == 0 { + return + } + value <<= (16 - bits) & 15 + value >>= (16 - bits) & 15 + b.bitContainer |= uint64(value) << (b.nBits & 63) + b.nBits += bits +} + +// flush will flush all pending full bytes. +// There will be at least 56 bits available for writing when this has been called. +// Using flush32 is faster, but leaves less space for writing. +func (b *bitWriter) flush() { + v := b.nBits >> 3 + switch v { + case 0: + return + case 1: + b.out = append(b.out, + byte(b.bitContainer), + ) + b.bitContainer >>= 1 << 3 + case 2: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + ) + b.bitContainer >>= 2 << 3 + case 3: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + ) + b.bitContainer >>= 3 << 3 + case 4: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + ) + b.bitContainer >>= 4 << 3 + case 5: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + ) + b.bitContainer >>= 5 << 3 + case 6: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + ) + b.bitContainer >>= 6 << 3 + case 7: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + byte(b.bitContainer>>48), + ) + b.bitContainer >>= 7 << 3 + case 8: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + byte(b.bitContainer>>48), + byte(b.bitContainer>>56), + ) + b.bitContainer = 0 + b.nBits = 0 + return + default: + panic(fmt.Errorf("bits (%d) > 64", b.nBits)) + } + b.nBits &= 7 +} + +// flush32 will flush out, so there are at least 32 bits available for writing. +func (b *bitWriter) flush32() { + if b.nBits < 32 { + return + } + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24)) + b.nBits -= 32 + b.bitContainer >>= 32 +} + +// flushAlign will flush remaining full bytes and align to next byte boundary. +func (b *bitWriter) flushAlign() { + nbBytes := (b.nBits + 7) >> 3 + for i := uint8(0); i < nbBytes; i++ { + b.out = append(b.out, byte(b.bitContainer>>(i*8))) + } + b.nBits = 0 + b.bitContainer = 0 +} + +// close will write the alignment bit and write the final byte(s) +// to the output. +func (b *bitWriter) close() error { + // End mark + b.addBits16Clean(1, 1) + // flush until next byte. + b.flushAlign() + return nil +} + +// reset and continue writing by appending to out. +func (b *bitWriter) reset(out []byte) { + b.bitContainer = 0 + b.nBits = 0 + b.out = out +} diff --git a/vendor/github.com/klauspost/compress/huff0/bytereader.go b/vendor/github.com/klauspost/compress/huff0/bytereader.go new file mode 100644 index 0000000..50bcdf6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/bytereader.go @@ -0,0 +1,54 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package huff0 + +// byteReader provides a byte reader that reads +// little endian values from a byte stream. +// The input stream is manually advanced. +// The reader performs no bounds checks. +type byteReader struct { + b []byte + off int +} + +// init will initialize the reader and set the input. +func (b *byteReader) init(in []byte) { + b.b = in + b.off = 0 +} + +// advance the stream b n bytes. +func (b *byteReader) advance(n uint) { + b.off += int(n) +} + +// Int32 returns a little endian int32 starting at current offset. +func (b byteReader) Int32() int32 { + v3 := int32(b.b[b.off+3]) + v2 := int32(b.b[b.off+2]) + v1 := int32(b.b[b.off+1]) + v0 := int32(b.b[b.off]) + return (v3 << 24) | (v2 << 16) | (v1 << 8) | v0 +} + +// Uint32 returns a little endian uint32 starting at current offset. +func (b byteReader) Uint32() uint32 { + v3 := uint32(b.b[b.off+3]) + v2 := uint32(b.b[b.off+2]) + v1 := uint32(b.b[b.off+1]) + v0 := uint32(b.b[b.off]) + return (v3 << 24) | (v2 << 16) | (v1 << 8) | v0 +} + +// unread returns the unread portion of the input. +func (b byteReader) unread() []byte { + return b.b[b.off:] +} + +// remain will return the number of bytes remaining. +func (b byteReader) remain() int { + return len(b.b) - b.off +} diff --git a/vendor/github.com/klauspost/compress/huff0/compress.go b/vendor/github.com/klauspost/compress/huff0/compress.go new file mode 100644 index 0000000..bc95ac6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/compress.go @@ -0,0 +1,729 @@ +package huff0 + +import ( + "fmt" + "math" + "runtime" + "sync" +) + +// Compress1X will compress the input. +// The output can be decoded using Decompress1X. +// Supply a Scratch object. The scratch object contains state about re-use, +// So when sharing across independent encodes, be sure to set the re-use policy. +func Compress1X(in []byte, s *Scratch) (out []byte, reUsed bool, err error) { + s, err = s.prepare(in) + if err != nil { + return nil, false, err + } + return compress(in, s, s.compress1X) +} + +// Compress4X will compress the input. The input is split into 4 independent blocks +// and compressed similar to Compress1X. +// The output can be decoded using Decompress4X. +// Supply a Scratch object. The scratch object contains state about re-use, +// So when sharing across independent encodes, be sure to set the re-use policy. +func Compress4X(in []byte, s *Scratch) (out []byte, reUsed bool, err error) { + s, err = s.prepare(in) + if err != nil { + return nil, false, err + } + if false { + // TODO: compress4Xp only slightly faster. + const parallelThreshold = 8 << 10 + if len(in) < parallelThreshold || runtime.GOMAXPROCS(0) == 1 { + return compress(in, s, s.compress4X) + } + return compress(in, s, s.compress4Xp) + } + return compress(in, s, s.compress4X) +} + +func compress(in []byte, s *Scratch, compressor func(src []byte) ([]byte, error)) (out []byte, reUsed bool, err error) { + // Nuke previous table if we cannot reuse anyway. + if s.Reuse == ReusePolicyNone { + s.prevTable = s.prevTable[:0] + } + + // Create histogram, if none was provided. + maxCount := s.maxCount + var canReuse = false + if maxCount == 0 { + maxCount, canReuse = s.countSimple(in) + } else { + canReuse = s.canUseTable(s.prevTable) + } + + // We want the output size to be less than this: + wantSize := len(in) + if s.WantLogLess > 0 { + wantSize -= wantSize >> s.WantLogLess + } + + // Reset for next run. + s.clearCount = true + s.maxCount = 0 + if maxCount >= len(in) { + if maxCount > len(in) { + return nil, false, fmt.Errorf("maxCount (%d) > length (%d)", maxCount, len(in)) + } + if len(in) == 1 { + return nil, false, ErrIncompressible + } + // One symbol, use RLE + return nil, false, ErrUseRLE + } + if maxCount == 1 || maxCount < (len(in)>>7) { + // Each symbol present maximum once or too well distributed. + return nil, false, ErrIncompressible + } + if s.Reuse == ReusePolicyMust && !canReuse { + // We must reuse, but we can't. + return nil, false, ErrIncompressible + } + if (s.Reuse == ReusePolicyPrefer || s.Reuse == ReusePolicyMust) && canReuse { + keepTable := s.cTable + keepTL := s.actualTableLog + s.cTable = s.prevTable + s.actualTableLog = s.prevTableLog + s.Out, err = compressor(in) + s.cTable = keepTable + s.actualTableLog = keepTL + if err == nil && len(s.Out) < wantSize { + s.OutData = s.Out + return s.Out, true, nil + } + if s.Reuse == ReusePolicyMust { + return nil, false, ErrIncompressible + } + // Do not attempt to re-use later. + s.prevTable = s.prevTable[:0] + } + + // Calculate new table. + err = s.buildCTable() + if err != nil { + return nil, false, err + } + + if false && !s.canUseTable(s.cTable) { + panic("invalid table generated") + } + + if s.Reuse == ReusePolicyAllow && canReuse { + hSize := len(s.Out) + oldSize := s.prevTable.estimateSize(s.count[:s.symbolLen]) + newSize := s.cTable.estimateSize(s.count[:s.symbolLen]) + if oldSize <= hSize+newSize || hSize+12 >= wantSize { + // Retain cTable even if we re-use. + keepTable := s.cTable + keepTL := s.actualTableLog + + s.cTable = s.prevTable + s.actualTableLog = s.prevTableLog + s.Out, err = compressor(in) + + // Restore ctable. + s.cTable = keepTable + s.actualTableLog = keepTL + if err != nil { + return nil, false, err + } + if len(s.Out) >= wantSize { + return nil, false, ErrIncompressible + } + s.OutData = s.Out + return s.Out, true, nil + } + } + + // Use new table + err = s.cTable.write(s) + if err != nil { + s.OutTable = nil + return nil, false, err + } + s.OutTable = s.Out + + // Compress using new table + s.Out, err = compressor(in) + if err != nil { + s.OutTable = nil + return nil, false, err + } + if len(s.Out) >= wantSize { + s.OutTable = nil + return nil, false, ErrIncompressible + } + // Move current table into previous. + s.prevTable, s.prevTableLog, s.cTable = s.cTable, s.actualTableLog, s.prevTable[:0] + s.OutData = s.Out[len(s.OutTable):] + return s.Out, false, nil +} + +// EstimateSizes will estimate the data sizes +func EstimateSizes(in []byte, s *Scratch) (tableSz, dataSz, reuseSz int, err error) { + s, err = s.prepare(in) + if err != nil { + return 0, 0, 0, err + } + + // Create histogram, if none was provided. + tableSz, dataSz, reuseSz = -1, -1, -1 + maxCount := s.maxCount + var canReuse = false + if maxCount == 0 { + maxCount, canReuse = s.countSimple(in) + } else { + canReuse = s.canUseTable(s.prevTable) + } + + // We want the output size to be less than this: + wantSize := len(in) + if s.WantLogLess > 0 { + wantSize -= wantSize >> s.WantLogLess + } + + // Reset for next run. + s.clearCount = true + s.maxCount = 0 + if maxCount >= len(in) { + if maxCount > len(in) { + return 0, 0, 0, fmt.Errorf("maxCount (%d) > length (%d)", maxCount, len(in)) + } + if len(in) == 1 { + return 0, 0, 0, ErrIncompressible + } + // One symbol, use RLE + return 0, 0, 0, ErrUseRLE + } + if maxCount == 1 || maxCount < (len(in)>>7) { + // Each symbol present maximum once or too well distributed. + return 0, 0, 0, ErrIncompressible + } + + // Calculate new table. + err = s.buildCTable() + if err != nil { + return 0, 0, 0, err + } + + if false && !s.canUseTable(s.cTable) { + panic("invalid table generated") + } + + tableSz, err = s.cTable.estTableSize(s) + if err != nil { + return 0, 0, 0, err + } + if canReuse { + reuseSz = s.prevTable.estimateSize(s.count[:s.symbolLen]) + } + dataSz = s.cTable.estimateSize(s.count[:s.symbolLen]) + + // Restore + return tableSz, dataSz, reuseSz, nil +} + +func (s *Scratch) compress1X(src []byte) ([]byte, error) { + return s.compress1xDo(s.Out, src) +} + +func (s *Scratch) compress1xDo(dst, src []byte) ([]byte, error) { + var bw = bitWriter{out: dst} + + // N is length divisible by 4. + n := len(src) + n -= n & 3 + cTable := s.cTable[:256] + + // Encode last bytes. + for i := len(src) & 3; i > 0; i-- { + bw.encSymbol(cTable, src[n+i-1]) + } + n -= 4 + if s.actualTableLog <= 8 { + for ; n >= 0; n -= 4 { + tmp := src[n : n+4] + // tmp should be len 4 + bw.flush32() + bw.encTwoSymbols(cTable, tmp[3], tmp[2]) + bw.encTwoSymbols(cTable, tmp[1], tmp[0]) + } + } else { + for ; n >= 0; n -= 4 { + tmp := src[n : n+4] + // tmp should be len 4 + bw.flush32() + bw.encTwoSymbols(cTable, tmp[3], tmp[2]) + bw.flush32() + bw.encTwoSymbols(cTable, tmp[1], tmp[0]) + } + } + err := bw.close() + return bw.out, err +} + +var sixZeros [6]byte + +func (s *Scratch) compress4X(src []byte) ([]byte, error) { + if len(src) < 12 { + return nil, ErrIncompressible + } + segmentSize := (len(src) + 3) / 4 + + // Add placeholder for output length + offsetIdx := len(s.Out) + s.Out = append(s.Out, sixZeros[:]...) + + for i := 0; i < 4; i++ { + toDo := src + if len(toDo) > segmentSize { + toDo = toDo[:segmentSize] + } + src = src[len(toDo):] + + var err error + idx := len(s.Out) + s.Out, err = s.compress1xDo(s.Out, toDo) + if err != nil { + return nil, err + } + if len(s.Out)-idx > math.MaxUint16 { + // We cannot store the size in the jump table + return nil, ErrIncompressible + } + // Write compressed length as little endian before block. + if i < 3 { + // Last length is not written. + length := len(s.Out) - idx + s.Out[i*2+offsetIdx] = byte(length) + s.Out[i*2+offsetIdx+1] = byte(length >> 8) + } + } + + return s.Out, nil +} + +// compress4Xp will compress 4 streams using separate goroutines. +func (s *Scratch) compress4Xp(src []byte) ([]byte, error) { + if len(src) < 12 { + return nil, ErrIncompressible + } + // Add placeholder for output length + s.Out = s.Out[:6] + + segmentSize := (len(src) + 3) / 4 + var wg sync.WaitGroup + var errs [4]error + wg.Add(4) + for i := 0; i < 4; i++ { + toDo := src + if len(toDo) > segmentSize { + toDo = toDo[:segmentSize] + } + src = src[len(toDo):] + + // Separate goroutine for each block. + go func(i int) { + s.tmpOut[i], errs[i] = s.compress1xDo(s.tmpOut[i][:0], toDo) + wg.Done() + }(i) + } + wg.Wait() + for i := 0; i < 4; i++ { + if errs[i] != nil { + return nil, errs[i] + } + o := s.tmpOut[i] + if len(o) > math.MaxUint16 { + // We cannot store the size in the jump table + return nil, ErrIncompressible + } + // Write compressed length as little endian before block. + if i < 3 { + // Last length is not written. + s.Out[i*2] = byte(len(o)) + s.Out[i*2+1] = byte(len(o) >> 8) + } + + // Write output. + s.Out = append(s.Out, o...) + } + return s.Out, nil +} + +// countSimple will create a simple histogram in s.count. +// Returns the biggest count. +// Does not update s.clearCount. +func (s *Scratch) countSimple(in []byte) (max int, reuse bool) { + reuse = true + for _, v := range in { + s.count[v]++ + } + m := uint32(0) + if len(s.prevTable) > 0 { + for i, v := range s.count[:] { + if v > m { + m = v + } + if v > 0 { + s.symbolLen = uint16(i) + 1 + if i >= len(s.prevTable) { + reuse = false + } else { + if s.prevTable[i].nBits == 0 { + reuse = false + } + } + } + } + return int(m), reuse + } + for i, v := range s.count[:] { + if v > m { + m = v + } + if v > 0 { + s.symbolLen = uint16(i) + 1 + } + } + return int(m), false +} + +func (s *Scratch) canUseTable(c cTable) bool { + if len(c) < int(s.symbolLen) { + return false + } + for i, v := range s.count[:s.symbolLen] { + if v != 0 && c[i].nBits == 0 { + return false + } + } + return true +} + +func (s *Scratch) validateTable(c cTable) bool { + if len(c) < int(s.symbolLen) { + return false + } + for i, v := range s.count[:s.symbolLen] { + if v != 0 { + if c[i].nBits == 0 { + return false + } + if c[i].nBits > s.actualTableLog { + return false + } + } + } + return true +} + +// minTableLog provides the minimum logSize to safely represent a distribution. +func (s *Scratch) minTableLog() uint8 { + minBitsSrc := highBit32(uint32(s.br.remain())) + 1 + minBitsSymbols := highBit32(uint32(s.symbolLen-1)) + 2 + if minBitsSrc < minBitsSymbols { + return uint8(minBitsSrc) + } + return uint8(minBitsSymbols) +} + +// optimalTableLog calculates and sets the optimal tableLog in s.actualTableLog +func (s *Scratch) optimalTableLog() { + tableLog := s.TableLog + minBits := s.minTableLog() + maxBitsSrc := uint8(highBit32(uint32(s.br.remain()-1))) - 1 + if maxBitsSrc < tableLog { + // Accuracy can be reduced + tableLog = maxBitsSrc + } + if minBits > tableLog { + tableLog = minBits + } + // Need a minimum to safely represent all symbol values + if tableLog < minTablelog { + tableLog = minTablelog + } + if tableLog > tableLogMax { + tableLog = tableLogMax + } + s.actualTableLog = tableLog +} + +type cTableEntry struct { + val uint16 + nBits uint8 + // We have 8 bits extra +} + +const huffNodesMask = huffNodesLen - 1 + +func (s *Scratch) buildCTable() error { + s.optimalTableLog() + s.huffSort() + if cap(s.cTable) < maxSymbolValue+1 { + s.cTable = make([]cTableEntry, s.symbolLen, maxSymbolValue+1) + } else { + s.cTable = s.cTable[:s.symbolLen] + for i := range s.cTable { + s.cTable[i] = cTableEntry{} + } + } + + var startNode = int16(s.symbolLen) + nonNullRank := s.symbolLen - 1 + + nodeNb := startNode + huffNode := s.nodes[1 : huffNodesLen+1] + + // This overlays the slice above, but allows "-1" index lookups. + // Different from reference implementation. + huffNode0 := s.nodes[0 : huffNodesLen+1] + + for huffNode[nonNullRank].count == 0 { + nonNullRank-- + } + + lowS := int16(nonNullRank) + nodeRoot := nodeNb + lowS - 1 + lowN := nodeNb + huffNode[nodeNb].count = huffNode[lowS].count + huffNode[lowS-1].count + huffNode[lowS].parent, huffNode[lowS-1].parent = uint16(nodeNb), uint16(nodeNb) + nodeNb++ + lowS -= 2 + for n := nodeNb; n <= nodeRoot; n++ { + huffNode[n].count = 1 << 30 + } + // fake entry, strong barrier + huffNode0[0].count = 1 << 31 + + // create parents + for nodeNb <= nodeRoot { + var n1, n2 int16 + if huffNode0[lowS+1].count < huffNode0[lowN+1].count { + n1 = lowS + lowS-- + } else { + n1 = lowN + lowN++ + } + if huffNode0[lowS+1].count < huffNode0[lowN+1].count { + n2 = lowS + lowS-- + } else { + n2 = lowN + lowN++ + } + + huffNode[nodeNb].count = huffNode0[n1+1].count + huffNode0[n2+1].count + huffNode0[n1+1].parent, huffNode0[n2+1].parent = uint16(nodeNb), uint16(nodeNb) + nodeNb++ + } + + // distribute weights (unlimited tree height) + huffNode[nodeRoot].nbBits = 0 + for n := nodeRoot - 1; n >= startNode; n-- { + huffNode[n].nbBits = huffNode[huffNode[n].parent].nbBits + 1 + } + for n := uint16(0); n <= nonNullRank; n++ { + huffNode[n].nbBits = huffNode[huffNode[n].parent].nbBits + 1 + } + s.actualTableLog = s.setMaxHeight(int(nonNullRank)) + maxNbBits := s.actualTableLog + + // fill result into tree (val, nbBits) + if maxNbBits > tableLogMax { + return fmt.Errorf("internal error: maxNbBits (%d) > tableLogMax (%d)", maxNbBits, tableLogMax) + } + var nbPerRank [tableLogMax + 1]uint16 + var valPerRank [16]uint16 + for _, v := range huffNode[:nonNullRank+1] { + nbPerRank[v.nbBits]++ + } + // determine stating value per rank + { + min := uint16(0) + for n := maxNbBits; n > 0; n-- { + // get starting value within each rank + valPerRank[n] = min + min += nbPerRank[n] + min >>= 1 + } + } + + // push nbBits per symbol, symbol order + for _, v := range huffNode[:nonNullRank+1] { + s.cTable[v.symbol].nBits = v.nbBits + } + + // assign value within rank, symbol order + t := s.cTable[:s.symbolLen] + for n, val := range t { + nbits := val.nBits & 15 + v := valPerRank[nbits] + t[n].val = v + valPerRank[nbits] = v + 1 + } + + return nil +} + +// huffSort will sort symbols, decreasing order. +func (s *Scratch) huffSort() { + type rankPos struct { + base uint32 + current uint32 + } + + // Clear nodes + nodes := s.nodes[:huffNodesLen+1] + s.nodes = nodes + nodes = nodes[1 : huffNodesLen+1] + + // Sort into buckets based on length of symbol count. + var rank [32]rankPos + for _, v := range s.count[:s.symbolLen] { + r := highBit32(v+1) & 31 + rank[r].base++ + } + // maxBitLength is log2(BlockSizeMax) + 1 + const maxBitLength = 18 + 1 + for n := maxBitLength; n > 0; n-- { + rank[n-1].base += rank[n].base + } + for n := range rank[:maxBitLength] { + rank[n].current = rank[n].base + } + for n, c := range s.count[:s.symbolLen] { + r := (highBit32(c+1) + 1) & 31 + pos := rank[r].current + rank[r].current++ + prev := nodes[(pos-1)&huffNodesMask] + for pos > rank[r].base && c > prev.count { + nodes[pos&huffNodesMask] = prev + pos-- + prev = nodes[(pos-1)&huffNodesMask] + } + nodes[pos&huffNodesMask] = nodeElt{count: c, symbol: byte(n)} + } +} + +func (s *Scratch) setMaxHeight(lastNonNull int) uint8 { + maxNbBits := s.actualTableLog + huffNode := s.nodes[1 : huffNodesLen+1] + //huffNode = huffNode[: huffNodesLen] + + largestBits := huffNode[lastNonNull].nbBits + + // early exit : no elt > maxNbBits + if largestBits <= maxNbBits { + return largestBits + } + totalCost := int(0) + baseCost := int(1) << (largestBits - maxNbBits) + n := uint32(lastNonNull) + + for huffNode[n].nbBits > maxNbBits { + totalCost += baseCost - (1 << (largestBits - huffNode[n].nbBits)) + huffNode[n].nbBits = maxNbBits + n-- + } + // n stops at huffNode[n].nbBits <= maxNbBits + + for huffNode[n].nbBits == maxNbBits { + n-- + } + // n end at index of smallest symbol using < maxNbBits + + // renorm totalCost + totalCost >>= largestBits - maxNbBits /* note : totalCost is necessarily a multiple of baseCost */ + + // repay normalized cost + { + const noSymbol = 0xF0F0F0F0 + var rankLast [tableLogMax + 2]uint32 + + for i := range rankLast[:] { + rankLast[i] = noSymbol + } + + // Get pos of last (smallest) symbol per rank + { + currentNbBits := maxNbBits + for pos := int(n); pos >= 0; pos-- { + if huffNode[pos].nbBits >= currentNbBits { + continue + } + currentNbBits = huffNode[pos].nbBits // < maxNbBits + rankLast[maxNbBits-currentNbBits] = uint32(pos) + } + } + + for totalCost > 0 { + nBitsToDecrease := uint8(highBit32(uint32(totalCost))) + 1 + + for ; nBitsToDecrease > 1; nBitsToDecrease-- { + highPos := rankLast[nBitsToDecrease] + lowPos := rankLast[nBitsToDecrease-1] + if highPos == noSymbol { + continue + } + if lowPos == noSymbol { + break + } + highTotal := huffNode[highPos].count + lowTotal := 2 * huffNode[lowPos].count + if highTotal <= lowTotal { + break + } + } + // only triggered when no more rank 1 symbol left => find closest one (note : there is necessarily at least one !) + // HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary + // FIXME: try to remove + for (nBitsToDecrease <= tableLogMax) && (rankLast[nBitsToDecrease] == noSymbol) { + nBitsToDecrease++ + } + totalCost -= 1 << (nBitsToDecrease - 1) + if rankLast[nBitsToDecrease-1] == noSymbol { + // this rank is no longer empty + rankLast[nBitsToDecrease-1] = rankLast[nBitsToDecrease] + } + huffNode[rankLast[nBitsToDecrease]].nbBits++ + if rankLast[nBitsToDecrease] == 0 { + /* special case, reached largest symbol */ + rankLast[nBitsToDecrease] = noSymbol + } else { + rankLast[nBitsToDecrease]-- + if huffNode[rankLast[nBitsToDecrease]].nbBits != maxNbBits-nBitsToDecrease { + rankLast[nBitsToDecrease] = noSymbol /* this rank is now empty */ + } + } + } + + for totalCost < 0 { /* Sometimes, cost correction overshoot */ + if rankLast[1] == noSymbol { /* special case : no rank 1 symbol (using maxNbBits-1); let's create one from largest rank 0 (using maxNbBits) */ + for huffNode[n].nbBits == maxNbBits { + n-- + } + huffNode[n+1].nbBits-- + rankLast[1] = n + 1 + totalCost++ + continue + } + huffNode[rankLast[1]+1].nbBits-- + rankLast[1]++ + totalCost++ + } + } + return maxNbBits +} + +type nodeElt struct { + count uint32 + parent uint16 + symbol byte + nbBits uint8 +} diff --git a/vendor/github.com/klauspost/compress/huff0/decompress.go b/vendor/github.com/klauspost/compress/huff0/decompress.go new file mode 100644 index 0000000..04f6529 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress.go @@ -0,0 +1,1270 @@ +package huff0 + +import ( + "errors" + "fmt" + "io" + "sync" + + "github.com/klauspost/compress/fse" +) + +type dTable struct { + single []dEntrySingle + double []dEntryDouble +} + +// single-symbols decoding +type dEntrySingle struct { + entry uint16 +} + +// double-symbols decoding +type dEntryDouble struct { + seq [4]byte + nBits uint8 + len uint8 +} + +// Uses special code for all tables that are < 8 bits. +const use8BitTables = true + +// ReadTable will read a table from the input. +// The size of the input may be larger than the table definition. +// Any content remaining after the table definition will be returned. +// If no Scratch is provided a new one is allocated. +// The returned Scratch can be used for encoding or decoding input using this table. +func ReadTable(in []byte, s *Scratch) (s2 *Scratch, remain []byte, err error) { + s, err = s.prepare(in) + if err != nil { + return s, nil, err + } + if len(in) <= 1 { + return s, nil, errors.New("input too small for table") + } + iSize := in[0] + in = in[1:] + if iSize >= 128 { + // Uncompressed + oSize := iSize - 127 + iSize = (oSize + 1) / 2 + if int(iSize) > len(in) { + return s, nil, errors.New("input too small for table") + } + for n := uint8(0); n < oSize; n += 2 { + v := in[n/2] + s.huffWeight[n] = v >> 4 + s.huffWeight[n+1] = v & 15 + } + s.symbolLen = uint16(oSize) + in = in[iSize:] + } else { + if len(in) < int(iSize) { + return s, nil, fmt.Errorf("input too small for table, want %d bytes, have %d", iSize, len(in)) + } + // FSE compressed weights + s.fse.DecompressLimit = 255 + hw := s.huffWeight[:] + s.fse.Out = hw + b, err := fse.Decompress(in[:iSize], s.fse) + s.fse.Out = nil + if err != nil { + return s, nil, err + } + if len(b) > 255 { + return s, nil, errors.New("corrupt input: output table too large") + } + s.symbolLen = uint16(len(b)) + in = in[iSize:] + } + + // collect weight stats + var rankStats [16]uint32 + weightTotal := uint32(0) + for _, v := range s.huffWeight[:s.symbolLen] { + if v > tableLogMax { + return s, nil, errors.New("corrupt input: weight too large") + } + v2 := v & 15 + rankStats[v2]++ + // (1 << (v2-1)) is slower since the compiler cannot prove that v2 isn't 0. + weightTotal += (1 << v2) >> 1 + } + if weightTotal == 0 { + return s, nil, errors.New("corrupt input: weights zero") + } + + // get last non-null symbol weight (implied, total must be 2^n) + { + tableLog := highBit32(weightTotal) + 1 + if tableLog > tableLogMax { + return s, nil, errors.New("corrupt input: tableLog too big") + } + s.actualTableLog = uint8(tableLog) + // determine last weight + { + total := uint32(1) << tableLog + rest := total - weightTotal + verif := uint32(1) << highBit32(rest) + lastWeight := highBit32(rest) + 1 + if verif != rest { + // last value must be a clean power of 2 + return s, nil, errors.New("corrupt input: last value not power of two") + } + s.huffWeight[s.symbolLen] = uint8(lastWeight) + s.symbolLen++ + rankStats[lastWeight]++ + } + } + + if (rankStats[1] < 2) || (rankStats[1]&1 != 0) { + // by construction : at least 2 elts of rank 1, must be even + return s, nil, errors.New("corrupt input: min elt size, even check failed ") + } + + // TODO: Choose between single/double symbol decoding + + // Calculate starting value for each rank + { + var nextRankStart uint32 + for n := uint8(1); n < s.actualTableLog+1; n++ { + current := nextRankStart + nextRankStart += rankStats[n] << (n - 1) + rankStats[n] = current + } + } + + // fill DTable (always full size) + tSize := 1 << tableLogMax + if len(s.dt.single) != tSize { + s.dt.single = make([]dEntrySingle, tSize) + } + cTable := s.prevTable + if cap(cTable) < maxSymbolValue+1 { + cTable = make([]cTableEntry, 0, maxSymbolValue+1) + } + cTable = cTable[:maxSymbolValue+1] + s.prevTable = cTable[:s.symbolLen] + s.prevTableLog = s.actualTableLog + + for n, w := range s.huffWeight[:s.symbolLen] { + if w == 0 { + cTable[n] = cTableEntry{ + val: 0, + nBits: 0, + } + continue + } + length := (uint32(1) << w) >> 1 + d := dEntrySingle{ + entry: uint16(s.actualTableLog+1-w) | (uint16(n) << 8), + } + + rank := &rankStats[w] + cTable[n] = cTableEntry{ + val: uint16(*rank >> (w - 1)), + nBits: uint8(d.entry), + } + + single := s.dt.single[*rank : *rank+length] + for i := range single { + single[i] = d + } + *rank += length + } + + return s, in, nil +} + +// Decompress1X will decompress a 1X encoded stream. +// The length of the supplied input must match the end of a block exactly. +// Before this is called, the table must be initialized with ReadTable unless +// the encoder re-used the table. +// deprecated: Use the stateless Decoder() to get a concurrent version. +func (s *Scratch) Decompress1X(in []byte) (out []byte, err error) { + if cap(s.Out) < s.MaxDecodedSize { + s.Out = make([]byte, s.MaxDecodedSize) + } + s.Out = s.Out[:0:s.MaxDecodedSize] + s.Out, err = s.Decoder().Decompress1X(s.Out, in) + return s.Out, err +} + +// Decompress4X will decompress a 4X encoded stream. +// Before this is called, the table must be initialized with ReadTable unless +// the encoder re-used the table. +// The length of the supplied input must match the end of a block exactly. +// The destination size of the uncompressed data must be known and provided. +// deprecated: Use the stateless Decoder() to get a concurrent version. +func (s *Scratch) Decompress4X(in []byte, dstSize int) (out []byte, err error) { + if dstSize > s.MaxDecodedSize { + return nil, ErrMaxDecodedSizeExceeded + } + if cap(s.Out) < dstSize { + s.Out = make([]byte, s.MaxDecodedSize) + } + s.Out = s.Out[:0:dstSize] + s.Out, err = s.Decoder().Decompress4X(s.Out, in) + return s.Out, err +} + +// Decoder will return a stateless decoder that can be used by multiple +// decompressors concurrently. +// Before this is called, the table must be initialized with ReadTable. +// The Decoder is still linked to the scratch buffer so that cannot be reused. +// However, it is safe to discard the scratch. +func (s *Scratch) Decoder() *Decoder { + return &Decoder{ + dt: s.dt, + actualTableLog: s.actualTableLog, + bufs: &s.decPool, + } +} + +// Decoder provides stateless decoding. +type Decoder struct { + dt dTable + actualTableLog uint8 + bufs *sync.Pool +} + +func (d *Decoder) buffer() *[4][256]byte { + buf, ok := d.bufs.Get().(*[4][256]byte) + if ok { + return buf + } + return &[4][256]byte{} +} + +// Decompress1X will decompress a 1X encoded stream. +// The cap of the output buffer will be the maximum decompressed size. +// The length of the supplied input must match the end of a block exactly. +func (d *Decoder) Decompress1X(dst, src []byte) ([]byte, error) { + if len(d.dt.single) == 0 { + return nil, errors.New("no table loaded") + } + if use8BitTables && d.actualTableLog <= 8 { + return d.decompress1X8Bit(dst, src) + } + var br bitReaderShifted + err := br.init(src) + if err != nil { + return dst, err + } + maxDecodedSize := cap(dst) + dst = dst[:0] + + // Avoid bounds check by always having full sized table. + const tlSize = 1 << tableLogMax + const tlMask = tlSize - 1 + dt := d.dt.single[:tlSize] + + // Use temp table to avoid bound checks/append penalty. + bufs := d.buffer() + buf := &bufs[0] + var off uint8 + + for br.off >= 8 { + br.fillFast() + v := dt[br.peekBitsFast(d.actualTableLog)&tlMask] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[br.peekBitsFast(d.actualTableLog)&tlMask] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + // Refill + br.fillFast() + + v = dt[br.peekBitsFast(d.actualTableLog)&tlMask] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[br.peekBitsFast(d.actualTableLog)&tlMask] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + br.close() + d.bufs.Put(bufs) + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + + if len(dst)+int(off) > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:off]...) + + // br < 8, so uint8 is fine + bitsLeft := uint8(br.off)*8 + 64 - br.bitsRead + for bitsLeft > 0 { + br.fill() + if false && br.bitsRead >= 32 { + if br.off >= 4 { + v := br.in[br.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + br.value = (br.value << 32) | uint64(low) + br.bitsRead -= 32 + br.off -= 4 + } else { + for br.off > 0 { + br.value = (br.value << 8) | uint64(br.in[br.off-1]) + br.bitsRead -= 8 + br.off-- + } + } + } + if len(dst) >= maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + v := d.dt.single[br.peekBitsFast(d.actualTableLog)&tlMask] + nBits := uint8(v.entry) + br.advance(nBits) + bitsLeft -= nBits + dst = append(dst, uint8(v.entry>>8)) + } + d.bufs.Put(bufs) + return dst, br.close() +} + +// decompress1X8Bit will decompress a 1X encoded stream with tablelog <= 8. +// The cap of the output buffer will be the maximum decompressed size. +// The length of the supplied input must match the end of a block exactly. +func (d *Decoder) decompress1X8Bit(dst, src []byte) ([]byte, error) { + if d.actualTableLog == 8 { + return d.decompress1X8BitExactly(dst, src) + } + var br bitReaderBytes + err := br.init(src) + if err != nil { + return dst, err + } + maxDecodedSize := cap(dst) + dst = dst[:0] + + // Avoid bounds check by always having full sized table. + dt := d.dt.single[:256] + + // Use temp table to avoid bound checks/append penalty. + bufs := d.buffer() + buf := &bufs[0] + var off uint8 + + switch d.actualTableLog { + case 8: + const shift = 8 - 8 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + br.close() + d.bufs.Put(bufs) + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 7: + const shift = 8 - 7 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + br.close() + d.bufs.Put(bufs) + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 6: + const shift = 8 - 6 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 5: + const shift = 8 - 5 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 4: + const shift = 8 - 4 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 3: + const shift = 8 - 3 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 2: + const shift = 8 - 2 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + case 1: + const shift = 8 - 1 + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>(56+shift))] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + default: + d.bufs.Put(bufs) + return nil, fmt.Errorf("invalid tablelog: %d", d.actualTableLog) + } + + if len(dst)+int(off) > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:off]...) + + // br < 4, so uint8 is fine + bitsLeft := int8(uint8(br.off)*8 + (64 - br.bitsRead)) + shift := (8 - d.actualTableLog) & 7 + + for bitsLeft > 0 { + if br.bitsRead >= 64-8 { + for br.off > 0 { + br.value |= uint64(br.in[br.off-1]) << (br.bitsRead - 8) + br.bitsRead -= 8 + br.off-- + } + } + if len(dst) >= maxDecodedSize { + br.close() + d.bufs.Put(bufs) + return nil, ErrMaxDecodedSizeExceeded + } + v := dt[br.peekByteFast()>>shift] + nBits := uint8(v.entry) + br.advance(nBits) + bitsLeft -= int8(nBits) + dst = append(dst, uint8(v.entry>>8)) + } + d.bufs.Put(bufs) + return dst, br.close() +} + +// decompress1X8Bit will decompress a 1X encoded stream with tablelog <= 8. +// The cap of the output buffer will be the maximum decompressed size. +// The length of the supplied input must match the end of a block exactly. +func (d *Decoder) decompress1X8BitExactly(dst, src []byte) ([]byte, error) { + var br bitReaderBytes + err := br.init(src) + if err != nil { + return dst, err + } + maxDecodedSize := cap(dst) + dst = dst[:0] + + // Avoid bounds check by always having full sized table. + dt := d.dt.single[:256] + + // Use temp table to avoid bound checks/append penalty. + bufs := d.buffer() + buf := &bufs[0] + var off uint8 + + const shift = 56 + + //fmt.Printf("mask: %b, tl:%d\n", mask, d.actualTableLog) + for br.off >= 4 { + br.fillFast() + v := dt[uint8(br.value>>shift)] + br.advance(uint8(v.entry)) + buf[off+0] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>shift)] + br.advance(uint8(v.entry)) + buf[off+1] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>shift)] + br.advance(uint8(v.entry)) + buf[off+2] = uint8(v.entry >> 8) + + v = dt[uint8(br.value>>shift)] + br.advance(uint8(v.entry)) + buf[off+3] = uint8(v.entry >> 8) + + off += 4 + if off == 0 { + if len(dst)+256 > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:]...) + } + } + + if len(dst)+int(off) > maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + dst = append(dst, buf[:off]...) + + // br < 4, so uint8 is fine + bitsLeft := int8(uint8(br.off)*8 + (64 - br.bitsRead)) + for bitsLeft > 0 { + if br.bitsRead >= 64-8 { + for br.off > 0 { + br.value |= uint64(br.in[br.off-1]) << (br.bitsRead - 8) + br.bitsRead -= 8 + br.off-- + } + } + if len(dst) >= maxDecodedSize { + d.bufs.Put(bufs) + br.close() + return nil, ErrMaxDecodedSizeExceeded + } + v := dt[br.peekByteFast()] + nBits := uint8(v.entry) + br.advance(nBits) + bitsLeft -= int8(nBits) + dst = append(dst, uint8(v.entry>>8)) + } + d.bufs.Put(bufs) + return dst, br.close() +} + +// Decompress4X will decompress a 4X encoded stream. +// The length of the supplied input must match the end of a block exactly. +// The *capacity* of the dst slice must match the destination size of +// the uncompressed data exactly. +func (d *Decoder) decompress4X8bit(dst, src []byte) ([]byte, error) { + if d.actualTableLog == 8 { + return d.decompress4X8bitExactly(dst, src) + } + + var br [4]bitReaderBytes + start := 6 + for i := 0; i < 3; i++ { + length := int(src[i*2]) | (int(src[i*2+1]) << 8) + if start+length >= len(src) { + return nil, errors.New("truncated input (or invalid offset)") + } + err := br[i].init(src[start : start+length]) + if err != nil { + return nil, err + } + start += length + } + err := br[3].init(src[start:]) + if err != nil { + return nil, err + } + + // destination, offset to match first output + dstSize := cap(dst) + dst = dst[:dstSize] + out := dst + dstEvery := (dstSize + 3) / 4 + + shift := (56 + (8 - d.actualTableLog)) & 63 + + const tlSize = 1 << 8 + single := d.dt.single[:tlSize] + + // Use temp table to avoid bound checks/append penalty. + buf := d.buffer() + var off uint8 + var decoded int + + // Decode 4 values from each decoder/loop. + const bufoff = 256 + for { + if br[0].off < 4 || br[1].off < 4 || br[2].off < 4 || br[3].off < 4 { + break + } + + { + // Interleave 2 decodes. + const stream = 0 + const stream2 = 1 + br1 := &br[stream] + br2 := &br[stream2] + br1.fillFast() + br2.fillFast() + + v := single[uint8(br1.value>>shift)].entry + v2 := single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off] = uint8(v >> 8) + buf[stream2][off] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+1] = uint8(v >> 8) + buf[stream2][off+1] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+2] = uint8(v >> 8) + buf[stream2][off+2] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+3] = uint8(v >> 8) + buf[stream2][off+3] = uint8(v2 >> 8) + } + + { + const stream = 2 + const stream2 = 3 + br1 := &br[stream] + br2 := &br[stream2] + br1.fillFast() + br2.fillFast() + + v := single[uint8(br1.value>>shift)].entry + v2 := single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off] = uint8(v >> 8) + buf[stream2][off] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+1] = uint8(v >> 8) + buf[stream2][off+1] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+2] = uint8(v >> 8) + buf[stream2][off+2] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+3] = uint8(v >> 8) + buf[stream2][off+3] = uint8(v2 >> 8) + } + + off += 4 + + if off == 0 { + if bufoff > dstEvery { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 1") + } + copy(out, buf[0][:]) + copy(out[dstEvery:], buf[1][:]) + copy(out[dstEvery*2:], buf[2][:]) + copy(out[dstEvery*3:], buf[3][:]) + out = out[bufoff:] + decoded += bufoff * 4 + // There must at least be 3 buffers left. + if len(out) < dstEvery*3 { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 2") + } + } + } + if off > 0 { + ioff := int(off) + if len(out) < dstEvery*3+ioff { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 3") + } + copy(out, buf[0][:off]) + copy(out[dstEvery:], buf[1][:off]) + copy(out[dstEvery*2:], buf[2][:off]) + copy(out[dstEvery*3:], buf[3][:off]) + decoded += int(off) * 4 + out = out[off:] + } + + // Decode remaining. + // Decode remaining. + remainBytes := dstEvery - (decoded / 4) + for i := range br { + offset := dstEvery * i + endsAt := offset + remainBytes + if endsAt > len(out) { + endsAt = len(out) + } + br := &br[i] + bitsLeft := br.remaining() + for bitsLeft > 0 { + if br.finished() { + d.bufs.Put(buf) + return nil, io.ErrUnexpectedEOF + } + if br.bitsRead >= 56 { + if br.off >= 4 { + v := br.in[br.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + br.value |= uint64(low) << (br.bitsRead - 32) + br.bitsRead -= 32 + br.off -= 4 + } else { + for br.off > 0 { + br.value |= uint64(br.in[br.off-1]) << (br.bitsRead - 8) + br.bitsRead -= 8 + br.off-- + } + } + } + // end inline... + if offset >= endsAt { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 4") + } + + // Read value and increment offset. + v := single[uint8(br.value>>shift)].entry + nBits := uint8(v) + br.advance(nBits) + bitsLeft -= uint(nBits) + out[offset] = uint8(v >> 8) + offset++ + } + if offset != endsAt { + d.bufs.Put(buf) + return nil, fmt.Errorf("corruption detected: short output block %d, end %d != %d", i, offset, endsAt) + } + decoded += offset - dstEvery*i + err = br.close() + if err != nil { + d.bufs.Put(buf) + return nil, err + } + } + d.bufs.Put(buf) + if dstSize != decoded { + return nil, errors.New("corruption detected: short output block") + } + return dst, nil +} + +// Decompress4X will decompress a 4X encoded stream. +// The length of the supplied input must match the end of a block exactly. +// The *capacity* of the dst slice must match the destination size of +// the uncompressed data exactly. +func (d *Decoder) decompress4X8bitExactly(dst, src []byte) ([]byte, error) { + var br [4]bitReaderBytes + start := 6 + for i := 0; i < 3; i++ { + length := int(src[i*2]) | (int(src[i*2+1]) << 8) + if start+length >= len(src) { + return nil, errors.New("truncated input (or invalid offset)") + } + err := br[i].init(src[start : start+length]) + if err != nil { + return nil, err + } + start += length + } + err := br[3].init(src[start:]) + if err != nil { + return nil, err + } + + // destination, offset to match first output + dstSize := cap(dst) + dst = dst[:dstSize] + out := dst + dstEvery := (dstSize + 3) / 4 + + const shift = 56 + const tlSize = 1 << 8 + const tlMask = tlSize - 1 + single := d.dt.single[:tlSize] + + // Use temp table to avoid bound checks/append penalty. + buf := d.buffer() + var off uint8 + var decoded int + + // Decode 4 values from each decoder/loop. + const bufoff = 256 + for { + if br[0].off < 4 || br[1].off < 4 || br[2].off < 4 || br[3].off < 4 { + break + } + + { + // Interleave 2 decodes. + const stream = 0 + const stream2 = 1 + br1 := &br[stream] + br2 := &br[stream2] + br1.fillFast() + br2.fillFast() + + v := single[uint8(br1.value>>shift)].entry + v2 := single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off] = uint8(v >> 8) + buf[stream2][off] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+1] = uint8(v >> 8) + buf[stream2][off+1] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+2] = uint8(v >> 8) + buf[stream2][off+2] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+3] = uint8(v >> 8) + buf[stream2][off+3] = uint8(v2 >> 8) + } + + { + const stream = 2 + const stream2 = 3 + br1 := &br[stream] + br2 := &br[stream2] + br1.fillFast() + br2.fillFast() + + v := single[uint8(br1.value>>shift)].entry + v2 := single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off] = uint8(v >> 8) + buf[stream2][off] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+1] = uint8(v >> 8) + buf[stream2][off+1] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+2] = uint8(v >> 8) + buf[stream2][off+2] = uint8(v2 >> 8) + + v = single[uint8(br1.value>>shift)].entry + v2 = single[uint8(br2.value>>shift)].entry + br1.bitsRead += uint8(v) + br1.value <<= v & 63 + br2.bitsRead += uint8(v2) + br2.value <<= v2 & 63 + buf[stream][off+3] = uint8(v >> 8) + buf[stream2][off+3] = uint8(v2 >> 8) + } + + off += 4 + + if off == 0 { + if bufoff > dstEvery { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 1") + } + copy(out, buf[0][:]) + copy(out[dstEvery:], buf[1][:]) + copy(out[dstEvery*2:], buf[2][:]) + copy(out[dstEvery*3:], buf[3][:]) + out = out[bufoff:] + decoded += bufoff * 4 + // There must at least be 3 buffers left. + if len(out) < dstEvery*3 { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 2") + } + } + } + if off > 0 { + ioff := int(off) + if len(out) < dstEvery*3+ioff { + return nil, errors.New("corruption detected: stream overrun 3") + } + copy(out, buf[0][:off]) + copy(out[dstEvery:], buf[1][:off]) + copy(out[dstEvery*2:], buf[2][:off]) + copy(out[dstEvery*3:], buf[3][:off]) + decoded += int(off) * 4 + out = out[off:] + } + + // Decode remaining. + remainBytes := dstEvery - (decoded / 4) + for i := range br { + offset := dstEvery * i + endsAt := offset + remainBytes + if endsAt > len(out) { + endsAt = len(out) + } + br := &br[i] + bitsLeft := br.remaining() + for bitsLeft > 0 { + if br.finished() { + d.bufs.Put(buf) + return nil, io.ErrUnexpectedEOF + } + if br.bitsRead >= 56 { + if br.off >= 4 { + v := br.in[br.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + br.value |= uint64(low) << (br.bitsRead - 32) + br.bitsRead -= 32 + br.off -= 4 + } else { + for br.off > 0 { + br.value |= uint64(br.in[br.off-1]) << (br.bitsRead - 8) + br.bitsRead -= 8 + br.off-- + } + } + } + // end inline... + if offset >= endsAt { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 4") + } + + // Read value and increment offset. + v := single[br.peekByteFast()].entry + nBits := uint8(v) + br.advance(nBits) + bitsLeft -= uint(nBits) + out[offset] = uint8(v >> 8) + offset++ + } + if offset != endsAt { + d.bufs.Put(buf) + return nil, fmt.Errorf("corruption detected: short output block %d, end %d != %d", i, offset, endsAt) + } + + decoded += offset - dstEvery*i + err = br.close() + if err != nil { + d.bufs.Put(buf) + return nil, err + } + } + d.bufs.Put(buf) + if dstSize != decoded { + return nil, errors.New("corruption detected: short output block") + } + return dst, nil +} + +// matches will compare a decoding table to a coding table. +// Errors are written to the writer. +// Nothing will be written if table is ok. +func (s *Scratch) matches(ct cTable, w io.Writer) { + if s == nil || len(s.dt.single) == 0 { + return + } + dt := s.dt.single[:1<>8) == byte(sym) { + fmt.Fprintf(w, "symbol %x has decoder, but no encoder\n", sym) + errs++ + break + } + } + if errs == 0 { + broken-- + } + continue + } + // Unused bits in input + ub := tablelog - enc.nBits + top := enc.val << ub + // decoder looks at top bits. + dec := dt[top] + if uint8(dec.entry) != enc.nBits { + fmt.Fprintf(w, "symbol 0x%x bit size mismatch (enc: %d, dec:%d).\n", sym, enc.nBits, uint8(dec.entry)) + errs++ + } + if uint8(dec.entry>>8) != uint8(sym) { + fmt.Fprintf(w, "symbol 0x%x decoder output mismatch (enc: %d, dec:%d).\n", sym, sym, uint8(dec.entry>>8)) + errs++ + } + if errs > 0 { + fmt.Fprintf(w, "%d errros in base, stopping\n", errs) + continue + } + // Ensure that all combinations are covered. + for i := uint16(0); i < (1 << ub); i++ { + vval := top | i + dec := dt[vval] + if uint8(dec.entry) != enc.nBits { + fmt.Fprintf(w, "symbol 0x%x bit size mismatch (enc: %d, dec:%d).\n", vval, enc.nBits, uint8(dec.entry)) + errs++ + } + if uint8(dec.entry>>8) != uint8(sym) { + fmt.Fprintf(w, "symbol 0x%x decoder output mismatch (enc: %d, dec:%d).\n", vval, sym, uint8(dec.entry>>8)) + errs++ + } + if errs > 20 { + fmt.Fprintf(w, "%d errros, stopping\n", errs) + break + } + } + if errs == 0 { + ok++ + broken-- + } + } + if broken > 0 { + fmt.Fprintf(w, "%d broken, %d ok\n", broken, ok) + } +} diff --git a/vendor/github.com/klauspost/compress/huff0/decompress_8b_amd64.s b/vendor/github.com/klauspost/compress/huff0/decompress_8b_amd64.s new file mode 100644 index 0000000..0d6cb1a --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress_8b_amd64.s @@ -0,0 +1,488 @@ +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" +#include "funcdata.h" +#include "go_asm.h" + +#define bufoff 256 // see decompress.go, we're using [4][256]byte table + +// func decompress4x_main_loop_x86(pbr0, pbr1, pbr2, pbr3 *bitReaderShifted, +// peekBits uint8, buf *byte, tbl *dEntrySingle) (int, bool) +TEXT ·decompress4x_8b_loop_x86(SB), NOSPLIT, $8 +#define off R8 +#define buffer DI +#define table SI + +#define br_bits_read R9 +#define br_value R10 +#define br_offset R11 +#define peek_bits R12 +#define exhausted DX + +#define br0 R13 +#define br1 R14 +#define br2 R15 +#define br3 BP + + MOVQ BP, 0(SP) + + XORQ exhausted, exhausted // exhausted = false + XORQ off, off // off = 0 + + MOVBQZX peekBits+32(FP), peek_bits + MOVQ buf+40(FP), buffer + MOVQ tbl+48(FP), table + + MOVQ pbr0+0(FP), br0 + MOVQ pbr1+8(FP), br1 + MOVQ pbr2+16(FP), br2 + MOVQ pbr3+24(FP), br3 + +main_loop: + + // const stream = 0 + // br0.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br0), br_bits_read + MOVQ bitReaderShifted_value(br0), br_value + MOVQ bitReaderShifted_off(br0), br_offset + + // if b.bitsRead >= 32 { + CMPQ br_bits_read, $32 + JB skip_fill0 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br0), AX + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + + // b.value |= uint64(low) << (b.bitsRead & 63) + MOVQ br_bits_read, CX + SHLQ CL, AX + ORQ AX, br_value + + // exhausted = exhausted || (br0.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill0: + + // val0 := br0.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br0.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val1 := br0.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br0.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 0(buffer)(off*1) + + // SECOND PART: + // val2 := br0.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v2 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br0.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val3 := br0.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v3 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br0.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off+2] = uint8(v2.entry >> 8) + // buf[stream][off+3] = uint8(v3.entry >> 8) + MOVW BX, 0+2(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br0) + MOVQ br_value, bitReaderShifted_value(br0) + MOVQ br_offset, bitReaderShifted_off(br0) + + // const stream = 1 + // br1.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br1), br_bits_read + MOVQ bitReaderShifted_value(br1), br_value + MOVQ bitReaderShifted_off(br1), br_offset + + // if b.bitsRead >= 32 { + CMPQ br_bits_read, $32 + JB skip_fill1 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br1), AX + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + + // b.value |= uint64(low) << (b.bitsRead & 63) + MOVQ br_bits_read, CX + SHLQ CL, AX + ORQ AX, br_value + + // exhausted = exhausted || (br1.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill1: + + // val0 := br1.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br1.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val1 := br1.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br1.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 256(buffer)(off*1) + + // SECOND PART: + // val2 := br1.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v2 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br1.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val3 := br1.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v3 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br1.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off+2] = uint8(v2.entry >> 8) + // buf[stream][off+3] = uint8(v3.entry >> 8) + MOVW BX, 256+2(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br1) + MOVQ br_value, bitReaderShifted_value(br1) + MOVQ br_offset, bitReaderShifted_off(br1) + + // const stream = 2 + // br2.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br2), br_bits_read + MOVQ bitReaderShifted_value(br2), br_value + MOVQ bitReaderShifted_off(br2), br_offset + + // if b.bitsRead >= 32 { + CMPQ br_bits_read, $32 + JB skip_fill2 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br2), AX + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + + // b.value |= uint64(low) << (b.bitsRead & 63) + MOVQ br_bits_read, CX + SHLQ CL, AX + ORQ AX, br_value + + // exhausted = exhausted || (br2.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill2: + + // val0 := br2.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br2.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val1 := br2.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br2.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 512(buffer)(off*1) + + // SECOND PART: + // val2 := br2.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v2 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br2.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val3 := br2.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v3 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br2.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off+2] = uint8(v2.entry >> 8) + // buf[stream][off+3] = uint8(v3.entry >> 8) + MOVW BX, 512+2(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br2) + MOVQ br_value, bitReaderShifted_value(br2) + MOVQ br_offset, bitReaderShifted_off(br2) + + // const stream = 3 + // br3.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br3), br_bits_read + MOVQ bitReaderShifted_value(br3), br_value + MOVQ bitReaderShifted_off(br3), br_offset + + // if b.bitsRead >= 32 { + CMPQ br_bits_read, $32 + JB skip_fill3 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br3), AX + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + + // b.value |= uint64(low) << (b.bitsRead & 63) + MOVQ br_bits_read, CX + SHLQ CL, AX + ORQ AX, br_value + + // exhausted = exhausted || (br3.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill3: + + // val0 := br3.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br3.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val1 := br3.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br3.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 768(buffer)(off*1) + + // SECOND PART: + // val2 := br3.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v2 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br3.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val3 := br3.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v3 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br3.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off+2] = uint8(v2.entry >> 8) + // buf[stream][off+3] = uint8(v3.entry >> 8) + MOVW BX, 768+2(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br3) + MOVQ br_value, bitReaderShifted_value(br3) + MOVQ br_offset, bitReaderShifted_off(br3) + + ADDQ $4, off // off += 2 + + TESTB DH, DH // any br[i].ofs < 4? + JNZ end + + CMPQ off, $bufoff + JL main_loop + +end: + MOVQ 0(SP), BP + + MOVB off, ret+56(FP) + RET + +#undef off +#undef buffer +#undef table + +#undef br_bits_read +#undef br_value +#undef br_offset +#undef peek_bits +#undef exhausted + +#undef br0 +#undef br1 +#undef br2 +#undef br3 diff --git a/vendor/github.com/klauspost/compress/huff0/decompress_8b_amd64.s.in b/vendor/github.com/klauspost/compress/huff0/decompress_8b_amd64.s.in new file mode 100644 index 0000000..6d477a2 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress_8b_amd64.s.in @@ -0,0 +1,197 @@ +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" +#include "funcdata.h" +#include "go_asm.h" + + +#define bufoff 256 // see decompress.go, we're using [4][256]byte table + +//func decompress4x_main_loop_x86(pbr0, pbr1, pbr2, pbr3 *bitReaderShifted, +// peekBits uint8, buf *byte, tbl *dEntrySingle) (int, bool) +TEXT ·decompress4x_8b_loop_x86(SB), NOSPLIT, $8 +#define off R8 +#define buffer DI +#define table SI + +#define br_bits_read R9 +#define br_value R10 +#define br_offset R11 +#define peek_bits R12 +#define exhausted DX + +#define br0 R13 +#define br1 R14 +#define br2 R15 +#define br3 BP + + MOVQ BP, 0(SP) + + XORQ exhausted, exhausted // exhausted = false + XORQ off, off // off = 0 + + MOVBQZX peekBits+32(FP), peek_bits + MOVQ buf+40(FP), buffer + MOVQ tbl+48(FP), table + + MOVQ pbr0+0(FP), br0 + MOVQ pbr1+8(FP), br1 + MOVQ pbr2+16(FP), br2 + MOVQ pbr3+24(FP), br3 + +main_loop: +{{ define "decode_2_values_x86" }} + // const stream = {{ var "id" }} + // br{{ var "id"}}.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br{{ var "id" }}), br_bits_read + MOVQ bitReaderShifted_value(br{{ var "id" }}), br_value + MOVQ bitReaderShifted_off(br{{ var "id" }}), br_offset + + // if b.bitsRead >= 32 { + CMPQ br_bits_read, $32 + JB skip_fill{{ var "id" }} + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br{{ var "id" }}), AX + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + + // b.value |= uint64(low) << (b.bitsRead & 63) + MOVQ br_bits_read, CX + SHLQ CL, AX + ORQ AX, br_value + + // exhausted = exhausted || (br{{ var "id"}}.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + // } +skip_fill{{ var "id" }}: + + // val0 := br{{ var "id"}}.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br{{ var "id"}}.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val1 := br{{ var "id"}}.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br{{ var "id"}}.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, {{ var "bufofs" }}(buffer)(off*1) + + // SECOND PART: + // val2 := br{{ var "id"}}.peekTopBits(peekBits) + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v2 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br{{ var "id"}}.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + // val3 := br{{ var "id"}}.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + + // v3 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br{{ var "id"}}.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + MOVBQZX AL, CX + SHLQ CX, br_value // value <<= n + ADDQ CX, br_bits_read // bits_read += n + + + // these two writes get coalesced + // buf[stream][off+2] = uint8(v2.entry >> 8) + // buf[stream][off+3] = uint8(v3.entry >> 8) + MOVW BX, {{ var "bufofs" }}+2(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br{{ var "id" }}) + MOVQ br_value, bitReaderShifted_value(br{{ var "id" }}) + MOVQ br_offset, bitReaderShifted_off(br{{ var "id" }}) +{{ end }} + + {{ set "id" "0" }} + {{ set "ofs" "0" }} + {{ set "bufofs" "0" }} {{/* id * bufoff */}} + {{ template "decode_2_values_x86" . }} + + {{ set "id" "1" }} + {{ set "ofs" "8" }} + {{ set "bufofs" "256" }} + {{ template "decode_2_values_x86" . }} + + {{ set "id" "2" }} + {{ set "ofs" "16" }} + {{ set "bufofs" "512" }} + {{ template "decode_2_values_x86" . }} + + {{ set "id" "3" }} + {{ set "ofs" "24" }} + {{ set "bufofs" "768" }} + {{ template "decode_2_values_x86" . }} + + ADDQ $4, off // off += 2 + + TESTB DH, DH // any br[i].ofs < 4? + JNZ end + + CMPQ off, $bufoff + JL main_loop +end: + MOVQ 0(SP), BP + + MOVB off, ret+56(FP) + RET +#undef off +#undef buffer +#undef table + +#undef br_bits_read +#undef br_value +#undef br_offset +#undef peek_bits +#undef exhausted + +#undef br0 +#undef br1 +#undef br2 +#undef br3 diff --git a/vendor/github.com/klauspost/compress/huff0/decompress_amd64.go b/vendor/github.com/klauspost/compress/huff0/decompress_amd64.go new file mode 100644 index 0000000..d47f664 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress_amd64.go @@ -0,0 +1,181 @@ +//go:build amd64 && !appengine && !noasm && gc +// +build amd64,!appengine,!noasm,gc + +// This file contains the specialisation of Decoder.Decompress4X +// that uses an asm implementation of its main loop. +package huff0 + +import ( + "errors" + "fmt" +) + +// decompress4x_main_loop_x86 is an x86 assembler implementation +// of Decompress4X when tablelog > 8. +// go:noescape +func decompress4x_main_loop_x86(pbr0, pbr1, pbr2, pbr3 *bitReaderShifted, + peekBits uint8, buf *byte, tbl *dEntrySingle) uint8 + +// decompress4x_8b_loop_x86 is an x86 assembler implementation +// of Decompress4X when tablelog <= 8 which decodes 4 entries +// per loop. +// go:noescape +func decompress4x_8b_loop_x86(pbr0, pbr1, pbr2, pbr3 *bitReaderShifted, + peekBits uint8, buf *byte, tbl *dEntrySingle) uint8 + +// fallback8BitSize is the size where using Go version is faster. +const fallback8BitSize = 800 + +// Decompress4X will decompress a 4X encoded stream. +// The length of the supplied input must match the end of a block exactly. +// The *capacity* of the dst slice must match the destination size of +// the uncompressed data exactly. +func (d *Decoder) Decompress4X(dst, src []byte) ([]byte, error) { + if len(d.dt.single) == 0 { + return nil, errors.New("no table loaded") + } + if len(src) < 6+(4*1) { + return nil, errors.New("input too small") + } + + use8BitTables := d.actualTableLog <= 8 + if cap(dst) < fallback8BitSize && use8BitTables { + return d.decompress4X8bit(dst, src) + } + var br [4]bitReaderShifted + // Decode "jump table" + start := 6 + for i := 0; i < 3; i++ { + length := int(src[i*2]) | (int(src[i*2+1]) << 8) + if start+length >= len(src) { + return nil, errors.New("truncated input (or invalid offset)") + } + err := br[i].init(src[start : start+length]) + if err != nil { + return nil, err + } + start += length + } + err := br[3].init(src[start:]) + if err != nil { + return nil, err + } + + // destination, offset to match first output + dstSize := cap(dst) + dst = dst[:dstSize] + out := dst + dstEvery := (dstSize + 3) / 4 + + const tlSize = 1 << tableLogMax + const tlMask = tlSize - 1 + single := d.dt.single[:tlSize] + + // Use temp table to avoid bound checks/append penalty. + buf := d.buffer() + var off uint8 + var decoded int + + const debug = false + + // see: bitReaderShifted.peekBitsFast() + peekBits := uint8((64 - d.actualTableLog) & 63) + + // Decode 2 values from each decoder/loop. + const bufoff = 256 + for { + if br[0].off < 4 || br[1].off < 4 || br[2].off < 4 || br[3].off < 4 { + break + } + + if use8BitTables { + off = decompress4x_8b_loop_x86(&br[0], &br[1], &br[2], &br[3], peekBits, &buf[0][0], &single[0]) + } else { + off = decompress4x_main_loop_x86(&br[0], &br[1], &br[2], &br[3], peekBits, &buf[0][0], &single[0]) + } + if debug { + fmt.Print("DEBUG: ") + fmt.Printf("off=%d,", off) + for i := 0; i < 4; i++ { + fmt.Printf(" br[%d]={bitsRead=%d, value=%x, off=%d}", + i, br[i].bitsRead, br[i].value, br[i].off) + } + fmt.Println("") + } + + if off != 0 { + break + } + + if bufoff > dstEvery { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 1") + } + copy(out, buf[0][:]) + copy(out[dstEvery:], buf[1][:]) + copy(out[dstEvery*2:], buf[2][:]) + copy(out[dstEvery*3:], buf[3][:]) + out = out[bufoff:] + decoded += bufoff * 4 + // There must at least be 3 buffers left. + if len(out) < dstEvery*3 { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 2") + } + } + if off > 0 { + ioff := int(off) + if len(out) < dstEvery*3+ioff { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 3") + } + copy(out, buf[0][:off]) + copy(out[dstEvery:], buf[1][:off]) + copy(out[dstEvery*2:], buf[2][:off]) + copy(out[dstEvery*3:], buf[3][:off]) + decoded += int(off) * 4 + out = out[off:] + } + + // Decode remaining. + remainBytes := dstEvery - (decoded / 4) + for i := range br { + offset := dstEvery * i + endsAt := offset + remainBytes + if endsAt > len(out) { + endsAt = len(out) + } + br := &br[i] + bitsLeft := br.remaining() + for bitsLeft > 0 { + br.fill() + if offset >= endsAt { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 4") + } + + // Read value and increment offset. + val := br.peekBitsFast(d.actualTableLog) + v := single[val&tlMask].entry + nBits := uint8(v) + br.advance(nBits) + bitsLeft -= uint(nBits) + out[offset] = uint8(v >> 8) + offset++ + } + if offset != endsAt { + d.bufs.Put(buf) + return nil, fmt.Errorf("corruption detected: short output block %d, end %d != %d", i, offset, endsAt) + } + decoded += offset - dstEvery*i + err = br.close() + if err != nil { + return nil, err + } + } + d.bufs.Put(buf) + if dstSize != decoded { + return nil, errors.New("corruption detected: short output block") + } + return dst, nil +} diff --git a/vendor/github.com/klauspost/compress/huff0/decompress_amd64.s b/vendor/github.com/klauspost/compress/huff0/decompress_amd64.s new file mode 100644 index 0000000..2edad3e --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress_amd64.s @@ -0,0 +1,506 @@ +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" +#include "funcdata.h" +#include "go_asm.h" + +#ifdef GOAMD64_v4 +#ifndef GOAMD64_v3 +#define GOAMD64_v3 +#endif +#endif + +#define bufoff 256 // see decompress.go, we're using [4][256]byte table + +// func decompress4x_main_loop_x86(pbr0, pbr1, pbr2, pbr3 *bitReaderShifted, +// peekBits uint8, buf *byte, tbl *dEntrySingle) (int, bool) +TEXT ·decompress4x_main_loop_x86(SB), NOSPLIT, $8 +#define off R8 +#define buffer DI +#define table SI + +#define br_bits_read R9 +#define br_value R10 +#define br_offset R11 +#define peek_bits R12 +#define exhausted DX + +#define br0 R13 +#define br1 R14 +#define br2 R15 +#define br3 BP + + MOVQ BP, 0(SP) + + XORQ exhausted, exhausted // exhausted = false + XORQ off, off // off = 0 + + MOVBQZX peekBits+32(FP), peek_bits + MOVQ buf+40(FP), buffer + MOVQ tbl+48(FP), table + + MOVQ pbr0+0(FP), br0 + MOVQ pbr1+8(FP), br1 + MOVQ pbr2+16(FP), br2 + MOVQ pbr3+24(FP), br3 + +main_loop: + + // const stream = 0 + // br0.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br0), br_bits_read + MOVQ bitReaderShifted_value(br0), br_value + MOVQ bitReaderShifted_off(br0), br_offset + + // We must have at least 2 * max tablelog left + CMPQ br_bits_read, $64-22 + JBE skip_fill0 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br0), AX + + // b.value |= uint64(low) << (b.bitsRead & 63) +#ifdef GOAMD64_v3 + SHLXQ br_bits_read, 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) << (b.bitsRead & 63) + +#else + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + MOVQ br_bits_read, CX + SHLQ CL, AX + +#endif + + ORQ AX, br_value + + // exhausted = exhausted || (br0.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill0: + + // val0 := br0.peekTopBits(peekBits) +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br0.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + // val1 := br0.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br0.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 0(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br0) + MOVQ br_value, bitReaderShifted_value(br0) + MOVQ br_offset, bitReaderShifted_off(br0) + + // const stream = 1 + // br1.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br1), br_bits_read + MOVQ bitReaderShifted_value(br1), br_value + MOVQ bitReaderShifted_off(br1), br_offset + + // We must have at least 2 * max tablelog left + CMPQ br_bits_read, $64-22 + JBE skip_fill1 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br1), AX + + // b.value |= uint64(low) << (b.bitsRead & 63) +#ifdef GOAMD64_v3 + SHLXQ br_bits_read, 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) << (b.bitsRead & 63) + +#else + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + MOVQ br_bits_read, CX + SHLQ CL, AX + +#endif + + ORQ AX, br_value + + // exhausted = exhausted || (br1.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill1: + + // val0 := br1.peekTopBits(peekBits) +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br1.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + // val1 := br1.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br1.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 256(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br1) + MOVQ br_value, bitReaderShifted_value(br1) + MOVQ br_offset, bitReaderShifted_off(br1) + + // const stream = 2 + // br2.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br2), br_bits_read + MOVQ bitReaderShifted_value(br2), br_value + MOVQ bitReaderShifted_off(br2), br_offset + + // We must have at least 2 * max tablelog left + CMPQ br_bits_read, $64-22 + JBE skip_fill2 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br2), AX + + // b.value |= uint64(low) << (b.bitsRead & 63) +#ifdef GOAMD64_v3 + SHLXQ br_bits_read, 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) << (b.bitsRead & 63) + +#else + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + MOVQ br_bits_read, CX + SHLQ CL, AX + +#endif + + ORQ AX, br_value + + // exhausted = exhausted || (br2.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill2: + + // val0 := br2.peekTopBits(peekBits) +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br2.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + // val1 := br2.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br2.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 512(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br2) + MOVQ br_value, bitReaderShifted_value(br2) + MOVQ br_offset, bitReaderShifted_off(br2) + + // const stream = 3 + // br3.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br3), br_bits_read + MOVQ bitReaderShifted_value(br3), br_value + MOVQ bitReaderShifted_off(br3), br_offset + + // We must have at least 2 * max tablelog left + CMPQ br_bits_read, $64-22 + JBE skip_fill3 + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br3), AX + + // b.value |= uint64(low) << (b.bitsRead & 63) +#ifdef GOAMD64_v3 + SHLXQ br_bits_read, 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) << (b.bitsRead & 63) + +#else + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + MOVQ br_bits_read, CX + SHLQ CL, AX + +#endif + + ORQ AX, br_value + + // exhausted = exhausted || (br3.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + + // } +skip_fill3: + + // val0 := br3.peekTopBits(peekBits) +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br3.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask + +#else + // val1 := br3.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask + +#endif + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br3.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n + +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n + +#endif + + ADDQ CX, br_bits_read // bits_read += n + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, 768(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br3) + MOVQ br_value, bitReaderShifted_value(br3) + MOVQ br_offset, bitReaderShifted_off(br3) + + ADDQ $2, off // off += 2 + + TESTB DH, DH // any br[i].ofs < 4? + JNZ end + + CMPQ off, $bufoff + JL main_loop + +end: + MOVQ 0(SP), BP + + MOVB off, ret+56(FP) + RET + +#undef off +#undef buffer +#undef table + +#undef br_bits_read +#undef br_value +#undef br_offset +#undef peek_bits +#undef exhausted + +#undef br0 +#undef br1 +#undef br2 +#undef br3 diff --git a/vendor/github.com/klauspost/compress/huff0/decompress_amd64.s.in b/vendor/github.com/klauspost/compress/huff0/decompress_amd64.s.in new file mode 100644 index 0000000..330d86a --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress_amd64.s.in @@ -0,0 +1,195 @@ +// +build !appengine +// +build gc +// +build !noasm + +#include "textflag.h" +#include "funcdata.h" +#include "go_asm.h" + +#ifdef GOAMD64_v4 +#ifndef GOAMD64_v3 +#define GOAMD64_v3 +#endif +#endif + +#define bufoff 256 // see decompress.go, we're using [4][256]byte table + +//func decompress4x_main_loop_x86(pbr0, pbr1, pbr2, pbr3 *bitReaderShifted, +// peekBits uint8, buf *byte, tbl *dEntrySingle) (int, bool) +TEXT ·decompress4x_main_loop_x86(SB), NOSPLIT, $8 +#define off R8 +#define buffer DI +#define table SI + +#define br_bits_read R9 +#define br_value R10 +#define br_offset R11 +#define peek_bits R12 +#define exhausted DX + +#define br0 R13 +#define br1 R14 +#define br2 R15 +#define br3 BP + + MOVQ BP, 0(SP) + + XORQ exhausted, exhausted // exhausted = false + XORQ off, off // off = 0 + + MOVBQZX peekBits+32(FP), peek_bits + MOVQ buf+40(FP), buffer + MOVQ tbl+48(FP), table + + MOVQ pbr0+0(FP), br0 + MOVQ pbr1+8(FP), br1 + MOVQ pbr2+16(FP), br2 + MOVQ pbr3+24(FP), br3 + +main_loop: +{{ define "decode_2_values_x86" }} + // const stream = {{ var "id" }} + // br{{ var "id"}}.fillFast() + MOVBQZX bitReaderShifted_bitsRead(br{{ var "id" }}), br_bits_read + MOVQ bitReaderShifted_value(br{{ var "id" }}), br_value + MOVQ bitReaderShifted_off(br{{ var "id" }}), br_offset + + // We must have at least 2 * max tablelog left + CMPQ br_bits_read, $64-22 + JBE skip_fill{{ var "id" }} + + SUBQ $32, br_bits_read // b.bitsRead -= 32 + SUBQ $4, br_offset // b.off -= 4 + + // v := b.in[b.off-4 : b.off] + // v = v[:4] + // low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + MOVQ bitReaderShifted_in(br{{ var "id" }}), AX + + // b.value |= uint64(low) << (b.bitsRead & 63) +#ifdef GOAMD64_v3 + SHLXQ br_bits_read, 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) << (b.bitsRead & 63) +#else + MOVL 0(br_offset)(AX*1), AX // AX = uint32(b.in[b.off:b.off+4]) + MOVQ br_bits_read, CX + SHLQ CL, AX +#endif + + ORQ AX, br_value + + // exhausted = exhausted || (br{{ var "id"}}.off < 4) + CMPQ br_offset, $4 + SETLT DL + ORB DL, DH + // } +skip_fill{{ var "id" }}: + + // val0 := br{{ var "id"}}.peekTopBits(peekBits) +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask +#else + MOVQ br_value, AX + MOVQ peek_bits, CX + SHRQ CL, AX // AX = (value >> peek_bits) & mask +#endif + + // v0 := table[val0&mask] + MOVW 0(table)(AX*2), AX // AX - v0 + + // br{{ var "id"}}.advance(uint8(v0.entry)) + MOVB AH, BL // BL = uint8(v0.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n +#endif + + ADDQ CX, br_bits_read // bits_read += n + + +#ifdef GOAMD64_v3 + SHRXQ peek_bits, br_value, AX // AX = (value >> peek_bits) & mask +#else + // val1 := br{{ var "id"}}.peekTopBits(peekBits) + MOVQ peek_bits, CX + MOVQ br_value, AX + SHRQ CL, AX // AX = (value >> peek_bits) & mask +#endif + + // v1 := table[val1&mask] + MOVW 0(table)(AX*2), AX // AX - v1 + + // br{{ var "id"}}.advance(uint8(v1.entry)) + MOVB AH, BH // BH = uint8(v1.entry >> 8) + +#ifdef GOAMD64_v3 + MOVBQZX AL, CX + SHLXQ AX, br_value, br_value // value <<= n +#else + MOVBQZX AL, CX + SHLQ CL, br_value // value <<= n +#endif + + ADDQ CX, br_bits_read // bits_read += n + + + // these two writes get coalesced + // buf[stream][off] = uint8(v0.entry >> 8) + // buf[stream][off+1] = uint8(v1.entry >> 8) + MOVW BX, {{ var "bufofs" }}(buffer)(off*1) + + // update the bitrader reader structure + MOVB br_bits_read, bitReaderShifted_bitsRead(br{{ var "id" }}) + MOVQ br_value, bitReaderShifted_value(br{{ var "id" }}) + MOVQ br_offset, bitReaderShifted_off(br{{ var "id" }}) +{{ end }} + + {{ set "id" "0" }} + {{ set "ofs" "0" }} + {{ set "bufofs" "0" }} {{/* id * bufoff */}} + {{ template "decode_2_values_x86" . }} + + {{ set "id" "1" }} + {{ set "ofs" "8" }} + {{ set "bufofs" "256" }} + {{ template "decode_2_values_x86" . }} + + {{ set "id" "2" }} + {{ set "ofs" "16" }} + {{ set "bufofs" "512" }} + {{ template "decode_2_values_x86" . }} + + {{ set "id" "3" }} + {{ set "ofs" "24" }} + {{ set "bufofs" "768" }} + {{ template "decode_2_values_x86" . }} + + ADDQ $2, off // off += 2 + + TESTB DH, DH // any br[i].ofs < 4? + JNZ end + + CMPQ off, $bufoff + JL main_loop +end: + MOVQ 0(SP), BP + + MOVB off, ret+56(FP) + RET +#undef off +#undef buffer +#undef table + +#undef br_bits_read +#undef br_value +#undef br_offset +#undef peek_bits +#undef exhausted + +#undef br0 +#undef br1 +#undef br2 +#undef br3 diff --git a/vendor/github.com/klauspost/compress/huff0/decompress_generic.go b/vendor/github.com/klauspost/compress/huff0/decompress_generic.go new file mode 100644 index 0000000..126b4d6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/decompress_generic.go @@ -0,0 +1,193 @@ +//go:build !amd64 || appengine || !gc || noasm +// +build !amd64 appengine !gc noasm + +// This file contains a generic implementation of Decoder.Decompress4X. +package huff0 + +import ( + "errors" + "fmt" +) + +// Decompress4X will decompress a 4X encoded stream. +// The length of the supplied input must match the end of a block exactly. +// The *capacity* of the dst slice must match the destination size of +// the uncompressed data exactly. +func (d *Decoder) Decompress4X(dst, src []byte) ([]byte, error) { + if len(d.dt.single) == 0 { + return nil, errors.New("no table loaded") + } + if len(src) < 6+(4*1) { + return nil, errors.New("input too small") + } + if use8BitTables && d.actualTableLog <= 8 { + return d.decompress4X8bit(dst, src) + } + + var br [4]bitReaderShifted + // Decode "jump table" + start := 6 + for i := 0; i < 3; i++ { + length := int(src[i*2]) | (int(src[i*2+1]) << 8) + if start+length >= len(src) { + return nil, errors.New("truncated input (or invalid offset)") + } + err := br[i].init(src[start : start+length]) + if err != nil { + return nil, err + } + start += length + } + err := br[3].init(src[start:]) + if err != nil { + return nil, err + } + + // destination, offset to match first output + dstSize := cap(dst) + dst = dst[:dstSize] + out := dst + dstEvery := (dstSize + 3) / 4 + + const tlSize = 1 << tableLogMax + const tlMask = tlSize - 1 + single := d.dt.single[:tlSize] + + // Use temp table to avoid bound checks/append penalty. + buf := d.buffer() + var off uint8 + var decoded int + + // Decode 2 values from each decoder/loop. + const bufoff = 256 + for { + if br[0].off < 4 || br[1].off < 4 || br[2].off < 4 || br[3].off < 4 { + break + } + + { + const stream = 0 + const stream2 = 1 + br[stream].fillFast() + br[stream2].fillFast() + + val := br[stream].peekBitsFast(d.actualTableLog) + val2 := br[stream2].peekBitsFast(d.actualTableLog) + v := single[val&tlMask] + v2 := single[val2&tlMask] + br[stream].advance(uint8(v.entry)) + br[stream2].advance(uint8(v2.entry)) + buf[stream][off] = uint8(v.entry >> 8) + buf[stream2][off] = uint8(v2.entry >> 8) + + val = br[stream].peekBitsFast(d.actualTableLog) + val2 = br[stream2].peekBitsFast(d.actualTableLog) + v = single[val&tlMask] + v2 = single[val2&tlMask] + br[stream].advance(uint8(v.entry)) + br[stream2].advance(uint8(v2.entry)) + buf[stream][off+1] = uint8(v.entry >> 8) + buf[stream2][off+1] = uint8(v2.entry >> 8) + } + + { + const stream = 2 + const stream2 = 3 + br[stream].fillFast() + br[stream2].fillFast() + + val := br[stream].peekBitsFast(d.actualTableLog) + val2 := br[stream2].peekBitsFast(d.actualTableLog) + v := single[val&tlMask] + v2 := single[val2&tlMask] + br[stream].advance(uint8(v.entry)) + br[stream2].advance(uint8(v2.entry)) + buf[stream][off] = uint8(v.entry >> 8) + buf[stream2][off] = uint8(v2.entry >> 8) + + val = br[stream].peekBitsFast(d.actualTableLog) + val2 = br[stream2].peekBitsFast(d.actualTableLog) + v = single[val&tlMask] + v2 = single[val2&tlMask] + br[stream].advance(uint8(v.entry)) + br[stream2].advance(uint8(v2.entry)) + buf[stream][off+1] = uint8(v.entry >> 8) + buf[stream2][off+1] = uint8(v2.entry >> 8) + } + + off += 2 + + if off == 0 { + if bufoff > dstEvery { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 1") + } + copy(out, buf[0][:]) + copy(out[dstEvery:], buf[1][:]) + copy(out[dstEvery*2:], buf[2][:]) + copy(out[dstEvery*3:], buf[3][:]) + out = out[bufoff:] + decoded += bufoff * 4 + // There must at least be 3 buffers left. + if len(out) < dstEvery*3 { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 2") + } + } + } + if off > 0 { + ioff := int(off) + if len(out) < dstEvery*3+ioff { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 3") + } + copy(out, buf[0][:off]) + copy(out[dstEvery:], buf[1][:off]) + copy(out[dstEvery*2:], buf[2][:off]) + copy(out[dstEvery*3:], buf[3][:off]) + decoded += int(off) * 4 + out = out[off:] + } + + // Decode remaining. + remainBytes := dstEvery - (decoded / 4) + for i := range br { + offset := dstEvery * i + endsAt := offset + remainBytes + if endsAt > len(out) { + endsAt = len(out) + } + br := &br[i] + bitsLeft := br.remaining() + for bitsLeft > 0 { + br.fill() + if offset >= endsAt { + d.bufs.Put(buf) + return nil, errors.New("corruption detected: stream overrun 4") + } + + // Read value and increment offset. + val := br.peekBitsFast(d.actualTableLog) + v := single[val&tlMask].entry + nBits := uint8(v) + br.advance(nBits) + bitsLeft -= uint(nBits) + out[offset] = uint8(v >> 8) + offset++ + } + if offset != endsAt { + d.bufs.Put(buf) + return nil, fmt.Errorf("corruption detected: short output block %d, end %d != %d", i, offset, endsAt) + } + decoded += offset - dstEvery*i + err = br.close() + if err != nil { + return nil, err + } + } + d.bufs.Put(buf) + if dstSize != decoded { + return nil, errors.New("corruption detected: short output block") + } + return dst, nil +} diff --git a/vendor/github.com/klauspost/compress/huff0/huff0.go b/vendor/github.com/klauspost/compress/huff0/huff0.go new file mode 100644 index 0000000..e8ad17a --- /dev/null +++ b/vendor/github.com/klauspost/compress/huff0/huff0.go @@ -0,0 +1,337 @@ +// Package huff0 provides fast huffman encoding as used in zstd. +// +// See README.md at https://github.com/klauspost/compress/tree/master/huff0 for details. +package huff0 + +import ( + "errors" + "fmt" + "math" + "math/bits" + "sync" + + "github.com/klauspost/compress/fse" +) + +const ( + maxSymbolValue = 255 + + // zstandard limits tablelog to 11, see: + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#huffman-tree-description + tableLogMax = 11 + tableLogDefault = 11 + minTablelog = 5 + huffNodesLen = 512 + + // BlockSizeMax is maximum input size for a single block uncompressed. + BlockSizeMax = 1<<18 - 1 +) + +var ( + // ErrIncompressible is returned when input is judged to be too hard to compress. + ErrIncompressible = errors.New("input is not compressible") + + // ErrUseRLE is returned from the compressor when the input is a single byte value repeated. + ErrUseRLE = errors.New("input is single value repeated") + + // ErrTooBig is return if input is too large for a single block. + ErrTooBig = errors.New("input too big") + + // ErrMaxDecodedSizeExceeded is return if input is too large for a single block. + ErrMaxDecodedSizeExceeded = errors.New("maximum output size exceeded") +) + +type ReusePolicy uint8 + +const ( + // ReusePolicyAllow will allow reuse if it produces smaller output. + ReusePolicyAllow ReusePolicy = iota + + // ReusePolicyPrefer will re-use aggressively if possible. + // This will not check if a new table will produce smaller output, + // except if the current table is impossible to use or + // compressed output is bigger than input. + ReusePolicyPrefer + + // ReusePolicyNone will disable re-use of tables. + // This is slightly faster than ReusePolicyAllow but may produce larger output. + ReusePolicyNone + + // ReusePolicyMust must allow reuse and produce smaller output. + ReusePolicyMust +) + +type Scratch struct { + count [maxSymbolValue + 1]uint32 + + // Per block parameters. + // These can be used to override compression parameters of the block. + // Do not touch, unless you know what you are doing. + + // Out is output buffer. + // If the scratch is re-used before the caller is done processing the output, + // set this field to nil. + // Otherwise the output buffer will be re-used for next Compression/Decompression step + // and allocation will be avoided. + Out []byte + + // OutTable will contain the table data only, if a new table has been generated. + // Slice of the returned data. + OutTable []byte + + // OutData will contain the compressed data. + // Slice of the returned data. + OutData []byte + + // MaxDecodedSize will set the maximum allowed output size. + // This value will automatically be set to BlockSizeMax if not set. + // Decoders will return ErrMaxDecodedSizeExceeded is this limit is exceeded. + MaxDecodedSize int + + br byteReader + + // MaxSymbolValue will override the maximum symbol value of the next block. + MaxSymbolValue uint8 + + // TableLog will attempt to override the tablelog for the next block. + // Must be <= 11 and >= 5. + TableLog uint8 + + // Reuse will specify the reuse policy + Reuse ReusePolicy + + // WantLogLess allows to specify a log 2 reduction that should at least be achieved, + // otherwise the block will be returned as incompressible. + // The reduction should then at least be (input size >> WantLogLess) + // If WantLogLess == 0 any improvement will do. + WantLogLess uint8 + + symbolLen uint16 // Length of active part of the symbol table. + maxCount int // count of the most probable symbol + clearCount bool // clear count + actualTableLog uint8 // Selected tablelog. + prevTableLog uint8 // Tablelog for previous table + prevTable cTable // Table used for previous compression. + cTable cTable // compression table + dt dTable // decompression table + nodes []nodeElt + tmpOut [4][]byte + fse *fse.Scratch + decPool sync.Pool // *[4][256]byte buffers. + huffWeight [maxSymbolValue + 1]byte +} + +// TransferCTable will transfer the previously used compression table. +func (s *Scratch) TransferCTable(src *Scratch) { + if cap(s.prevTable) < len(src.prevTable) { + s.prevTable = make(cTable, 0, maxSymbolValue+1) + } + s.prevTable = s.prevTable[:len(src.prevTable)] + copy(s.prevTable, src.prevTable) + s.prevTableLog = src.prevTableLog +} + +func (s *Scratch) prepare(in []byte) (*Scratch, error) { + if len(in) > BlockSizeMax { + return nil, ErrTooBig + } + if s == nil { + s = &Scratch{} + } + if s.MaxSymbolValue == 0 { + s.MaxSymbolValue = maxSymbolValue + } + if s.TableLog == 0 { + s.TableLog = tableLogDefault + } + if s.TableLog > tableLogMax || s.TableLog < minTablelog { + return nil, fmt.Errorf(" invalid tableLog %d (%d -> %d)", s.TableLog, minTablelog, tableLogMax) + } + if s.MaxDecodedSize <= 0 || s.MaxDecodedSize > BlockSizeMax { + s.MaxDecodedSize = BlockSizeMax + } + if s.clearCount && s.maxCount == 0 { + for i := range s.count { + s.count[i] = 0 + } + s.clearCount = false + } + if cap(s.Out) == 0 { + s.Out = make([]byte, 0, len(in)) + } + s.Out = s.Out[:0] + + s.OutTable = nil + s.OutData = nil + if cap(s.nodes) < huffNodesLen+1 { + s.nodes = make([]nodeElt, 0, huffNodesLen+1) + } + s.nodes = s.nodes[:0] + if s.fse == nil { + s.fse = &fse.Scratch{} + } + s.br.init(in) + + return s, nil +} + +type cTable []cTableEntry + +func (c cTable) write(s *Scratch) error { + var ( + // precomputed conversion table + bitsToWeight [tableLogMax + 1]byte + huffLog = s.actualTableLog + // last weight is not saved. + maxSymbolValue = uint8(s.symbolLen - 1) + huffWeight = s.huffWeight[:256] + ) + const ( + maxFSETableLog = 6 + ) + // convert to weight + bitsToWeight[0] = 0 + for n := uint8(1); n < huffLog+1; n++ { + bitsToWeight[n] = huffLog + 1 - n + } + + // Acquire histogram for FSE. + hist := s.fse.Histogram() + hist = hist[:256] + for i := range hist[:16] { + hist[i] = 0 + } + for n := uint8(0); n < maxSymbolValue; n++ { + v := bitsToWeight[c[n].nBits] & 15 + huffWeight[n] = v + hist[v]++ + } + + // FSE compress if feasible. + if maxSymbolValue >= 2 { + huffMaxCnt := uint32(0) + huffMax := uint8(0) + for i, v := range hist[:16] { + if v == 0 { + continue + } + huffMax = byte(i) + if v > huffMaxCnt { + huffMaxCnt = v + } + } + s.fse.HistogramFinished(huffMax, int(huffMaxCnt)) + s.fse.TableLog = maxFSETableLog + b, err := fse.Compress(huffWeight[:maxSymbolValue], s.fse) + if err == nil && len(b) < int(s.symbolLen>>1) { + s.Out = append(s.Out, uint8(len(b))) + s.Out = append(s.Out, b...) + return nil + } + // Unable to compress (RLE/uncompressible) + } + // write raw values as 4-bits (max : 15) + if maxSymbolValue > (256 - 128) { + // should not happen : likely means source cannot be compressed + return ErrIncompressible + } + op := s.Out + // special case, pack weights 4 bits/weight. + op = append(op, 128|(maxSymbolValue-1)) + // be sure it doesn't cause msan issue in final combination + huffWeight[maxSymbolValue] = 0 + for n := uint16(0); n < uint16(maxSymbolValue); n += 2 { + op = append(op, (huffWeight[n]<<4)|huffWeight[n+1]) + } + s.Out = op + return nil +} + +func (c cTable) estTableSize(s *Scratch) (sz int, err error) { + var ( + // precomputed conversion table + bitsToWeight [tableLogMax + 1]byte + huffLog = s.actualTableLog + // last weight is not saved. + maxSymbolValue = uint8(s.symbolLen - 1) + huffWeight = s.huffWeight[:256] + ) + const ( + maxFSETableLog = 6 + ) + // convert to weight + bitsToWeight[0] = 0 + for n := uint8(1); n < huffLog+1; n++ { + bitsToWeight[n] = huffLog + 1 - n + } + + // Acquire histogram for FSE. + hist := s.fse.Histogram() + hist = hist[:256] + for i := range hist[:16] { + hist[i] = 0 + } + for n := uint8(0); n < maxSymbolValue; n++ { + v := bitsToWeight[c[n].nBits] & 15 + huffWeight[n] = v + hist[v]++ + } + + // FSE compress if feasible. + if maxSymbolValue >= 2 { + huffMaxCnt := uint32(0) + huffMax := uint8(0) + for i, v := range hist[:16] { + if v == 0 { + continue + } + huffMax = byte(i) + if v > huffMaxCnt { + huffMaxCnt = v + } + } + s.fse.HistogramFinished(huffMax, int(huffMaxCnt)) + s.fse.TableLog = maxFSETableLog + b, err := fse.Compress(huffWeight[:maxSymbolValue], s.fse) + if err == nil && len(b) < int(s.symbolLen>>1) { + sz += 1 + len(b) + return sz, nil + } + // Unable to compress (RLE/uncompressible) + } + // write raw values as 4-bits (max : 15) + if maxSymbolValue > (256 - 128) { + // should not happen : likely means source cannot be compressed + return 0, ErrIncompressible + } + // special case, pack weights 4 bits/weight. + sz += 1 + int(maxSymbolValue/2) + return sz, nil +} + +// estimateSize returns the estimated size in bytes of the input represented in the +// histogram supplied. +func (c cTable) estimateSize(hist []uint32) int { + nbBits := uint32(7) + for i, v := range c[:len(hist)] { + nbBits += uint32(v.nBits) * hist[i] + } + return int(nbBits >> 3) +} + +// minSize returns the minimum possible size considering the shannon limit. +func (s *Scratch) minSize(total int) int { + nbBits := float64(7) + fTotal := float64(total) + for _, v := range s.count[:s.symbolLen] { + n := float64(v) + if n > 0 { + nbBits += math.Log2(fTotal/n) * n + } + } + return int(nbBits) >> 3 +} + +func highBit32(val uint32) (n uint32) { + return uint32(bits.Len32(val) - 1) +} diff --git a/vendor/github.com/klauspost/compress/internal/snapref/LICENSE b/vendor/github.com/klauspost/compress/internal/snapref/LICENSE new file mode 100644 index 0000000..6050c10 --- /dev/null +++ b/vendor/github.com/klauspost/compress/internal/snapref/LICENSE @@ -0,0 +1,27 @@ +Copyright (c) 2011 The Snappy-Go Authors. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/github.com/klauspost/compress/internal/snapref/decode.go b/vendor/github.com/klauspost/compress/internal/snapref/decode.go new file mode 100644 index 0000000..40796a4 --- /dev/null +++ b/vendor/github.com/klauspost/compress/internal/snapref/decode.go @@ -0,0 +1,264 @@ +// Copyright 2011 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package snapref + +import ( + "encoding/binary" + "errors" + "io" +) + +var ( + // ErrCorrupt reports that the input is invalid. + ErrCorrupt = errors.New("snappy: corrupt input") + // ErrTooLarge reports that the uncompressed length is too large. + ErrTooLarge = errors.New("snappy: decoded block is too large") + // ErrUnsupported reports that the input isn't supported. + ErrUnsupported = errors.New("snappy: unsupported input") + + errUnsupportedLiteralLength = errors.New("snappy: unsupported literal length") +) + +// DecodedLen returns the length of the decoded block. +func DecodedLen(src []byte) (int, error) { + v, _, err := decodedLen(src) + return v, err +} + +// decodedLen returns the length of the decoded block and the number of bytes +// that the length header occupied. +func decodedLen(src []byte) (blockLen, headerLen int, err error) { + v, n := binary.Uvarint(src) + if n <= 0 || v > 0xffffffff { + return 0, 0, ErrCorrupt + } + + const wordSize = 32 << (^uint(0) >> 32 & 1) + if wordSize == 32 && v > 0x7fffffff { + return 0, 0, ErrTooLarge + } + return int(v), n, nil +} + +const ( + decodeErrCodeCorrupt = 1 + decodeErrCodeUnsupportedLiteralLength = 2 +) + +// Decode returns the decoded form of src. The returned slice may be a sub- +// slice of dst if dst was large enough to hold the entire decoded block. +// Otherwise, a newly allocated slice will be returned. +// +// The dst and src must not overlap. It is valid to pass a nil dst. +// +// Decode handles the Snappy block format, not the Snappy stream format. +func Decode(dst, src []byte) ([]byte, error) { + dLen, s, err := decodedLen(src) + if err != nil { + return nil, err + } + if dLen <= len(dst) { + dst = dst[:dLen] + } else { + dst = make([]byte, dLen) + } + switch decode(dst, src[s:]) { + case 0: + return dst, nil + case decodeErrCodeUnsupportedLiteralLength: + return nil, errUnsupportedLiteralLength + } + return nil, ErrCorrupt +} + +// NewReader returns a new Reader that decompresses from r, using the framing +// format described at +// https://github.com/google/snappy/blob/master/framing_format.txt +func NewReader(r io.Reader) *Reader { + return &Reader{ + r: r, + decoded: make([]byte, maxBlockSize), + buf: make([]byte, maxEncodedLenOfMaxBlockSize+checksumSize), + } +} + +// Reader is an io.Reader that can read Snappy-compressed bytes. +// +// Reader handles the Snappy stream format, not the Snappy block format. +type Reader struct { + r io.Reader + err error + decoded []byte + buf []byte + // decoded[i:j] contains decoded bytes that have not yet been passed on. + i, j int + readHeader bool +} + +// Reset discards any buffered data, resets all state, and switches the Snappy +// reader to read from r. This permits reusing a Reader rather than allocating +// a new one. +func (r *Reader) Reset(reader io.Reader) { + r.r = reader + r.err = nil + r.i = 0 + r.j = 0 + r.readHeader = false +} + +func (r *Reader) readFull(p []byte, allowEOF bool) (ok bool) { + if _, r.err = io.ReadFull(r.r, p); r.err != nil { + if r.err == io.ErrUnexpectedEOF || (r.err == io.EOF && !allowEOF) { + r.err = ErrCorrupt + } + return false + } + return true +} + +func (r *Reader) fill() error { + for r.i >= r.j { + if !r.readFull(r.buf[:4], true) { + return r.err + } + chunkType := r.buf[0] + if !r.readHeader { + if chunkType != chunkTypeStreamIdentifier { + r.err = ErrCorrupt + return r.err + } + r.readHeader = true + } + chunkLen := int(r.buf[1]) | int(r.buf[2])<<8 | int(r.buf[3])<<16 + if chunkLen > len(r.buf) { + r.err = ErrUnsupported + return r.err + } + + // The chunk types are specified at + // https://github.com/google/snappy/blob/master/framing_format.txt + switch chunkType { + case chunkTypeCompressedData: + // Section 4.2. Compressed data (chunk type 0x00). + if chunkLen < checksumSize { + r.err = ErrCorrupt + return r.err + } + buf := r.buf[:chunkLen] + if !r.readFull(buf, false) { + return r.err + } + checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24 + buf = buf[checksumSize:] + + n, err := DecodedLen(buf) + if err != nil { + r.err = err + return r.err + } + if n > len(r.decoded) { + r.err = ErrCorrupt + return r.err + } + if _, err := Decode(r.decoded, buf); err != nil { + r.err = err + return r.err + } + if crc(r.decoded[:n]) != checksum { + r.err = ErrCorrupt + return r.err + } + r.i, r.j = 0, n + continue + + case chunkTypeUncompressedData: + // Section 4.3. Uncompressed data (chunk type 0x01). + if chunkLen < checksumSize { + r.err = ErrCorrupt + return r.err + } + buf := r.buf[:checksumSize] + if !r.readFull(buf, false) { + return r.err + } + checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24 + // Read directly into r.decoded instead of via r.buf. + n := chunkLen - checksumSize + if n > len(r.decoded) { + r.err = ErrCorrupt + return r.err + } + if !r.readFull(r.decoded[:n], false) { + return r.err + } + if crc(r.decoded[:n]) != checksum { + r.err = ErrCorrupt + return r.err + } + r.i, r.j = 0, n + continue + + case chunkTypeStreamIdentifier: + // Section 4.1. Stream identifier (chunk type 0xff). + if chunkLen != len(magicBody) { + r.err = ErrCorrupt + return r.err + } + if !r.readFull(r.buf[:len(magicBody)], false) { + return r.err + } + for i := 0; i < len(magicBody); i++ { + if r.buf[i] != magicBody[i] { + r.err = ErrCorrupt + return r.err + } + } + continue + } + + if chunkType <= 0x7f { + // Section 4.5. Reserved unskippable chunks (chunk types 0x02-0x7f). + r.err = ErrUnsupported + return r.err + } + // Section 4.4 Padding (chunk type 0xfe). + // Section 4.6. Reserved skippable chunks (chunk types 0x80-0xfd). + if !r.readFull(r.buf[:chunkLen], false) { + return r.err + } + } + + return nil +} + +// Read satisfies the io.Reader interface. +func (r *Reader) Read(p []byte) (int, error) { + if r.err != nil { + return 0, r.err + } + + if err := r.fill(); err != nil { + return 0, err + } + + n := copy(p, r.decoded[r.i:r.j]) + r.i += n + return n, nil +} + +// ReadByte satisfies the io.ByteReader interface. +func (r *Reader) ReadByte() (byte, error) { + if r.err != nil { + return 0, r.err + } + + if err := r.fill(); err != nil { + return 0, err + } + + c := r.decoded[r.i] + r.i++ + return c, nil +} diff --git a/vendor/github.com/klauspost/compress/internal/snapref/decode_other.go b/vendor/github.com/klauspost/compress/internal/snapref/decode_other.go new file mode 100644 index 0000000..77395a6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/internal/snapref/decode_other.go @@ -0,0 +1,113 @@ +// Copyright 2016 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package snapref + +// decode writes the decoding of src to dst. It assumes that the varint-encoded +// length of the decompressed bytes has already been read, and that len(dst) +// equals that length. +// +// It returns 0 on success or a decodeErrCodeXxx error code on failure. +func decode(dst, src []byte) int { + var d, s, offset, length int + for s < len(src) { + switch src[s] & 0x03 { + case tagLiteral: + x := uint32(src[s] >> 2) + switch { + case x < 60: + s++ + case x == 60: + s += 2 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-1]) + case x == 61: + s += 3 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-2]) | uint32(src[s-1])<<8 + case x == 62: + s += 4 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16 + case x == 63: + s += 5 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24 + } + length = int(x) + 1 + if length <= 0 { + return decodeErrCodeUnsupportedLiteralLength + } + if length > len(dst)-d || length > len(src)-s { + return decodeErrCodeCorrupt + } + copy(dst[d:], src[s:s+length]) + d += length + s += length + continue + + case tagCopy1: + s += 2 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + length = 4 + int(src[s-2])>>2&0x7 + offset = int(uint32(src[s-2])&0xe0<<3 | uint32(src[s-1])) + + case tagCopy2: + s += 3 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + length = 1 + int(src[s-3])>>2 + offset = int(uint32(src[s-2]) | uint32(src[s-1])<<8) + + case tagCopy4: + s += 5 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + return decodeErrCodeCorrupt + } + length = 1 + int(src[s-5])>>2 + offset = int(uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24) + } + + if offset <= 0 || d < offset || length > len(dst)-d { + return decodeErrCodeCorrupt + } + // Copy from an earlier sub-slice of dst to a later sub-slice. + // If no overlap, use the built-in copy: + if offset >= length { + copy(dst[d:d+length], dst[d-offset:]) + d += length + continue + } + + // Unlike the built-in copy function, this byte-by-byte copy always runs + // forwards, even if the slices overlap. Conceptually, this is: + // + // d += forwardCopy(dst[d:d+length], dst[d-offset:]) + // + // We align the slices into a and b and show the compiler they are the same size. + // This allows the loop to run without bounds checks. + a := dst[d : d+length] + b := dst[d-offset:] + b = b[:len(a)] + for i := range a { + a[i] = b[i] + } + d += length + } + if d != len(dst) { + return decodeErrCodeCorrupt + } + return 0 +} diff --git a/vendor/github.com/klauspost/compress/internal/snapref/encode.go b/vendor/github.com/klauspost/compress/internal/snapref/encode.go new file mode 100644 index 0000000..13c6040 --- /dev/null +++ b/vendor/github.com/klauspost/compress/internal/snapref/encode.go @@ -0,0 +1,289 @@ +// Copyright 2011 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package snapref + +import ( + "encoding/binary" + "errors" + "io" +) + +// Encode returns the encoded form of src. The returned slice may be a sub- +// slice of dst if dst was large enough to hold the entire encoded block. +// Otherwise, a newly allocated slice will be returned. +// +// The dst and src must not overlap. It is valid to pass a nil dst. +// +// Encode handles the Snappy block format, not the Snappy stream format. +func Encode(dst, src []byte) []byte { + if n := MaxEncodedLen(len(src)); n < 0 { + panic(ErrTooLarge) + } else if len(dst) < n { + dst = make([]byte, n) + } + + // The block starts with the varint-encoded length of the decompressed bytes. + d := binary.PutUvarint(dst, uint64(len(src))) + + for len(src) > 0 { + p := src + src = nil + if len(p) > maxBlockSize { + p, src = p[:maxBlockSize], p[maxBlockSize:] + } + if len(p) < minNonLiteralBlockSize { + d += emitLiteral(dst[d:], p) + } else { + d += encodeBlock(dst[d:], p) + } + } + return dst[:d] +} + +// inputMargin is the minimum number of extra input bytes to keep, inside +// encodeBlock's inner loop. On some architectures, this margin lets us +// implement a fast path for emitLiteral, where the copy of short (<= 16 byte) +// literals can be implemented as a single load to and store from a 16-byte +// register. That literal's actual length can be as short as 1 byte, so this +// can copy up to 15 bytes too much, but that's OK as subsequent iterations of +// the encoding loop will fix up the copy overrun, and this inputMargin ensures +// that we don't overrun the dst and src buffers. +const inputMargin = 16 - 1 + +// minNonLiteralBlockSize is the minimum size of the input to encodeBlock that +// could be encoded with a copy tag. This is the minimum with respect to the +// algorithm used by encodeBlock, not a minimum enforced by the file format. +// +// The encoded output must start with at least a 1 byte literal, as there are +// no previous bytes to copy. A minimal (1 byte) copy after that, generated +// from an emitCopy call in encodeBlock's main loop, would require at least +// another inputMargin bytes, for the reason above: we want any emitLiteral +// calls inside encodeBlock's main loop to use the fast path if possible, which +// requires being able to overrun by inputMargin bytes. Thus, +// minNonLiteralBlockSize equals 1 + 1 + inputMargin. +// +// The C++ code doesn't use this exact threshold, but it could, as discussed at +// https://groups.google.com/d/topic/snappy-compression/oGbhsdIJSJ8/discussion +// The difference between Go (2+inputMargin) and C++ (inputMargin) is purely an +// optimization. It should not affect the encoded form. This is tested by +// TestSameEncodingAsCppShortCopies. +const minNonLiteralBlockSize = 1 + 1 + inputMargin + +// MaxEncodedLen returns the maximum length of a snappy block, given its +// uncompressed length. +// +// It will return a negative value if srcLen is too large to encode. +func MaxEncodedLen(srcLen int) int { + n := uint64(srcLen) + if n > 0xffffffff { + return -1 + } + // Compressed data can be defined as: + // compressed := item* literal* + // item := literal* copy + // + // The trailing literal sequence has a space blowup of at most 62/60 + // since a literal of length 60 needs one tag byte + one extra byte + // for length information. + // + // Item blowup is trickier to measure. Suppose the "copy" op copies + // 4 bytes of data. Because of a special check in the encoding code, + // we produce a 4-byte copy only if the offset is < 65536. Therefore + // the copy op takes 3 bytes to encode, and this type of item leads + // to at most the 62/60 blowup for representing literals. + // + // Suppose the "copy" op copies 5 bytes of data. If the offset is big + // enough, it will take 5 bytes to encode the copy op. Therefore the + // worst case here is a one-byte literal followed by a five-byte copy. + // That is, 6 bytes of input turn into 7 bytes of "compressed" data. + // + // This last factor dominates the blowup, so the final estimate is: + n = 32 + n + n/6 + if n > 0xffffffff { + return -1 + } + return int(n) +} + +var errClosed = errors.New("snappy: Writer is closed") + +// NewWriter returns a new Writer that compresses to w. +// +// The Writer returned does not buffer writes. There is no need to Flush or +// Close such a Writer. +// +// Deprecated: the Writer returned is not suitable for many small writes, only +// for few large writes. Use NewBufferedWriter instead, which is efficient +// regardless of the frequency and shape of the writes, and remember to Close +// that Writer when done. +func NewWriter(w io.Writer) *Writer { + return &Writer{ + w: w, + obuf: make([]byte, obufLen), + } +} + +// NewBufferedWriter returns a new Writer that compresses to w, using the +// framing format described at +// https://github.com/google/snappy/blob/master/framing_format.txt +// +// The Writer returned buffers writes. Users must call Close to guarantee all +// data has been forwarded to the underlying io.Writer. They may also call +// Flush zero or more times before calling Close. +func NewBufferedWriter(w io.Writer) *Writer { + return &Writer{ + w: w, + ibuf: make([]byte, 0, maxBlockSize), + obuf: make([]byte, obufLen), + } +} + +// Writer is an io.Writer that can write Snappy-compressed bytes. +// +// Writer handles the Snappy stream format, not the Snappy block format. +type Writer struct { + w io.Writer + err error + + // ibuf is a buffer for the incoming (uncompressed) bytes. + // + // Its use is optional. For backwards compatibility, Writers created by the + // NewWriter function have ibuf == nil, do not buffer incoming bytes, and + // therefore do not need to be Flush'ed or Close'd. + ibuf []byte + + // obuf is a buffer for the outgoing (compressed) bytes. + obuf []byte + + // wroteStreamHeader is whether we have written the stream header. + wroteStreamHeader bool +} + +// Reset discards the writer's state and switches the Snappy writer to write to +// w. This permits reusing a Writer rather than allocating a new one. +func (w *Writer) Reset(writer io.Writer) { + w.w = writer + w.err = nil + if w.ibuf != nil { + w.ibuf = w.ibuf[:0] + } + w.wroteStreamHeader = false +} + +// Write satisfies the io.Writer interface. +func (w *Writer) Write(p []byte) (nRet int, errRet error) { + if w.ibuf == nil { + // Do not buffer incoming bytes. This does not perform or compress well + // if the caller of Writer.Write writes many small slices. This + // behavior is therefore deprecated, but still supported for backwards + // compatibility with code that doesn't explicitly Flush or Close. + return w.write(p) + } + + // The remainder of this method is based on bufio.Writer.Write from the + // standard library. + + for len(p) > (cap(w.ibuf)-len(w.ibuf)) && w.err == nil { + var n int + if len(w.ibuf) == 0 { + // Large write, empty buffer. + // Write directly from p to avoid copy. + n, _ = w.write(p) + } else { + n = copy(w.ibuf[len(w.ibuf):cap(w.ibuf)], p) + w.ibuf = w.ibuf[:len(w.ibuf)+n] + w.Flush() + } + nRet += n + p = p[n:] + } + if w.err != nil { + return nRet, w.err + } + n := copy(w.ibuf[len(w.ibuf):cap(w.ibuf)], p) + w.ibuf = w.ibuf[:len(w.ibuf)+n] + nRet += n + return nRet, nil +} + +func (w *Writer) write(p []byte) (nRet int, errRet error) { + if w.err != nil { + return 0, w.err + } + for len(p) > 0 { + obufStart := len(magicChunk) + if !w.wroteStreamHeader { + w.wroteStreamHeader = true + copy(w.obuf, magicChunk) + obufStart = 0 + } + + var uncompressed []byte + if len(p) > maxBlockSize { + uncompressed, p = p[:maxBlockSize], p[maxBlockSize:] + } else { + uncompressed, p = p, nil + } + checksum := crc(uncompressed) + + // Compress the buffer, discarding the result if the improvement + // isn't at least 12.5%. + compressed := Encode(w.obuf[obufHeaderLen:], uncompressed) + chunkType := uint8(chunkTypeCompressedData) + chunkLen := 4 + len(compressed) + obufEnd := obufHeaderLen + len(compressed) + if len(compressed) >= len(uncompressed)-len(uncompressed)/8 { + chunkType = chunkTypeUncompressedData + chunkLen = 4 + len(uncompressed) + obufEnd = obufHeaderLen + } + + // Fill in the per-chunk header that comes before the body. + w.obuf[len(magicChunk)+0] = chunkType + w.obuf[len(magicChunk)+1] = uint8(chunkLen >> 0) + w.obuf[len(magicChunk)+2] = uint8(chunkLen >> 8) + w.obuf[len(magicChunk)+3] = uint8(chunkLen >> 16) + w.obuf[len(magicChunk)+4] = uint8(checksum >> 0) + w.obuf[len(magicChunk)+5] = uint8(checksum >> 8) + w.obuf[len(magicChunk)+6] = uint8(checksum >> 16) + w.obuf[len(magicChunk)+7] = uint8(checksum >> 24) + + if _, err := w.w.Write(w.obuf[obufStart:obufEnd]); err != nil { + w.err = err + return nRet, err + } + if chunkType == chunkTypeUncompressedData { + if _, err := w.w.Write(uncompressed); err != nil { + w.err = err + return nRet, err + } + } + nRet += len(uncompressed) + } + return nRet, nil +} + +// Flush flushes the Writer to its underlying io.Writer. +func (w *Writer) Flush() error { + if w.err != nil { + return w.err + } + if len(w.ibuf) == 0 { + return nil + } + w.write(w.ibuf) + w.ibuf = w.ibuf[:0] + return w.err +} + +// Close calls Flush and then closes the Writer. +func (w *Writer) Close() error { + w.Flush() + ret := w.err + if w.err == nil { + w.err = errClosed + } + return ret +} diff --git a/vendor/github.com/klauspost/compress/internal/snapref/encode_other.go b/vendor/github.com/klauspost/compress/internal/snapref/encode_other.go new file mode 100644 index 0000000..511bba6 --- /dev/null +++ b/vendor/github.com/klauspost/compress/internal/snapref/encode_other.go @@ -0,0 +1,236 @@ +// Copyright 2016 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package snapref + +func load32(b []byte, i int) uint32 { + b = b[i : i+4 : len(b)] // Help the compiler eliminate bounds checks on the next line. + return uint32(b[0]) | uint32(b[1])<<8 | uint32(b[2])<<16 | uint32(b[3])<<24 +} + +func load64(b []byte, i int) uint64 { + b = b[i : i+8 : len(b)] // Help the compiler eliminate bounds checks on the next line. + return uint64(b[0]) | uint64(b[1])<<8 | uint64(b[2])<<16 | uint64(b[3])<<24 | + uint64(b[4])<<32 | uint64(b[5])<<40 | uint64(b[6])<<48 | uint64(b[7])<<56 +} + +// emitLiteral writes a literal chunk and returns the number of bytes written. +// +// It assumes that: +// dst is long enough to hold the encoded bytes +// 1 <= len(lit) && len(lit) <= 65536 +func emitLiteral(dst, lit []byte) int { + i, n := 0, uint(len(lit)-1) + switch { + case n < 60: + dst[0] = uint8(n)<<2 | tagLiteral + i = 1 + case n < 1<<8: + dst[0] = 60<<2 | tagLiteral + dst[1] = uint8(n) + i = 2 + default: + dst[0] = 61<<2 | tagLiteral + dst[1] = uint8(n) + dst[2] = uint8(n >> 8) + i = 3 + } + return i + copy(dst[i:], lit) +} + +// emitCopy writes a copy chunk and returns the number of bytes written. +// +// It assumes that: +// dst is long enough to hold the encoded bytes +// 1 <= offset && offset <= 65535 +// 4 <= length && length <= 65535 +func emitCopy(dst []byte, offset, length int) int { + i := 0 + // The maximum length for a single tagCopy1 or tagCopy2 op is 64 bytes. The + // threshold for this loop is a little higher (at 68 = 64 + 4), and the + // length emitted down below is is a little lower (at 60 = 64 - 4), because + // it's shorter to encode a length 67 copy as a length 60 tagCopy2 followed + // by a length 7 tagCopy1 (which encodes as 3+2 bytes) than to encode it as + // a length 64 tagCopy2 followed by a length 3 tagCopy2 (which encodes as + // 3+3 bytes). The magic 4 in the 64±4 is because the minimum length for a + // tagCopy1 op is 4 bytes, which is why a length 3 copy has to be an + // encodes-as-3-bytes tagCopy2 instead of an encodes-as-2-bytes tagCopy1. + for length >= 68 { + // Emit a length 64 copy, encoded as 3 bytes. + dst[i+0] = 63<<2 | tagCopy2 + dst[i+1] = uint8(offset) + dst[i+2] = uint8(offset >> 8) + i += 3 + length -= 64 + } + if length > 64 { + // Emit a length 60 copy, encoded as 3 bytes. + dst[i+0] = 59<<2 | tagCopy2 + dst[i+1] = uint8(offset) + dst[i+2] = uint8(offset >> 8) + i += 3 + length -= 60 + } + if length >= 12 || offset >= 2048 { + // Emit the remaining copy, encoded as 3 bytes. + dst[i+0] = uint8(length-1)<<2 | tagCopy2 + dst[i+1] = uint8(offset) + dst[i+2] = uint8(offset >> 8) + return i + 3 + } + // Emit the remaining copy, encoded as 2 bytes. + dst[i+0] = uint8(offset>>8)<<5 | uint8(length-4)<<2 | tagCopy1 + dst[i+1] = uint8(offset) + return i + 2 +} + +// extendMatch returns the largest k such that k <= len(src) and that +// src[i:i+k-j] and src[j:k] have the same contents. +// +// It assumes that: +// 0 <= i && i < j && j <= len(src) +func extendMatch(src []byte, i, j int) int { + for ; j < len(src) && src[i] == src[j]; i, j = i+1, j+1 { + } + return j +} + +func hash(u, shift uint32) uint32 { + return (u * 0x1e35a7bd) >> shift +} + +// encodeBlock encodes a non-empty src to a guaranteed-large-enough dst. It +// assumes that the varint-encoded length of the decompressed bytes has already +// been written. +// +// It also assumes that: +// len(dst) >= MaxEncodedLen(len(src)) && +// minNonLiteralBlockSize <= len(src) && len(src) <= maxBlockSize +func encodeBlock(dst, src []byte) (d int) { + // Initialize the hash table. Its size ranges from 1<<8 to 1<<14 inclusive. + // The table element type is uint16, as s < sLimit and sLimit < len(src) + // and len(src) <= maxBlockSize and maxBlockSize == 65536. + const ( + maxTableSize = 1 << 14 + // tableMask is redundant, but helps the compiler eliminate bounds + // checks. + tableMask = maxTableSize - 1 + ) + shift := uint32(32 - 8) + for tableSize := 1 << 8; tableSize < maxTableSize && tableSize < len(src); tableSize *= 2 { + shift-- + } + // In Go, all array elements are zero-initialized, so there is no advantage + // to a smaller tableSize per se. However, it matches the C++ algorithm, + // and in the asm versions of this code, we can get away with zeroing only + // the first tableSize elements. + var table [maxTableSize]uint16 + + // sLimit is when to stop looking for offset/length copies. The inputMargin + // lets us use a fast path for emitLiteral in the main loop, while we are + // looking for copies. + sLimit := len(src) - inputMargin + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := 0 + + // The encoded form must start with a literal, as there are no previous + // bytes to copy, so we start looking for hash matches at s == 1. + s := 1 + nextHash := hash(load32(src, s), shift) + + for { + // Copied from the C++ snappy implementation: + // + // Heuristic match skipping: If 32 bytes are scanned with no matches + // found, start looking only at every other byte. If 32 more bytes are + // scanned (or skipped), look at every third byte, etc.. When a match + // is found, immediately go back to looking at every byte. This is a + // small loss (~5% performance, ~0.1% density) for compressible data + // due to more bookkeeping, but for non-compressible data (such as + // JPEG) it's a huge win since the compressor quickly "realizes" the + // data is incompressible and doesn't bother looking for matches + // everywhere. + // + // The "skip" variable keeps track of how many bytes there are since + // the last match; dividing it by 32 (ie. right-shifting by five) gives + // the number of bytes to move ahead for each iteration. + skip := 32 + + nextS := s + candidate := 0 + for { + s = nextS + bytesBetweenHashLookups := skip >> 5 + nextS = s + bytesBetweenHashLookups + skip += bytesBetweenHashLookups + if nextS > sLimit { + goto emitRemainder + } + candidate = int(table[nextHash&tableMask]) + table[nextHash&tableMask] = uint16(s) + nextHash = hash(load32(src, nextS), shift) + if load32(src, s) == load32(src, candidate) { + break + } + } + + // A 4-byte match has been found. We'll later see if more than 4 bytes + // match. But, prior to the match, src[nextEmit:s] are unmatched. Emit + // them as literal bytes. + d += emitLiteral(dst[d:], src[nextEmit:s]) + + // Call emitCopy, and then see if another emitCopy could be our next + // move. Repeat until we find no match for the input immediately after + // what was consumed by the last emitCopy call. + // + // If we exit this loop normally then we need to call emitLiteral next, + // though we don't yet know how big the literal will be. We handle that + // by proceeding to the next iteration of the main loop. We also can + // exit this loop via goto if we get close to exhausting the input. + for { + // Invariant: we have a 4-byte match at s, and no need to emit any + // literal bytes prior to s. + base := s + + // Extend the 4-byte match as long as possible. + // + // This is an inlined version of: + // s = extendMatch(src, candidate+4, s+4) + s += 4 + for i := candidate + 4; s < len(src) && src[i] == src[s]; i, s = i+1, s+1 { + } + + d += emitCopy(dst[d:], base-candidate, s-base) + nextEmit = s + if s >= sLimit { + goto emitRemainder + } + + // We could immediately start working at s now, but to improve + // compression we first update the hash table at s-1 and at s. If + // another emitCopy is not our next move, also calculate nextHash + // at s+1. At least on GOARCH=amd64, these three hash calculations + // are faster as one load64 call (with some shifts) instead of + // three load32 calls. + x := load64(src, s-1) + prevHash := hash(uint32(x>>0), shift) + table[prevHash&tableMask] = uint16(s - 1) + currHash := hash(uint32(x>>8), shift) + candidate = int(table[currHash&tableMask]) + table[currHash&tableMask] = uint16(s) + if uint32(x>>8) != load32(src, candidate) { + nextHash = hash(uint32(x>>16), shift) + s++ + break + } + } + } + +emitRemainder: + if nextEmit < len(src) { + d += emitLiteral(dst[d:], src[nextEmit:]) + } + return d +} diff --git a/vendor/github.com/klauspost/compress/internal/snapref/snappy.go b/vendor/github.com/klauspost/compress/internal/snapref/snappy.go new file mode 100644 index 0000000..34d01f4 --- /dev/null +++ b/vendor/github.com/klauspost/compress/internal/snapref/snappy.go @@ -0,0 +1,98 @@ +// Copyright 2011 The Snappy-Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package snapref implements the Snappy compression format. It aims for very +// high speeds and reasonable compression. +// +// There are actually two Snappy formats: block and stream. They are related, +// but different: trying to decompress block-compressed data as a Snappy stream +// will fail, and vice versa. The block format is the Decode and Encode +// functions and the stream format is the Reader and Writer types. +// +// The block format, the more common case, is used when the complete size (the +// number of bytes) of the original data is known upfront, at the time +// compression starts. The stream format, also known as the framing format, is +// for when that isn't always true. +// +// The canonical, C++ implementation is at https://github.com/google/snappy and +// it only implements the block format. +package snapref + +import ( + "hash/crc32" +) + +/* +Each encoded block begins with the varint-encoded length of the decoded data, +followed by a sequence of chunks. Chunks begin and end on byte boundaries. The +first byte of each chunk is broken into its 2 least and 6 most significant bits +called l and m: l ranges in [0, 4) and m ranges in [0, 64). l is the chunk tag. +Zero means a literal tag. All other values mean a copy tag. + +For literal tags: + - If m < 60, the next 1 + m bytes are literal bytes. + - Otherwise, let n be the little-endian unsigned integer denoted by the next + m - 59 bytes. The next 1 + n bytes after that are literal bytes. + +For copy tags, length bytes are copied from offset bytes ago, in the style of +Lempel-Ziv compression algorithms. In particular: + - For l == 1, the offset ranges in [0, 1<<11) and the length in [4, 12). + The length is 4 + the low 3 bits of m. The high 3 bits of m form bits 8-10 + of the offset. The next byte is bits 0-7 of the offset. + - For l == 2, the offset ranges in [0, 1<<16) and the length in [1, 65). + The length is 1 + m. The offset is the little-endian unsigned integer + denoted by the next 2 bytes. + - For l == 3, this tag is a legacy format that is no longer issued by most + encoders. Nonetheless, the offset ranges in [0, 1<<32) and the length in + [1, 65). The length is 1 + m. The offset is the little-endian unsigned + integer denoted by the next 4 bytes. +*/ +const ( + tagLiteral = 0x00 + tagCopy1 = 0x01 + tagCopy2 = 0x02 + tagCopy4 = 0x03 +) + +const ( + checksumSize = 4 + chunkHeaderSize = 4 + magicChunk = "\xff\x06\x00\x00" + magicBody + magicBody = "sNaPpY" + + // maxBlockSize is the maximum size of the input to encodeBlock. It is not + // part of the wire format per se, but some parts of the encoder assume + // that an offset fits into a uint16. + // + // Also, for the framing format (Writer type instead of Encode function), + // https://github.com/google/snappy/blob/master/framing_format.txt says + // that "the uncompressed data in a chunk must be no longer than 65536 + // bytes". + maxBlockSize = 65536 + + // maxEncodedLenOfMaxBlockSize equals MaxEncodedLen(maxBlockSize), but is + // hard coded to be a const instead of a variable, so that obufLen can also + // be a const. Their equivalence is confirmed by + // TestMaxEncodedLenOfMaxBlockSize. + maxEncodedLenOfMaxBlockSize = 76490 + + obufHeaderLen = len(magicChunk) + checksumSize + chunkHeaderSize + obufLen = obufHeaderLen + maxEncodedLenOfMaxBlockSize +) + +const ( + chunkTypeCompressedData = 0x00 + chunkTypeUncompressedData = 0x01 + chunkTypePadding = 0xfe + chunkTypeStreamIdentifier = 0xff +) + +var crcTable = crc32.MakeTable(crc32.Castagnoli) + +// crc implements the checksum specified in section 3 of +// https://github.com/google/snappy/blob/master/framing_format.txt +func crc(b []byte) uint32 { + c := crc32.Update(0, crcTable, b) + return uint32(c>>15|c<<17) + 0xa282ead8 +} diff --git a/vendor/github.com/klauspost/compress/s2sx.mod b/vendor/github.com/klauspost/compress/s2sx.mod new file mode 100644 index 0000000..2263853 --- /dev/null +++ b/vendor/github.com/klauspost/compress/s2sx.mod @@ -0,0 +1,4 @@ +module github.com/klauspost/compress + +go 1.16 + diff --git a/vendor/github.com/klauspost/compress/s2sx.sum b/vendor/github.com/klauspost/compress/s2sx.sum new file mode 100644 index 0000000..e69de29 diff --git a/vendor/github.com/klauspost/compress/zstd/README.md b/vendor/github.com/klauspost/compress/zstd/README.md new file mode 100644 index 0000000..e3445ac --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/README.md @@ -0,0 +1,455 @@ +# zstd + +[Zstandard](https://facebook.github.io/zstd/) is a real-time compression algorithm, providing high compression ratios. +It offers a very wide range of compression / speed trade-off, while being backed by a very fast decoder. +A high performance compression algorithm is implemented. For now focused on speed. + +This package provides [compression](#Compressor) to and [decompression](#Decompressor) of Zstandard content. + +This package is pure Go and without use of "unsafe". + +The `zstd` package is provided as open source software using a Go standard license. + +Currently the package is heavily optimized for 64 bit processors and will be significantly slower on 32 bit processors. + +## Installation + +Install using `go get -u github.com/klauspost/compress`. The package is located in `github.com/klauspost/compress/zstd`. + +[![Go Reference](https://pkg.go.dev/badge/github.com/klauspost/compress/zstd.svg)](https://pkg.go.dev/github.com/klauspost/compress/zstd) + +## Compressor + +### Status: + +STABLE - there may always be subtle bugs, a wide variety of content has been tested and the library is actively +used by several projects. This library is being [fuzz-tested](https://github.com/klauspost/compress-fuzz) for all updates. + +There may still be specific combinations of data types/size/settings that could lead to edge cases, +so as always, testing is recommended. + +For now, a high speed (fastest) and medium-fast (default) compressor has been implemented. + +* The "Fastest" compression ratio is roughly equivalent to zstd level 1. +* The "Default" compression ratio is roughly equivalent to zstd level 3 (default). +* The "Better" compression ratio is roughly equivalent to zstd level 7. +* The "Best" compression ratio is roughly equivalent to zstd level 11. + +In terms of speed, it is typically 2x as fast as the stdlib deflate/gzip in its fastest mode. +The compression ratio compared to stdlib is around level 3, but usually 3x as fast. + + +### Usage + +An Encoder can be used for either compressing a stream via the +`io.WriteCloser` interface supported by the Encoder or as multiple independent +tasks via the `EncodeAll` function. +Smaller encodes are encouraged to use the EncodeAll function. +Use `NewWriter` to create a new instance that can be used for both. + +To create a writer with default options, do like this: + +```Go +// Compress input to output. +func Compress(in io.Reader, out io.Writer) error { + enc, err := zstd.NewWriter(out) + if err != nil { + return err + } + _, err = io.Copy(enc, in) + if err != nil { + enc.Close() + return err + } + return enc.Close() +} +``` + +Now you can encode by writing data to `enc`. The output will be finished writing when `Close()` is called. +Even if your encode fails, you should still call `Close()` to release any resources that may be held up. + +The above is fine for big encodes. However, whenever possible try to *reuse* the writer. + +To reuse the encoder, you can use the `Reset(io.Writer)` function to change to another output. +This will allow the encoder to reuse all resources and avoid wasteful allocations. + +Currently stream encoding has 'light' concurrency, meaning up to 2 goroutines can be working on part +of a stream. This is independent of the `WithEncoderConcurrency(n)`, but that is likely to change +in the future. So if you want to limit concurrency for future updates, specify the concurrency +you would like. + +If you would like stream encoding to be done without spawning async goroutines, use `WithEncoderConcurrency(1)` +which will compress input as each block is completed, blocking on writes until each has completed. + +You can specify your desired compression level using `WithEncoderLevel()` option. Currently only pre-defined +compression settings can be specified. + +#### Future Compatibility Guarantees + +This will be an evolving project. When using this package it is important to note that both the compression efficiency and speed may change. + +The goal will be to keep the default efficiency at the default zstd (level 3). +However the encoding should never be assumed to remain the same, +and you should not use hashes of compressed output for similarity checks. + +The Encoder can be assumed to produce the same output from the exact same code version. +However, the may be modes in the future that break this, +although they will not be enabled without an explicit option. + +This encoder is not designed to (and will probably never) output the exact same bitstream as the reference encoder. + +Also note, that the cgo decompressor currently does not [report all errors on invalid input](https://github.com/DataDog/zstd/issues/59), +[omits error checks](https://github.com/DataDog/zstd/issues/61), [ignores checksums](https://github.com/DataDog/zstd/issues/43) +and seems to ignore concatenated streams, even though [it is part of the spec](https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#frames). + +#### Blocks + +For compressing small blocks, the returned encoder has a function called `EncodeAll(src, dst []byte) []byte`. + +`EncodeAll` will encode all input in src and append it to dst. +This function can be called concurrently. +Each call will only run on a same goroutine as the caller. + +Encoded blocks can be concatenated and the result will be the combined input stream. +Data compressed with EncodeAll can be decoded with the Decoder, using either a stream or `DecodeAll`. + +Especially when encoding blocks you should take special care to reuse the encoder. +This will effectively make it run without allocations after a warmup period. +To make it run completely without allocations, supply a destination buffer with space for all content. + +```Go +import "github.com/klauspost/compress/zstd" + +// Create a writer that caches compressors. +// For this operation type we supply a nil Reader. +var encoder, _ = zstd.NewWriter(nil) + +// Compress a buffer. +// If you have a destination buffer, the allocation in the call can also be eliminated. +func Compress(src []byte) []byte { + return encoder.EncodeAll(src, make([]byte, 0, len(src))) +} +``` + +You can control the maximum number of concurrent encodes using the `WithEncoderConcurrency(n)` +option when creating the writer. + +Using the Encoder for both a stream and individual blocks concurrently is safe. + +### Performance + +I have collected some speed examples to compare speed and compression against other compressors. + +* `file` is the input file. +* `out` is the compressor used. `zskp` is this package. `zstd` is the Datadog cgo library. `gzstd/gzkp` is gzip standard and this library. +* `level` is the compression level used. For `zskp` level 1 is "fastest", level 2 is "default"; 3 is "better", 4 is "best". +* `insize`/`outsize` is the input/output size. +* `millis` is the number of milliseconds used for compression. +* `mb/s` is megabytes (2^20 bytes) per second. + +``` +Silesia Corpus: +http://sun.aei.polsl.pl/~sdeor/corpus/silesia.zip + +This package: +file out level insize outsize millis mb/s +silesia.tar zskp 1 211947520 73821326 634 318.47 +silesia.tar zskp 2 211947520 67655404 1508 133.96 +silesia.tar zskp 3 211947520 64746933 3000 67.37 +silesia.tar zskp 4 211947520 60073508 16926 11.94 + +cgo zstd: +silesia.tar zstd 1 211947520 73605392 543 371.56 +silesia.tar zstd 3 211947520 66793289 864 233.68 +silesia.tar zstd 6 211947520 62916450 1913 105.66 +silesia.tar zstd 9 211947520 60212393 5063 39.92 + +gzip, stdlib/this package: +silesia.tar gzstd 1 211947520 80007735 1498 134.87 +silesia.tar gzkp 1 211947520 80088272 1009 200.31 + +GOB stream of binary data. Highly compressible. +https://files.klauspost.com/compress/gob-stream.7z + +file out level insize outsize millis mb/s +gob-stream zskp 1 1911399616 233948096 3230 564.34 +gob-stream zskp 2 1911399616 203997694 4997 364.73 +gob-stream zskp 3 1911399616 173526523 13435 135.68 +gob-stream zskp 4 1911399616 162195235 47559 38.33 + +gob-stream zstd 1 1911399616 249810424 2637 691.26 +gob-stream zstd 3 1911399616 208192146 3490 522.31 +gob-stream zstd 6 1911399616 193632038 6687 272.56 +gob-stream zstd 9 1911399616 177620386 16175 112.70 + +gob-stream gzstd 1 1911399616 357382013 9046 201.49 +gob-stream gzkp 1 1911399616 359136669 4885 373.08 + +The test data for the Large Text Compression Benchmark is the first +10^9 bytes of the English Wikipedia dump on Mar. 3, 2006. +http://mattmahoney.net/dc/textdata.html + +file out level insize outsize millis mb/s +enwik9 zskp 1 1000000000 343833605 3687 258.64 +enwik9 zskp 2 1000000000 317001237 7672 124.29 +enwik9 zskp 3 1000000000 291915823 15923 59.89 +enwik9 zskp 4 1000000000 261710291 77697 12.27 + +enwik9 zstd 1 1000000000 358072021 3110 306.65 +enwik9 zstd 3 1000000000 313734672 4784 199.35 +enwik9 zstd 6 1000000000 295138875 10290 92.68 +enwik9 zstd 9 1000000000 278348700 28549 33.40 + +enwik9 gzstd 1 1000000000 382578136 8608 110.78 +enwik9 gzkp 1 1000000000 382781160 5628 169.45 + +Highly compressible JSON file. +https://files.klauspost.com/compress/github-june-2days-2019.json.zst + +file out level insize outsize millis mb/s +github-june-2days-2019.json zskp 1 6273951764 697439532 9789 611.17 +github-june-2days-2019.json zskp 2 6273951764 610876538 18553 322.49 +github-june-2days-2019.json zskp 3 6273951764 517662858 44186 135.41 +github-june-2days-2019.json zskp 4 6273951764 464617114 165373 36.18 + +github-june-2days-2019.json zstd 1 6273951764 766284037 8450 708.00 +github-june-2days-2019.json zstd 3 6273951764 661889476 10927 547.57 +github-june-2days-2019.json zstd 6 6273951764 642756859 22996 260.18 +github-june-2days-2019.json zstd 9 6273951764 601974523 52413 114.16 + +github-june-2days-2019.json gzstd 1 6273951764 1164397768 26793 223.32 +github-june-2days-2019.json gzkp 1 6273951764 1120631856 17693 338.16 + +VM Image, Linux mint with a few installed applications: +https://files.klauspost.com/compress/rawstudio-mint14.7z + +file out level insize outsize millis mb/s +rawstudio-mint14.tar zskp 1 8558382592 3718400221 18206 448.29 +rawstudio-mint14.tar zskp 2 8558382592 3326118337 37074 220.15 +rawstudio-mint14.tar zskp 3 8558382592 3163842361 87306 93.49 +rawstudio-mint14.tar zskp 4 8558382592 2970480650 783862 10.41 + +rawstudio-mint14.tar zstd 1 8558382592 3609250104 17136 476.27 +rawstudio-mint14.tar zstd 3 8558382592 3341679997 29262 278.92 +rawstudio-mint14.tar zstd 6 8558382592 3235846406 77904 104.77 +rawstudio-mint14.tar zstd 9 8558382592 3160778861 140946 57.91 + +rawstudio-mint14.tar gzstd 1 8558382592 3926234992 51345 158.96 +rawstudio-mint14.tar gzkp 1 8558382592 3960117298 36722 222.26 + +CSV data: +https://files.klauspost.com/compress/nyc-taxi-data-10M.csv.zst + +file out level insize outsize millis mb/s +nyc-taxi-data-10M.csv zskp 1 3325605752 641319332 9462 335.17 +nyc-taxi-data-10M.csv zskp 2 3325605752 588976126 17570 180.50 +nyc-taxi-data-10M.csv zskp 3 3325605752 529329260 32432 97.79 +nyc-taxi-data-10M.csv zskp 4 3325605752 474949772 138025 22.98 + +nyc-taxi-data-10M.csv zstd 1 3325605752 687399637 8233 385.18 +nyc-taxi-data-10M.csv zstd 3 3325605752 598514411 10065 315.07 +nyc-taxi-data-10M.csv zstd 6 3325605752 570522953 20038 158.27 +nyc-taxi-data-10M.csv zstd 9 3325605752 517554797 64565 49.12 + +nyc-taxi-data-10M.csv gzstd 1 3325605752 928654908 21270 149.11 +nyc-taxi-data-10M.csv gzkp 1 3325605752 922273214 13929 227.68 +``` + +## Decompressor + +Staus: STABLE - there may still be subtle bugs, but a wide variety of content has been tested. + +This library is being continuously [fuzz-tested](https://github.com/klauspost/compress-fuzz), +kindly supplied by [fuzzit.dev](https://fuzzit.dev/). +The main purpose of the fuzz testing is to ensure that it is not possible to crash the decoder, +or run it past its limits with ANY input provided. + +### Usage + +The package has been designed for two main usages, big streams of data and smaller in-memory buffers. +There are two main usages of the package for these. Both of them are accessed by creating a `Decoder`. + +For streaming use a simple setup could look like this: + +```Go +import "github.com/klauspost/compress/zstd" + +func Decompress(in io.Reader, out io.Writer) error { + d, err := zstd.NewReader(in) + if err != nil { + return err + } + defer d.Close() + + // Copy content... + _, err = io.Copy(out, d) + return err +} +``` + +It is important to use the "Close" function when you no longer need the Reader to stop running goroutines, +when running with default settings. +Goroutines will exit once an error has been returned, including `io.EOF` at the end of a stream. + +Streams are decoded concurrently in 4 asynchronous stages to give the best possible throughput. +However, if you prefer synchronous decompression, use `WithDecoderConcurrency(1)` which will decompress data +as it is being requested only. + +For decoding buffers, it could look something like this: + +```Go +import "github.com/klauspost/compress/zstd" + +// Create a reader that caches decompressors. +// For this operation type we supply a nil Reader. +var decoder, _ = zstd.NewReader(nil, WithDecoderConcurrency(0)) + +// Decompress a buffer. We don't supply a destination buffer, +// so it will be allocated by the decoder. +func Decompress(src []byte) ([]byte, error) { + return decoder.DecodeAll(src, nil) +} +``` + +Both of these cases should provide the functionality needed. +The decoder can be used for *concurrent* decompression of multiple buffers. +By default 4 decompressors will be created. + +It will only allow a certain number of concurrent operations to run. +To tweak that yourself use the `WithDecoderConcurrency(n)` option when creating the decoder. +It is possible to use `WithDecoderConcurrency(0)` to create GOMAXPROCS decoders. + +### Dictionaries + +Data compressed with [dictionaries](https://github.com/facebook/zstd#the-case-for-small-data-compression) can be decompressed. + +Dictionaries are added individually to Decoders. +Dictionaries are generated by the `zstd --train` command and contains an initial state for the decoder. +To add a dictionary use the `WithDecoderDicts(dicts ...[]byte)` option with the dictionary data. +Several dictionaries can be added at once. + +The dictionary will be used automatically for the data that specifies them. +A re-used Decoder will still contain the dictionaries registered. + +When registering multiple dictionaries with the same ID, the last one will be used. + +It is possible to use dictionaries when compressing data. + +To enable a dictionary use `WithEncoderDict(dict []byte)`. Here only one dictionary will be used +and it will likely be used even if it doesn't improve compression. + +The used dictionary must be used to decompress the content. + +For any real gains, the dictionary should be built with similar data. +If an unsuitable dictionary is used the output may be slightly larger than using no dictionary. +Use the [zstd commandline tool](https://github.com/facebook/zstd/releases) to build a dictionary from sample data. +For information see [zstd dictionary information](https://github.com/facebook/zstd#the-case-for-small-data-compression). + +For now there is a fixed startup performance penalty for compressing content with dictionaries. +This will likely be improved over time. Just be aware to test performance when implementing. + +### Allocation-less operation + +The decoder has been designed to operate without allocations after a warmup. + +This means that you should *store* the decoder for best performance. +To re-use a stream decoder, use the `Reset(r io.Reader) error` to switch to another stream. +A decoder can safely be re-used even if the previous stream failed. + +To release the resources, you must call the `Close()` function on a decoder. +After this it can *no longer be reused*, but all running goroutines will be stopped. +So you *must* use this if you will no longer need the Reader. + +For decompressing smaller buffers a single decoder can be used. +When decoding buffers, you can supply a destination slice with length 0 and your expected capacity. +In this case no unneeded allocations should be made. + +### Concurrency + +The buffer decoder does everything on the same goroutine and does nothing concurrently. +It can however decode several buffers concurrently. Use `WithDecoderConcurrency(n)` to limit that. + +The stream decoder will create goroutines that: + +1) Reads input and splits the input into blocks. +2) Decompression of literals. +3) Decompression of sequences. +4) Reconstruction of output stream. + +So effectively this also means the decoder will "read ahead" and prepare data to always be available for output. + +The concurrency level will, for streams, determine how many blocks ahead the compression will start. + +Since "blocks" are quite dependent on the output of the previous block stream decoding will only have limited concurrency. + +In practice this means that concurrency is often limited to utilizing about 3 cores effectively. + +### Benchmarks + +These are some examples of performance compared to [datadog cgo library](https://github.com/DataDog/zstd). + +The first two are streaming decodes and the last are smaller inputs. + +``` +BenchmarkDecoderSilesia-8 3 385000067 ns/op 550.51 MB/s 5498 B/op 8 allocs/op +BenchmarkDecoderSilesiaCgo-8 6 197666567 ns/op 1072.25 MB/s 270672 B/op 8 allocs/op + +BenchmarkDecoderEnwik9-8 1 2027001600 ns/op 493.34 MB/s 10496 B/op 18 allocs/op +BenchmarkDecoderEnwik9Cgo-8 2 979499200 ns/op 1020.93 MB/s 270672 B/op 8 allocs/op + +Concurrent performance: + +BenchmarkDecoder_DecodeAllParallel/kppkn.gtb.zst-16 28915 42469 ns/op 4340.07 MB/s 114 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/geo.protodata.zst-16 116505 9965 ns/op 11900.16 MB/s 16 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/plrabn12.txt.zst-16 8952 134272 ns/op 3588.70 MB/s 915 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/lcet10.txt.zst-16 11820 102538 ns/op 4161.90 MB/s 594 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/asyoulik.txt.zst-16 34782 34184 ns/op 3661.88 MB/s 60 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/alice29.txt.zst-16 27712 43447 ns/op 3500.58 MB/s 99 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/html_x_4.zst-16 62826 18750 ns/op 21845.10 MB/s 104 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/paper-100k.pdf.zst-16 631545 1794 ns/op 57078.74 MB/s 2 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/fireworks.jpeg.zst-16 1690140 712 ns/op 172938.13 MB/s 1 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/urls.10K.zst-16 10432 113593 ns/op 6180.73 MB/s 1143 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/html.zst-16 113206 10671 ns/op 9596.27 MB/s 15 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallel/comp-data.bin.zst-16 1530615 779 ns/op 5229.49 MB/s 0 B/op 0 allocs/op + +BenchmarkDecoder_DecodeAllParallelCgo/kppkn.gtb.zst-16 65217 16192 ns/op 11383.34 MB/s 46 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/geo.protodata.zst-16 292671 4039 ns/op 29363.19 MB/s 6 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/plrabn12.txt.zst-16 26314 46021 ns/op 10470.43 MB/s 293 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/lcet10.txt.zst-16 33897 34900 ns/op 12227.96 MB/s 205 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/asyoulik.txt.zst-16 104348 11433 ns/op 10949.01 MB/s 20 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/alice29.txt.zst-16 75949 15510 ns/op 9805.60 MB/s 32 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/html_x_4.zst-16 173910 6756 ns/op 60624.29 MB/s 37 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/paper-100k.pdf.zst-16 923076 1339 ns/op 76474.87 MB/s 1 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/fireworks.jpeg.zst-16 922920 1351 ns/op 91102.57 MB/s 2 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/urls.10K.zst-16 27649 43618 ns/op 16096.19 MB/s 407 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/html.zst-16 279073 4160 ns/op 24614.18 MB/s 6 B/op 0 allocs/op +BenchmarkDecoder_DecodeAllParallelCgo/comp-data.bin.zst-16 749938 1579 ns/op 2581.71 MB/s 0 B/op 0 allocs/op +``` + +This reflects the performance around May 2020, but this may be out of date. + +## Zstd inside ZIP files + +It is possible to use zstandard to compress individual files inside zip archives. +While this isn't widely supported it can be useful for internal files. + +To support the compression and decompression of these files you must register a compressor and decompressor. + +It is highly recommended registering the (de)compressors on individual zip Reader/Writer and NOT +use the global registration functions. The main reason for this is that 2 registrations from +different packages will result in a panic. + +It is a good idea to only have a single compressor and decompressor, since they can be used for multiple zip +files concurrently, and using a single instance will allow reusing some resources. + +See [this example](https://pkg.go.dev/github.com/klauspost/compress/zstd#example-ZipCompressor) for +how to compress and decompress files inside zip archives. + +# Contributions + +Contributions are always welcome. +For new features/fixes, remember to add tests and for performance enhancements include benchmarks. + +For general feedback and experience reports, feel free to open an issue or write me on [Twitter](https://twitter.com/sh0dan). + +This package includes the excellent [`github.com/cespare/xxhash`](https://github.com/cespare/xxhash) package Copyright (c) 2016 Caleb Spare. diff --git a/vendor/github.com/klauspost/compress/zstd/bitreader.go b/vendor/github.com/klauspost/compress/zstd/bitreader.go new file mode 100644 index 0000000..d7cd15b --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/bitreader.go @@ -0,0 +1,147 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "encoding/binary" + "errors" + "fmt" + "io" + "math/bits" +) + +// bitReader reads a bitstream in reverse. +// The last set bit indicates the start of the stream and is used +// for aligning the input. +type bitReader struct { + in []byte + off uint // next byte to read is at in[off - 1] + value uint64 // Maybe use [16]byte, but shifting is awkward. + bitsRead uint8 +} + +// init initializes and resets the bit reader. +func (b *bitReader) init(in []byte) error { + if len(in) < 1 { + return errors.New("corrupt stream: too short") + } + b.in = in + b.off = uint(len(in)) + // The highest bit of the last byte indicates where to start + v := in[len(in)-1] + if v == 0 { + return errors.New("corrupt stream, did not find end of stream") + } + b.bitsRead = 64 + b.value = 0 + if len(in) >= 8 { + b.fillFastStart() + } else { + b.fill() + b.fill() + } + b.bitsRead += 8 - uint8(highBits(uint32(v))) + return nil +} + +// getBits will return n bits. n can be 0. +func (b *bitReader) getBits(n uint8) int { + if n == 0 /*|| b.bitsRead >= 64 */ { + return 0 + } + return int(b.get32BitsFast(n)) +} + +// get32BitsFast requires that at least one bit is requested every time. +// There are no checks if the buffer is filled. +func (b *bitReader) get32BitsFast(n uint8) uint32 { + const regMask = 64 - 1 + v := uint32((b.value << (b.bitsRead & regMask)) >> ((regMask + 1 - n) & regMask)) + b.bitsRead += n + return v +} + +func (b *bitReader) get16BitsFast(n uint8) uint16 { + const regMask = 64 - 1 + v := uint16((b.value << (b.bitsRead & regMask)) >> ((regMask + 1 - n) & regMask)) + b.bitsRead += n + return v +} + +// fillFast() will make sure at least 32 bits are available. +// There must be at least 4 bytes available. +func (b *bitReader) fillFast() { + if b.bitsRead < 32 { + return + } + // 2 bounds checks. + v := b.in[b.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value = (b.value << 32) | uint64(low) + b.bitsRead -= 32 + b.off -= 4 +} + +// fillFastStart() assumes the bitreader is empty and there is at least 8 bytes to read. +func (b *bitReader) fillFastStart() { + // Do single re-slice to avoid bounds checks. + b.value = binary.LittleEndian.Uint64(b.in[b.off-8:]) + b.bitsRead = 0 + b.off -= 8 +} + +// fill() will make sure at least 32 bits are available. +func (b *bitReader) fill() { + if b.bitsRead < 32 { + return + } + if b.off >= 4 { + v := b.in[b.off-4:] + v = v[:4] + low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24) + b.value = (b.value << 32) | uint64(low) + b.bitsRead -= 32 + b.off -= 4 + return + } + for b.off > 0 { + b.value = (b.value << 8) | uint64(b.in[b.off-1]) + b.bitsRead -= 8 + b.off-- + } +} + +// finished returns true if all bits have been read from the bit stream. +func (b *bitReader) finished() bool { + return b.off == 0 && b.bitsRead >= 64 +} + +// overread returns true if more bits have been requested than is on the stream. +func (b *bitReader) overread() bool { + return b.bitsRead > 64 +} + +// remain returns the number of bits remaining. +func (b *bitReader) remain() uint { + return b.off*8 + 64 - uint(b.bitsRead) +} + +// close the bitstream and returns an error if out-of-buffer reads occurred. +func (b *bitReader) close() error { + // Release reference. + b.in = nil + if !b.finished() { + return fmt.Errorf("%d extra bits on block, should be 0", b.remain()) + } + if b.bitsRead > 64 { + return io.ErrUnexpectedEOF + } + return nil +} + +func highBits(val uint32) (n uint32) { + return uint32(bits.Len32(val) - 1) +} diff --git a/vendor/github.com/klauspost/compress/zstd/bitwriter.go b/vendor/github.com/klauspost/compress/zstd/bitwriter.go new file mode 100644 index 0000000..b366182 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/bitwriter.go @@ -0,0 +1,189 @@ +// Copyright 2018 Klaus Post. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// Based on work Copyright (c) 2013, Yann Collet, released under BSD License. + +package zstd + +import "fmt" + +// bitWriter will write bits. +// First bit will be LSB of the first byte of output. +type bitWriter struct { + bitContainer uint64 + nBits uint8 + out []byte +} + +// bitMask16 is bitmasks. Has extra to avoid bounds check. +var bitMask16 = [32]uint16{ + 0, 1, 3, 7, 0xF, 0x1F, + 0x3F, 0x7F, 0xFF, 0x1FF, 0x3FF, 0x7FF, + 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0xFFFF, + 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, + 0xFFFF, 0xFFFF} /* up to 16 bits */ + +var bitMask32 = [32]uint32{ + 0, 1, 3, 7, 0xF, 0x1F, 0x3F, 0x7F, 0xFF, + 0x1FF, 0x3FF, 0x7FF, 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, + 0x1ffff, 0x3ffff, 0x7FFFF, 0xfFFFF, 0x1fFFFF, 0x3fFFFF, 0x7fFFFF, 0xffFFFF, + 0x1ffFFFF, 0x3ffFFFF, 0x7ffFFFF, 0xfffFFFF, 0x1fffFFFF, 0x3fffFFFF, 0x7fffFFFF, +} // up to 32 bits + +// addBits16NC will add up to 16 bits. +// It will not check if there is space for them, +// so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits16NC(value uint16, bits uint8) { + b.bitContainer |= uint64(value&bitMask16[bits&31]) << (b.nBits & 63) + b.nBits += bits +} + +// addBits32NC will add up to 31 bits. +// It will not check if there is space for them, +// so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits32NC(value uint32, bits uint8) { + b.bitContainer |= uint64(value&bitMask32[bits&31]) << (b.nBits & 63) + b.nBits += bits +} + +// addBits64NC will add up to 64 bits. +// There must be space for 32 bits. +func (b *bitWriter) addBits64NC(value uint64, bits uint8) { + if bits <= 31 { + b.addBits32Clean(uint32(value), bits) + return + } + b.addBits32Clean(uint32(value), 32) + b.flush32() + b.addBits32Clean(uint32(value>>32), bits-32) +} + +// addBits32Clean will add up to 32 bits. +// It will not check if there is space for them. +// The input must not contain more bits than specified. +func (b *bitWriter) addBits32Clean(value uint32, bits uint8) { + b.bitContainer |= uint64(value) << (b.nBits & 63) + b.nBits += bits +} + +// addBits16Clean will add up to 16 bits. value may not contain more set bits than indicated. +// It will not check if there is space for them, so the caller must ensure that it has flushed recently. +func (b *bitWriter) addBits16Clean(value uint16, bits uint8) { + b.bitContainer |= uint64(value) << (b.nBits & 63) + b.nBits += bits +} + +// flush will flush all pending full bytes. +// There will be at least 56 bits available for writing when this has been called. +// Using flush32 is faster, but leaves less space for writing. +func (b *bitWriter) flush() { + v := b.nBits >> 3 + switch v { + case 0: + case 1: + b.out = append(b.out, + byte(b.bitContainer), + ) + case 2: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + ) + case 3: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + ) + case 4: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + ) + case 5: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + ) + case 6: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + ) + case 7: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + byte(b.bitContainer>>48), + ) + case 8: + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24), + byte(b.bitContainer>>32), + byte(b.bitContainer>>40), + byte(b.bitContainer>>48), + byte(b.bitContainer>>56), + ) + default: + panic(fmt.Errorf("bits (%d) > 64", b.nBits)) + } + b.bitContainer >>= v << 3 + b.nBits &= 7 +} + +// flush32 will flush out, so there are at least 32 bits available for writing. +func (b *bitWriter) flush32() { + if b.nBits < 32 { + return + } + b.out = append(b.out, + byte(b.bitContainer), + byte(b.bitContainer>>8), + byte(b.bitContainer>>16), + byte(b.bitContainer>>24)) + b.nBits -= 32 + b.bitContainer >>= 32 +} + +// flushAlign will flush remaining full bytes and align to next byte boundary. +func (b *bitWriter) flushAlign() { + nbBytes := (b.nBits + 7) >> 3 + for i := uint8(0); i < nbBytes; i++ { + b.out = append(b.out, byte(b.bitContainer>>(i*8))) + } + b.nBits = 0 + b.bitContainer = 0 +} + +// close will write the alignment bit and write the final byte(s) +// to the output. +func (b *bitWriter) close() error { + // End mark + b.addBits16Clean(1, 1) + // flush until next byte. + b.flushAlign() + return nil +} + +// reset and continue writing by appending to out. +func (b *bitWriter) reset(out []byte) { + b.bitContainer = 0 + b.nBits = 0 + b.out = out +} diff --git a/vendor/github.com/klauspost/compress/zstd/blockdec.go b/vendor/github.com/klauspost/compress/zstd/blockdec.go new file mode 100644 index 0000000..7d567a5 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/blockdec.go @@ -0,0 +1,698 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "errors" + "fmt" + "io" + "sync" + + "github.com/klauspost/compress/huff0" + "github.com/klauspost/compress/zstd/internal/xxhash" +) + +type blockType uint8 + +//go:generate stringer -type=blockType,literalsBlockType,seqCompMode,tableIndex + +const ( + blockTypeRaw blockType = iota + blockTypeRLE + blockTypeCompressed + blockTypeReserved +) + +type literalsBlockType uint8 + +const ( + literalsBlockRaw literalsBlockType = iota + literalsBlockRLE + literalsBlockCompressed + literalsBlockTreeless +) + +const ( + // maxCompressedBlockSize is the biggest allowed compressed block size (128KB) + maxCompressedBlockSize = 128 << 10 + + // Maximum possible block size (all Raw+Uncompressed). + maxBlockSize = (1 << 21) - 1 + + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#literals_section_header + maxCompressedLiteralSize = 1 << 18 + maxRLELiteralSize = 1 << 20 + maxMatchLen = 131074 + maxSequences = 0x7f00 + 0xffff + + // We support slightly less than the reference decoder to be able to + // use ints on 32 bit archs. + maxOffsetBits = 30 +) + +var ( + huffDecoderPool = sync.Pool{New: func() interface{} { + return &huff0.Scratch{} + }} + + fseDecoderPool = sync.Pool{New: func() interface{} { + return &fseDecoder{} + }} +) + +type blockDec struct { + // Raw source data of the block. + data []byte + dataStorage []byte + + // Destination of the decoded data. + dst []byte + + // Buffer for literals data. + literalBuf []byte + + // Window size of the block. + WindowSize uint64 + + err error + + // Check against this crc + checkCRC []byte + + // Frame to use for singlethreaded decoding. + // Should not be used by the decoder itself since parent may be another frame. + localFrame *frameDec + + sequence []seqVals + + async struct { + newHist *history + literals []byte + seqData []byte + seqSize int // Size of uncompressed sequences + fcs uint64 + } + + // Block is RLE, this is the size. + RLESize uint32 + tmp [4]byte + + Type blockType + + // Is this the last block of a frame? + Last bool + + // Use less memory + lowMem bool +} + +func (b *blockDec) String() string { + if b == nil { + return "" + } + return fmt.Sprintf("Steam Size: %d, Type: %v, Last: %t, Window: %d", len(b.data), b.Type, b.Last, b.WindowSize) +} + +func newBlockDec(lowMem bool) *blockDec { + b := blockDec{ + lowMem: lowMem, + } + return &b +} + +// reset will reset the block. +// Input must be a start of a block and will be at the end of the block when returned. +func (b *blockDec) reset(br byteBuffer, windowSize uint64) error { + b.WindowSize = windowSize + tmp, err := br.readSmall(3) + if err != nil { + println("Reading block header:", err) + return err + } + bh := uint32(tmp[0]) | (uint32(tmp[1]) << 8) | (uint32(tmp[2]) << 16) + b.Last = bh&1 != 0 + b.Type = blockType((bh >> 1) & 3) + // find size. + cSize := int(bh >> 3) + maxSize := maxBlockSize + switch b.Type { + case blockTypeReserved: + return ErrReservedBlockType + case blockTypeRLE: + if cSize > maxCompressedBlockSize || cSize > int(b.WindowSize) { + if debugDecoder { + printf("rle block too big: csize:%d block: %+v\n", uint64(cSize), b) + } + return ErrWindowSizeExceeded + } + b.RLESize = uint32(cSize) + if b.lowMem { + maxSize = cSize + } + cSize = 1 + case blockTypeCompressed: + if debugDecoder { + println("Data size on stream:", cSize) + } + b.RLESize = 0 + maxSize = maxCompressedBlockSize + if windowSize < maxCompressedBlockSize && b.lowMem { + maxSize = int(windowSize) + } + if cSize > maxCompressedBlockSize || uint64(cSize) > b.WindowSize { + if debugDecoder { + printf("compressed block too big: csize:%d block: %+v\n", uint64(cSize), b) + } + return ErrCompressedSizeTooBig + } + // Empty compressed blocks must at least be 2 bytes + // for Literals_Block_Type and one for Sequences_Section_Header. + if cSize < 2 { + return ErrBlockTooSmall + } + case blockTypeRaw: + if cSize > maxCompressedBlockSize || cSize > int(b.WindowSize) { + if debugDecoder { + printf("rle block too big: csize:%d block: %+v\n", uint64(cSize), b) + } + return ErrWindowSizeExceeded + } + + b.RLESize = 0 + // We do not need a destination for raw blocks. + maxSize = -1 + default: + panic("Invalid block type") + } + + // Read block data. + if cap(b.dataStorage) < cSize { + if b.lowMem || cSize > maxCompressedBlockSize { + b.dataStorage = make([]byte, 0, cSize) + } else { + b.dataStorage = make([]byte, 0, maxCompressedBlockSize) + } + } + if cap(b.dst) <= maxSize { + b.dst = make([]byte, 0, maxSize+1) + } + b.data, err = br.readBig(cSize, b.dataStorage) + if err != nil { + if debugDecoder { + println("Reading block:", err, "(", cSize, ")", len(b.data)) + printf("%T", br) + } + return err + } + return nil +} + +// sendEOF will make the decoder send EOF on this frame. +func (b *blockDec) sendErr(err error) { + b.Last = true + b.Type = blockTypeReserved + b.err = err +} + +// Close will release resources. +// Closed blockDec cannot be reset. +func (b *blockDec) Close() { +} + +// decodeBuf +func (b *blockDec) decodeBuf(hist *history) error { + switch b.Type { + case blockTypeRLE: + if cap(b.dst) < int(b.RLESize) { + if b.lowMem { + b.dst = make([]byte, b.RLESize) + } else { + b.dst = make([]byte, maxBlockSize) + } + } + b.dst = b.dst[:b.RLESize] + v := b.data[0] + for i := range b.dst { + b.dst[i] = v + } + hist.appendKeep(b.dst) + return nil + case blockTypeRaw: + hist.appendKeep(b.data) + return nil + case blockTypeCompressed: + saved := b.dst + // Append directly to history + if hist.ignoreBuffer == 0 { + b.dst = hist.b + hist.b = nil + } else { + b.dst = b.dst[:0] + } + err := b.decodeCompressed(hist) + if debugDecoder { + println("Decompressed to total", len(b.dst), "bytes, hash:", xxhash.Sum64(b.dst), "error:", err) + } + if hist.ignoreBuffer == 0 { + hist.b = b.dst + b.dst = saved + } else { + hist.appendKeep(b.dst) + } + return err + case blockTypeReserved: + // Used for returning errors. + return b.err + default: + panic("Invalid block type") + } +} + +func (b *blockDec) decodeLiterals(in []byte, hist *history) (remain []byte, err error) { + // There must be at least one byte for Literals_Block_Type and one for Sequences_Section_Header + if len(in) < 2 { + return in, ErrBlockTooSmall + } + + litType := literalsBlockType(in[0] & 3) + var litRegenSize int + var litCompSize int + sizeFormat := (in[0] >> 2) & 3 + var fourStreams bool + var literals []byte + switch litType { + case literalsBlockRaw, literalsBlockRLE: + switch sizeFormat { + case 0, 2: + // Regenerated_Size uses 5 bits (0-31). Literals_Section_Header uses 1 byte. + litRegenSize = int(in[0] >> 3) + in = in[1:] + case 1: + // Regenerated_Size uses 12 bits (0-4095). Literals_Section_Header uses 2 bytes. + litRegenSize = int(in[0]>>4) + (int(in[1]) << 4) + in = in[2:] + case 3: + // Regenerated_Size uses 20 bits (0-1048575). Literals_Section_Header uses 3 bytes. + if len(in) < 3 { + println("too small: litType:", litType, " sizeFormat", sizeFormat, len(in)) + return in, ErrBlockTooSmall + } + litRegenSize = int(in[0]>>4) + (int(in[1]) << 4) + (int(in[2]) << 12) + in = in[3:] + } + case literalsBlockCompressed, literalsBlockTreeless: + switch sizeFormat { + case 0, 1: + // Both Regenerated_Size and Compressed_Size use 10 bits (0-1023). + if len(in) < 3 { + println("too small: litType:", litType, " sizeFormat", sizeFormat, len(in)) + return in, ErrBlockTooSmall + } + n := uint64(in[0]>>4) + (uint64(in[1]) << 4) + (uint64(in[2]) << 12) + litRegenSize = int(n & 1023) + litCompSize = int(n >> 10) + fourStreams = sizeFormat == 1 + in = in[3:] + case 2: + fourStreams = true + if len(in) < 4 { + println("too small: litType:", litType, " sizeFormat", sizeFormat, len(in)) + return in, ErrBlockTooSmall + } + n := uint64(in[0]>>4) + (uint64(in[1]) << 4) + (uint64(in[2]) << 12) + (uint64(in[3]) << 20) + litRegenSize = int(n & 16383) + litCompSize = int(n >> 14) + in = in[4:] + case 3: + fourStreams = true + if len(in) < 5 { + println("too small: litType:", litType, " sizeFormat", sizeFormat, len(in)) + return in, ErrBlockTooSmall + } + n := uint64(in[0]>>4) + (uint64(in[1]) << 4) + (uint64(in[2]) << 12) + (uint64(in[3]) << 20) + (uint64(in[4]) << 28) + litRegenSize = int(n & 262143) + litCompSize = int(n >> 18) + in = in[5:] + } + } + if debugDecoder { + println("literals type:", litType, "litRegenSize:", litRegenSize, "litCompSize:", litCompSize, "sizeFormat:", sizeFormat, "4X:", fourStreams) + } + if litRegenSize > int(b.WindowSize) || litRegenSize > maxCompressedBlockSize { + return in, ErrWindowSizeExceeded + } + + switch litType { + case literalsBlockRaw: + if len(in) < litRegenSize { + println("too small: litType:", litType, " sizeFormat", sizeFormat, "remain:", len(in), "want:", litRegenSize) + return in, ErrBlockTooSmall + } + literals = in[:litRegenSize] + in = in[litRegenSize:] + //printf("Found %d uncompressed literals\n", litRegenSize) + case literalsBlockRLE: + if len(in) < 1 { + println("too small: litType:", litType, " sizeFormat", sizeFormat, "remain:", len(in), "want:", 1) + return in, ErrBlockTooSmall + } + if cap(b.literalBuf) < litRegenSize { + if b.lowMem { + b.literalBuf = make([]byte, litRegenSize) + } else { + if litRegenSize > maxCompressedLiteralSize { + // Exceptional + b.literalBuf = make([]byte, litRegenSize) + } else { + b.literalBuf = make([]byte, litRegenSize, maxCompressedLiteralSize) + } + } + } + literals = b.literalBuf[:litRegenSize] + v := in[0] + for i := range literals { + literals[i] = v + } + in = in[1:] + if debugDecoder { + printf("Found %d RLE compressed literals\n", litRegenSize) + } + case literalsBlockTreeless: + if len(in) < litCompSize { + println("too small: litType:", litType, " sizeFormat", sizeFormat, "remain:", len(in), "want:", litCompSize) + return in, ErrBlockTooSmall + } + // Store compressed literals, so we defer decoding until we get history. + literals = in[:litCompSize] + in = in[litCompSize:] + if debugDecoder { + printf("Found %d compressed literals\n", litCompSize) + } + huff := hist.huffTree + if huff == nil { + return in, errors.New("literal block was treeless, but no history was defined") + } + // Ensure we have space to store it. + if cap(b.literalBuf) < litRegenSize { + if b.lowMem { + b.literalBuf = make([]byte, 0, litRegenSize) + } else { + b.literalBuf = make([]byte, 0, maxCompressedLiteralSize) + } + } + var err error + // Use our out buffer. + huff.MaxDecodedSize = maxCompressedBlockSize + if fourStreams { + literals, err = huff.Decoder().Decompress4X(b.literalBuf[:0:litRegenSize], literals) + } else { + literals, err = huff.Decoder().Decompress1X(b.literalBuf[:0:litRegenSize], literals) + } + // Make sure we don't leak our literals buffer + if err != nil { + println("decompressing literals:", err) + return in, err + } + if len(literals) != litRegenSize { + return in, fmt.Errorf("literal output size mismatch want %d, got %d", litRegenSize, len(literals)) + } + + case literalsBlockCompressed: + if len(in) < litCompSize { + println("too small: litType:", litType, " sizeFormat", sizeFormat, "remain:", len(in), "want:", litCompSize) + return in, ErrBlockTooSmall + } + literals = in[:litCompSize] + in = in[litCompSize:] + // Ensure we have space to store it. + if cap(b.literalBuf) < litRegenSize { + if b.lowMem { + b.literalBuf = make([]byte, 0, litRegenSize) + } else { + b.literalBuf = make([]byte, 0, maxCompressedBlockSize) + } + } + huff := hist.huffTree + if huff == nil || (hist.dict != nil && huff == hist.dict.litEnc) { + huff = huffDecoderPool.Get().(*huff0.Scratch) + if huff == nil { + huff = &huff0.Scratch{} + } + } + var err error + huff, literals, err = huff0.ReadTable(literals, huff) + if err != nil { + println("reading huffman table:", err) + return in, err + } + hist.huffTree = huff + huff.MaxDecodedSize = maxCompressedBlockSize + // Use our out buffer. + if fourStreams { + literals, err = huff.Decoder().Decompress4X(b.literalBuf[:0:litRegenSize], literals) + } else { + literals, err = huff.Decoder().Decompress1X(b.literalBuf[:0:litRegenSize], literals) + } + if err != nil { + println("decoding compressed literals:", err) + return in, err + } + // Make sure we don't leak our literals buffer + if len(literals) != litRegenSize { + return in, fmt.Errorf("literal output size mismatch want %d, got %d", litRegenSize, len(literals)) + } + if debugDecoder { + printf("Decompressed %d literals into %d bytes\n", litCompSize, litRegenSize) + } + } + hist.decoders.literals = literals + return in, nil +} + +// decodeCompressed will start decompressing a block. +func (b *blockDec) decodeCompressed(hist *history) error { + in := b.data + in, err := b.decodeLiterals(in, hist) + if err != nil { + return err + } + err = b.prepareSequences(in, hist) + if err != nil { + return err + } + if hist.decoders.nSeqs == 0 { + b.dst = append(b.dst, hist.decoders.literals...) + return nil + } + err = hist.decoders.decodeSync(hist) + if err != nil { + return err + } + b.dst = hist.decoders.out + hist.recentOffsets = hist.decoders.prevOffset + return nil +} + +func (b *blockDec) prepareSequences(in []byte, hist *history) (err error) { + if debugDecoder { + printf("prepareSequences: %d byte(s) input\n", len(in)) + } + // Decode Sequences + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#sequences-section + if len(in) < 1 { + return ErrBlockTooSmall + } + var nSeqs int + seqHeader := in[0] + switch { + case seqHeader < 128: + nSeqs = int(seqHeader) + in = in[1:] + case seqHeader < 255: + if len(in) < 2 { + return ErrBlockTooSmall + } + nSeqs = int(seqHeader-128)<<8 | int(in[1]) + in = in[2:] + case seqHeader == 255: + if len(in) < 3 { + return ErrBlockTooSmall + } + nSeqs = 0x7f00 + int(in[1]) + (int(in[2]) << 8) + in = in[3:] + } + if nSeqs == 0 && len(in) != 0 { + // When no sequences, there should not be any more data... + if debugDecoder { + printf("prepareSequences: 0 sequences, but %d byte(s) left on stream\n", len(in)) + } + return ErrUnexpectedBlockSize + } + + var seqs = &hist.decoders + seqs.nSeqs = nSeqs + if nSeqs > 0 { + if len(in) < 1 { + return ErrBlockTooSmall + } + br := byteReader{b: in, off: 0} + compMode := br.Uint8() + br.advance(1) + if debugDecoder { + printf("Compression modes: 0b%b", compMode) + } + for i := uint(0); i < 3; i++ { + mode := seqCompMode((compMode >> (6 - i*2)) & 3) + if debugDecoder { + println("Table", tableIndex(i), "is", mode) + } + var seq *sequenceDec + switch tableIndex(i) { + case tableLiteralLengths: + seq = &seqs.litLengths + case tableOffsets: + seq = &seqs.offsets + case tableMatchLengths: + seq = &seqs.matchLengths + default: + panic("unknown table") + } + switch mode { + case compModePredefined: + if seq.fse != nil && !seq.fse.preDefined { + fseDecoderPool.Put(seq.fse) + } + seq.fse = &fsePredef[i] + case compModeRLE: + if br.remain() < 1 { + return ErrBlockTooSmall + } + v := br.Uint8() + br.advance(1) + if seq.fse == nil || seq.fse.preDefined { + seq.fse = fseDecoderPool.Get().(*fseDecoder) + } + symb, err := decSymbolValue(v, symbolTableX[i]) + if err != nil { + printf("RLE Transform table (%v) error: %v", tableIndex(i), err) + return err + } + seq.fse.setRLE(symb) + if debugDecoder { + printf("RLE set to %+v, code: %v", symb, v) + } + case compModeFSE: + println("Reading table for", tableIndex(i)) + if seq.fse == nil || seq.fse.preDefined { + seq.fse = fseDecoderPool.Get().(*fseDecoder) + } + err := seq.fse.readNCount(&br, uint16(maxTableSymbol[i])) + if err != nil { + println("Read table error:", err) + return err + } + err = seq.fse.transform(symbolTableX[i]) + if err != nil { + println("Transform table error:", err) + return err + } + if debugDecoder { + println("Read table ok", "symbolLen:", seq.fse.symbolLen) + } + case compModeRepeat: + seq.repeat = true + } + if br.overread() { + return io.ErrUnexpectedEOF + } + } + in = br.unread() + } + if debugDecoder { + println("Literals:", len(seqs.literals), "hash:", xxhash.Sum64(seqs.literals), "and", seqs.nSeqs, "sequences.") + } + + if nSeqs == 0 { + if len(b.sequence) > 0 { + b.sequence = b.sequence[:0] + } + return nil + } + br := seqs.br + if br == nil { + br = &bitReader{} + } + if err := br.init(in); err != nil { + return err + } + + if err := seqs.initialize(br, hist, b.dst); err != nil { + println("initializing sequences:", err) + return err + } + return nil +} + +func (b *blockDec) decodeSequences(hist *history) error { + if cap(b.sequence) < hist.decoders.nSeqs { + if b.lowMem { + b.sequence = make([]seqVals, 0, hist.decoders.nSeqs) + } else { + b.sequence = make([]seqVals, 0, 0x7F00+0xffff) + } + } + b.sequence = b.sequence[:hist.decoders.nSeqs] + if hist.decoders.nSeqs == 0 { + hist.decoders.seqSize = len(hist.decoders.literals) + return nil + } + hist.decoders.windowSize = hist.windowSize + hist.decoders.prevOffset = hist.recentOffsets + err := hist.decoders.decode(b.sequence) + hist.recentOffsets = hist.decoders.prevOffset + return err +} + +func (b *blockDec) executeSequences(hist *history) error { + hbytes := hist.b + if len(hbytes) > hist.windowSize { + hbytes = hbytes[len(hbytes)-hist.windowSize:] + // We do not need history anymore. + if hist.dict != nil { + hist.dict.content = nil + } + } + hist.decoders.windowSize = hist.windowSize + hist.decoders.out = b.dst[:0] + err := hist.decoders.execute(b.sequence, hbytes) + if err != nil { + return err + } + return b.updateHistory(hist) +} + +func (b *blockDec) updateHistory(hist *history) error { + if len(b.data) > maxCompressedBlockSize { + return fmt.Errorf("compressed block size too large (%d)", len(b.data)) + } + // Set output and release references. + b.dst = hist.decoders.out + hist.recentOffsets = hist.decoders.prevOffset + + if b.Last { + // if last block we don't care about history. + println("Last block, no history returned") + hist.b = hist.b[:0] + return nil + } else { + hist.append(b.dst) + if debugDecoder { + println("Finished block with ", len(b.sequence), "sequences. Added", len(b.dst), "to history, now length", len(hist.b)) + } + } + hist.decoders.out, hist.decoders.literals = nil, nil + + return nil +} diff --git a/vendor/github.com/klauspost/compress/zstd/blockenc.go b/vendor/github.com/klauspost/compress/zstd/blockenc.go new file mode 100644 index 0000000..12e8f6f --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/blockenc.go @@ -0,0 +1,871 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "errors" + "fmt" + "math" + "math/bits" + + "github.com/klauspost/compress/huff0" +) + +type blockEnc struct { + size int + literals []byte + sequences []seq + coders seqCoders + litEnc *huff0.Scratch + dictLitEnc *huff0.Scratch + wr bitWriter + + extraLits int + output []byte + recentOffsets [3]uint32 + prevRecentOffsets [3]uint32 + + last bool + lowMem bool +} + +// init should be used once the block has been created. +// If called more than once, the effect is the same as calling reset. +func (b *blockEnc) init() { + if b.lowMem { + // 1K literals + if cap(b.literals) < 1<<10 { + b.literals = make([]byte, 0, 1<<10) + } + const defSeqs = 20 + if cap(b.sequences) < defSeqs { + b.sequences = make([]seq, 0, defSeqs) + } + // 1K + if cap(b.output) < 1<<10 { + b.output = make([]byte, 0, 1<<10) + } + } else { + if cap(b.literals) < maxCompressedBlockSize { + b.literals = make([]byte, 0, maxCompressedBlockSize) + } + const defSeqs = 2000 + if cap(b.sequences) < defSeqs { + b.sequences = make([]seq, 0, defSeqs) + } + if cap(b.output) < maxCompressedBlockSize { + b.output = make([]byte, 0, maxCompressedBlockSize) + } + } + + if b.coders.mlEnc == nil { + b.coders.mlEnc = &fseEncoder{} + b.coders.mlPrev = &fseEncoder{} + b.coders.ofEnc = &fseEncoder{} + b.coders.ofPrev = &fseEncoder{} + b.coders.llEnc = &fseEncoder{} + b.coders.llPrev = &fseEncoder{} + } + b.litEnc = &huff0.Scratch{WantLogLess: 4} + b.reset(nil) +} + +// initNewEncode can be used to reset offsets and encoders to the initial state. +func (b *blockEnc) initNewEncode() { + b.recentOffsets = [3]uint32{1, 4, 8} + b.litEnc.Reuse = huff0.ReusePolicyNone + b.coders.setPrev(nil, nil, nil) +} + +// reset will reset the block for a new encode, but in the same stream, +// meaning that state will be carried over, but the block content is reset. +// If a previous block is provided, the recent offsets are carried over. +func (b *blockEnc) reset(prev *blockEnc) { + b.extraLits = 0 + b.literals = b.literals[:0] + b.size = 0 + b.sequences = b.sequences[:0] + b.output = b.output[:0] + b.last = false + if prev != nil { + b.recentOffsets = prev.prevRecentOffsets + } + b.dictLitEnc = nil +} + +// reset will reset the block for a new encode, but in the same stream, +// meaning that state will be carried over, but the block content is reset. +// If a previous block is provided, the recent offsets are carried over. +func (b *blockEnc) swapEncoders(prev *blockEnc) { + b.coders.swap(&prev.coders) + b.litEnc, prev.litEnc = prev.litEnc, b.litEnc +} + +// blockHeader contains the information for a block header. +type blockHeader uint32 + +// setLast sets the 'last' indicator on a block. +func (h *blockHeader) setLast(b bool) { + if b { + *h = *h | 1 + } else { + const mask = (1 << 24) - 2 + *h = *h & mask + } +} + +// setSize will store the compressed size of a block. +func (h *blockHeader) setSize(v uint32) { + const mask = 7 + *h = (*h)&mask | blockHeader(v<<3) +} + +// setType sets the block type. +func (h *blockHeader) setType(t blockType) { + const mask = 1 | (((1 << 24) - 1) ^ 7) + *h = (*h & mask) | blockHeader(t<<1) +} + +// appendTo will append the block header to a slice. +func (h blockHeader) appendTo(b []byte) []byte { + return append(b, uint8(h), uint8(h>>8), uint8(h>>16)) +} + +// String returns a string representation of the block. +func (h blockHeader) String() string { + return fmt.Sprintf("Type: %d, Size: %d, Last:%t", (h>>1)&3, h>>3, h&1 == 1) +} + +// literalsHeader contains literals header information. +type literalsHeader uint64 + +// setType can be used to set the type of literal block. +func (h *literalsHeader) setType(t literalsBlockType) { + const mask = math.MaxUint64 - 3 + *h = (*h & mask) | literalsHeader(t) +} + +// setSize can be used to set a single size, for uncompressed and RLE content. +func (h *literalsHeader) setSize(regenLen int) { + inBits := bits.Len32(uint32(regenLen)) + // Only retain 2 bits + const mask = 3 + lh := uint64(*h & mask) + switch { + case inBits < 5: + lh |= (uint64(regenLen) << 3) | (1 << 60) + if debugEncoder { + got := int(lh>>3) & 0xff + if got != regenLen { + panic(fmt.Sprint("litRegenSize = ", regenLen, "(want) != ", got, "(got)")) + } + } + case inBits < 12: + lh |= (1 << 2) | (uint64(regenLen) << 4) | (2 << 60) + case inBits < 20: + lh |= (3 << 2) | (uint64(regenLen) << 4) | (3 << 60) + default: + panic(fmt.Errorf("internal error: block too big (%d)", regenLen)) + } + *h = literalsHeader(lh) +} + +// setSizes will set the size of a compressed literals section and the input length. +func (h *literalsHeader) setSizes(compLen, inLen int, single bool) { + compBits, inBits := bits.Len32(uint32(compLen)), bits.Len32(uint32(inLen)) + // Only retain 2 bits + const mask = 3 + lh := uint64(*h & mask) + switch { + case compBits <= 10 && inBits <= 10: + if !single { + lh |= 1 << 2 + } + lh |= (uint64(inLen) << 4) | (uint64(compLen) << (10 + 4)) | (3 << 60) + if debugEncoder { + const mmask = (1 << 24) - 1 + n := (lh >> 4) & mmask + if int(n&1023) != inLen { + panic(fmt.Sprint("regensize:", int(n&1023), "!=", inLen, inBits)) + } + if int(n>>10) != compLen { + panic(fmt.Sprint("compsize:", int(n>>10), "!=", compLen, compBits)) + } + } + case compBits <= 14 && inBits <= 14: + lh |= (2 << 2) | (uint64(inLen) << 4) | (uint64(compLen) << (14 + 4)) | (4 << 60) + if single { + panic("single stream used with more than 10 bits length.") + } + case compBits <= 18 && inBits <= 18: + lh |= (3 << 2) | (uint64(inLen) << 4) | (uint64(compLen) << (18 + 4)) | (5 << 60) + if single { + panic("single stream used with more than 10 bits length.") + } + default: + panic("internal error: block too big") + } + *h = literalsHeader(lh) +} + +// appendTo will append the literals header to a byte slice. +func (h literalsHeader) appendTo(b []byte) []byte { + size := uint8(h >> 60) + switch size { + case 1: + b = append(b, uint8(h)) + case 2: + b = append(b, uint8(h), uint8(h>>8)) + case 3: + b = append(b, uint8(h), uint8(h>>8), uint8(h>>16)) + case 4: + b = append(b, uint8(h), uint8(h>>8), uint8(h>>16), uint8(h>>24)) + case 5: + b = append(b, uint8(h), uint8(h>>8), uint8(h>>16), uint8(h>>24), uint8(h>>32)) + default: + panic(fmt.Errorf("internal error: literalsHeader has invalid size (%d)", size)) + } + return b +} + +// size returns the output size with currently set values. +func (h literalsHeader) size() int { + return int(h >> 60) +} + +func (h literalsHeader) String() string { + return fmt.Sprintf("Type: %d, SizeFormat: %d, Size: 0x%d, Bytes:%d", literalsBlockType(h&3), (h>>2)&3, h&((1<<60)-1)>>4, h>>60) +} + +// pushOffsets will push the recent offsets to the backup store. +func (b *blockEnc) pushOffsets() { + b.prevRecentOffsets = b.recentOffsets +} + +// pushOffsets will push the recent offsets to the backup store. +func (b *blockEnc) popOffsets() { + b.recentOffsets = b.prevRecentOffsets +} + +// matchOffset will adjust recent offsets and return the adjusted one, +// if it matches a previous offset. +func (b *blockEnc) matchOffset(offset, lits uint32) uint32 { + // Check if offset is one of the recent offsets. + // Adjusts the output offset accordingly. + // Gives a tiny bit of compression, typically around 1%. + if true { + if lits > 0 { + switch offset { + case b.recentOffsets[0]: + offset = 1 + case b.recentOffsets[1]: + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset = 2 + case b.recentOffsets[2]: + b.recentOffsets[2] = b.recentOffsets[1] + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset = 3 + default: + b.recentOffsets[2] = b.recentOffsets[1] + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset += 3 + } + } else { + switch offset { + case b.recentOffsets[1]: + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset = 1 + case b.recentOffsets[2]: + b.recentOffsets[2] = b.recentOffsets[1] + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset = 2 + case b.recentOffsets[0] - 1: + b.recentOffsets[2] = b.recentOffsets[1] + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset = 3 + default: + b.recentOffsets[2] = b.recentOffsets[1] + b.recentOffsets[1] = b.recentOffsets[0] + b.recentOffsets[0] = offset + offset += 3 + } + } + } else { + offset += 3 + } + return offset +} + +// encodeRaw can be used to set the output to a raw representation of supplied bytes. +func (b *blockEnc) encodeRaw(a []byte) { + var bh blockHeader + bh.setLast(b.last) + bh.setSize(uint32(len(a))) + bh.setType(blockTypeRaw) + b.output = bh.appendTo(b.output[:0]) + b.output = append(b.output, a...) + if debugEncoder { + println("Adding RAW block, length", len(a), "last:", b.last) + } +} + +// encodeRaw can be used to set the output to a raw representation of supplied bytes. +func (b *blockEnc) encodeRawTo(dst, src []byte) []byte { + var bh blockHeader + bh.setLast(b.last) + bh.setSize(uint32(len(src))) + bh.setType(blockTypeRaw) + dst = bh.appendTo(dst) + dst = append(dst, src...) + if debugEncoder { + println("Adding RAW block, length", len(src), "last:", b.last) + } + return dst +} + +// encodeLits can be used if the block is only litLen. +func (b *blockEnc) encodeLits(lits []byte, raw bool) error { + var bh blockHeader + bh.setLast(b.last) + bh.setSize(uint32(len(lits))) + + // Don't compress extremely small blocks + if len(lits) < 8 || (len(lits) < 32 && b.dictLitEnc == nil) || raw { + if debugEncoder { + println("Adding RAW block, length", len(lits), "last:", b.last) + } + bh.setType(blockTypeRaw) + b.output = bh.appendTo(b.output) + b.output = append(b.output, lits...) + return nil + } + + var ( + out []byte + reUsed, single bool + err error + ) + if b.dictLitEnc != nil { + b.litEnc.TransferCTable(b.dictLitEnc) + b.litEnc.Reuse = huff0.ReusePolicyAllow + b.dictLitEnc = nil + } + if len(lits) >= 1024 { + // Use 4 Streams. + out, reUsed, err = huff0.Compress4X(lits, b.litEnc) + } else if len(lits) > 32 { + // Use 1 stream + single = true + out, reUsed, err = huff0.Compress1X(lits, b.litEnc) + } else { + err = huff0.ErrIncompressible + } + + switch err { + case huff0.ErrIncompressible: + if debugEncoder { + println("Adding RAW block, length", len(lits), "last:", b.last) + } + bh.setType(blockTypeRaw) + b.output = bh.appendTo(b.output) + b.output = append(b.output, lits...) + return nil + case huff0.ErrUseRLE: + if debugEncoder { + println("Adding RLE block, length", len(lits)) + } + bh.setType(blockTypeRLE) + b.output = bh.appendTo(b.output) + b.output = append(b.output, lits[0]) + return nil + case nil: + default: + return err + } + // Compressed... + // Now, allow reuse + b.litEnc.Reuse = huff0.ReusePolicyAllow + bh.setType(blockTypeCompressed) + var lh literalsHeader + if reUsed { + if debugEncoder { + println("Reused tree, compressed to", len(out)) + } + lh.setType(literalsBlockTreeless) + } else { + if debugEncoder { + println("New tree, compressed to", len(out), "tree size:", len(b.litEnc.OutTable)) + } + lh.setType(literalsBlockCompressed) + } + // Set sizes + lh.setSizes(len(out), len(lits), single) + bh.setSize(uint32(len(out) + lh.size() + 1)) + + // Write block headers. + b.output = bh.appendTo(b.output) + b.output = lh.appendTo(b.output) + // Add compressed data. + b.output = append(b.output, out...) + // No sequences. + b.output = append(b.output, 0) + return nil +} + +// fuzzFseEncoder can be used to fuzz the FSE encoder. +func fuzzFseEncoder(data []byte) int { + if len(data) > maxSequences || len(data) < 2 { + return 0 + } + enc := fseEncoder{} + hist := enc.Histogram() + maxSym := uint8(0) + for i, v := range data { + v = v & 63 + data[i] = v + hist[v]++ + if v > maxSym { + maxSym = v + } + } + if maxSym == 0 { + // All 0 + return 0 + } + maxCount := func(a []uint32) int { + var max uint32 + for _, v := range a { + if v > max { + max = v + } + } + return int(max) + } + cnt := maxCount(hist[:maxSym]) + if cnt == len(data) { + // RLE + return 0 + } + enc.HistogramFinished(maxSym, cnt) + err := enc.normalizeCount(len(data)) + if err != nil { + return 0 + } + _, err = enc.writeCount(nil) + if err != nil { + panic(err) + } + return 1 +} + +// encode will encode the block and append the output in b.output. +// Previous offset codes must be pushed if more blocks are expected. +func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error { + if len(b.sequences) == 0 { + return b.encodeLits(b.literals, rawAllLits) + } + // We want some difference to at least account for the headers. + saved := b.size - len(b.literals) - (b.size >> 5) + if saved < 16 { + if org == nil { + return errIncompressible + } + b.popOffsets() + return b.encodeLits(org, rawAllLits) + } + + var bh blockHeader + var lh literalsHeader + bh.setLast(b.last) + bh.setType(blockTypeCompressed) + // Store offset of the block header. Needed when we know the size. + bhOffset := len(b.output) + b.output = bh.appendTo(b.output) + + var ( + out []byte + reUsed, single bool + err error + ) + if b.dictLitEnc != nil { + b.litEnc.TransferCTable(b.dictLitEnc) + b.litEnc.Reuse = huff0.ReusePolicyAllow + b.dictLitEnc = nil + } + if len(b.literals) >= 1024 && !raw { + // Use 4 Streams. + out, reUsed, err = huff0.Compress4X(b.literals, b.litEnc) + } else if len(b.literals) > 32 && !raw { + // Use 1 stream + single = true + out, reUsed, err = huff0.Compress1X(b.literals, b.litEnc) + } else { + err = huff0.ErrIncompressible + } + + switch err { + case huff0.ErrIncompressible: + lh.setType(literalsBlockRaw) + lh.setSize(len(b.literals)) + b.output = lh.appendTo(b.output) + b.output = append(b.output, b.literals...) + if debugEncoder { + println("Adding literals RAW, length", len(b.literals)) + } + case huff0.ErrUseRLE: + lh.setType(literalsBlockRLE) + lh.setSize(len(b.literals)) + b.output = lh.appendTo(b.output) + b.output = append(b.output, b.literals[0]) + if debugEncoder { + println("Adding literals RLE") + } + case nil: + // Compressed litLen... + if reUsed { + if debugEncoder { + println("reused tree") + } + lh.setType(literalsBlockTreeless) + } else { + if debugEncoder { + println("new tree, size:", len(b.litEnc.OutTable)) + } + lh.setType(literalsBlockCompressed) + if debugEncoder { + _, _, err := huff0.ReadTable(out, nil) + if err != nil { + panic(err) + } + } + } + lh.setSizes(len(out), len(b.literals), single) + if debugEncoder { + printf("Compressed %d literals to %d bytes", len(b.literals), len(out)) + println("Adding literal header:", lh) + } + b.output = lh.appendTo(b.output) + b.output = append(b.output, out...) + b.litEnc.Reuse = huff0.ReusePolicyAllow + if debugEncoder { + println("Adding literals compressed") + } + default: + if debugEncoder { + println("Adding literals ERROR:", err) + } + return err + } + // Sequence compression + + // Write the number of sequences + switch { + case len(b.sequences) < 128: + b.output = append(b.output, uint8(len(b.sequences))) + case len(b.sequences) < 0x7f00: // TODO: this could be wrong + n := len(b.sequences) + b.output = append(b.output, 128+uint8(n>>8), uint8(n)) + default: + n := len(b.sequences) - 0x7f00 + b.output = append(b.output, 255, uint8(n), uint8(n>>8)) + } + if debugEncoder { + println("Encoding", len(b.sequences), "sequences") + } + b.genCodes() + llEnc := b.coders.llEnc + ofEnc := b.coders.ofEnc + mlEnc := b.coders.mlEnc + err = llEnc.normalizeCount(len(b.sequences)) + if err != nil { + return err + } + err = ofEnc.normalizeCount(len(b.sequences)) + if err != nil { + return err + } + err = mlEnc.normalizeCount(len(b.sequences)) + if err != nil { + return err + } + + // Choose the best compression mode for each type. + // Will evaluate the new vs predefined and previous. + chooseComp := func(cur, prev, preDef *fseEncoder) (*fseEncoder, seqCompMode) { + // See if predefined/previous is better + hist := cur.count[:cur.symbolLen] + nSize := cur.approxSize(hist) + cur.maxHeaderSize() + predefSize := preDef.approxSize(hist) + prevSize := prev.approxSize(hist) + + // Add a small penalty for new encoders. + // Don't bother with extremely small (<2 byte gains). + nSize = nSize + (nSize+2*8*16)>>4 + switch { + case predefSize <= prevSize && predefSize <= nSize || forcePreDef: + if debugEncoder { + println("Using predefined", predefSize>>3, "<=", nSize>>3) + } + return preDef, compModePredefined + case prevSize <= nSize: + if debugEncoder { + println("Using previous", prevSize>>3, "<=", nSize>>3) + } + return prev, compModeRepeat + default: + if debugEncoder { + println("Using new, predef", predefSize>>3, ". previous:", prevSize>>3, ">", nSize>>3, "header max:", cur.maxHeaderSize()>>3, "bytes") + println("tl:", cur.actualTableLog, "symbolLen:", cur.symbolLen, "norm:", cur.norm[:cur.symbolLen], "hist", cur.count[:cur.symbolLen]) + } + return cur, compModeFSE + } + } + + // Write compression mode + var mode uint8 + if llEnc.useRLE { + mode |= uint8(compModeRLE) << 6 + llEnc.setRLE(b.sequences[0].llCode) + if debugEncoder { + println("llEnc.useRLE") + } + } else { + var m seqCompMode + llEnc, m = chooseComp(llEnc, b.coders.llPrev, &fsePredefEnc[tableLiteralLengths]) + mode |= uint8(m) << 6 + } + if ofEnc.useRLE { + mode |= uint8(compModeRLE) << 4 + ofEnc.setRLE(b.sequences[0].ofCode) + if debugEncoder { + println("ofEnc.useRLE") + } + } else { + var m seqCompMode + ofEnc, m = chooseComp(ofEnc, b.coders.ofPrev, &fsePredefEnc[tableOffsets]) + mode |= uint8(m) << 4 + } + + if mlEnc.useRLE { + mode |= uint8(compModeRLE) << 2 + mlEnc.setRLE(b.sequences[0].mlCode) + if debugEncoder { + println("mlEnc.useRLE, code: ", b.sequences[0].mlCode, "value", b.sequences[0].matchLen) + } + } else { + var m seqCompMode + mlEnc, m = chooseComp(mlEnc, b.coders.mlPrev, &fsePredefEnc[tableMatchLengths]) + mode |= uint8(m) << 2 + } + b.output = append(b.output, mode) + if debugEncoder { + printf("Compression modes: 0b%b", mode) + } + b.output, err = llEnc.writeCount(b.output) + if err != nil { + return err + } + start := len(b.output) + b.output, err = ofEnc.writeCount(b.output) + if err != nil { + return err + } + if false { + println("block:", b.output[start:], "tablelog", ofEnc.actualTableLog, "maxcount:", ofEnc.maxCount) + fmt.Printf("selected TableLog: %d, Symbol length: %d\n", ofEnc.actualTableLog, ofEnc.symbolLen) + for i, v := range ofEnc.norm[:ofEnc.symbolLen] { + fmt.Printf("%3d: %5d -> %4d \n", i, ofEnc.count[i], v) + } + } + b.output, err = mlEnc.writeCount(b.output) + if err != nil { + return err + } + + // Maybe in block? + wr := &b.wr + wr.reset(b.output) + + var ll, of, ml cState + + // Current sequence + seq := len(b.sequences) - 1 + s := b.sequences[seq] + llEnc.setBits(llBitsTable[:]) + mlEnc.setBits(mlBitsTable[:]) + ofEnc.setBits(nil) + + llTT, ofTT, mlTT := llEnc.ct.symbolTT[:256], ofEnc.ct.symbolTT[:256], mlEnc.ct.symbolTT[:256] + + // We have 3 bounds checks here (and in the loop). + // Since we are iterating backwards it is kinda hard to avoid. + llB, ofB, mlB := llTT[s.llCode], ofTT[s.ofCode], mlTT[s.mlCode] + ll.init(wr, &llEnc.ct, llB) + of.init(wr, &ofEnc.ct, ofB) + wr.flush32() + ml.init(wr, &mlEnc.ct, mlB) + + // Each of these lookups also generates a bounds check. + wr.addBits32NC(s.litLen, llB.outBits) + wr.addBits32NC(s.matchLen, mlB.outBits) + wr.flush32() + wr.addBits32NC(s.offset, ofB.outBits) + if debugSequences { + println("Encoded seq", seq, s, "codes:", s.llCode, s.mlCode, s.ofCode, "states:", ll.state, ml.state, of.state, "bits:", llB, mlB, ofB) + } + seq-- + // Store sequences in reverse... + for seq >= 0 { + s = b.sequences[seq] + + ofB := ofTT[s.ofCode] + wr.flush32() // tablelog max is below 8 for each, so it will fill max 24 bits. + //of.encode(ofB) + nbBitsOut := (uint32(of.state) + ofB.deltaNbBits) >> 16 + dstState := int32(of.state>>(nbBitsOut&15)) + int32(ofB.deltaFindState) + wr.addBits16NC(of.state, uint8(nbBitsOut)) + of.state = of.stateTable[dstState] + + // Accumulate extra bits. + outBits := ofB.outBits & 31 + extraBits := uint64(s.offset & bitMask32[outBits]) + extraBitsN := outBits + + mlB := mlTT[s.mlCode] + //ml.encode(mlB) + nbBitsOut = (uint32(ml.state) + mlB.deltaNbBits) >> 16 + dstState = int32(ml.state>>(nbBitsOut&15)) + int32(mlB.deltaFindState) + wr.addBits16NC(ml.state, uint8(nbBitsOut)) + ml.state = ml.stateTable[dstState] + + outBits = mlB.outBits & 31 + extraBits = extraBits<> 16 + dstState = int32(ll.state>>(nbBitsOut&15)) + int32(llB.deltaFindState) + wr.addBits16NC(ll.state, uint8(nbBitsOut)) + ll.state = ll.stateTable[dstState] + + outBits = llB.outBits & 31 + extraBits = extraBits<= b.size { + // Maybe even add a bigger margin. + b.litEnc.Reuse = huff0.ReusePolicyNone + return errIncompressible + } + + // Size is output minus block header. + bh.setSize(uint32(len(b.output)-bhOffset) - 3) + if debugEncoder { + println("Rewriting block header", bh) + } + _ = bh.appendTo(b.output[bhOffset:bhOffset]) + b.coders.setPrev(llEnc, mlEnc, ofEnc) + return nil +} + +var errIncompressible = errors.New("incompressible") + +func (b *blockEnc) genCodes() { + if len(b.sequences) == 0 { + // nothing to do + return + } + if len(b.sequences) > math.MaxUint16 { + panic("can only encode up to 64K sequences") + } + // No bounds checks after here: + llH := b.coders.llEnc.Histogram() + ofH := b.coders.ofEnc.Histogram() + mlH := b.coders.mlEnc.Histogram() + for i := range llH { + llH[i] = 0 + } + for i := range ofH { + ofH[i] = 0 + } + for i := range mlH { + mlH[i] = 0 + } + + var llMax, ofMax, mlMax uint8 + for i := range b.sequences { + seq := &b.sequences[i] + v := llCode(seq.litLen) + seq.llCode = v + llH[v]++ + if v > llMax { + llMax = v + } + + v = ofCode(seq.offset) + seq.ofCode = v + ofH[v]++ + if v > ofMax { + ofMax = v + } + + v = mlCode(seq.matchLen) + seq.mlCode = v + mlH[v]++ + if v > mlMax { + mlMax = v + if debugAsserts && mlMax > maxMatchLengthSymbol { + panic(fmt.Errorf("mlMax > maxMatchLengthSymbol (%d), matchlen: %d", mlMax, seq.matchLen)) + } + } + } + maxCount := func(a []uint32) int { + var max uint32 + for _, v := range a { + if v > max { + max = v + } + } + return int(max) + } + if debugAsserts && mlMax > maxMatchLengthSymbol { + panic(fmt.Errorf("mlMax > maxMatchLengthSymbol (%d)", mlMax)) + } + if debugAsserts && ofMax > maxOffsetBits { + panic(fmt.Errorf("ofMax > maxOffsetBits (%d)", ofMax)) + } + if debugAsserts && llMax > maxLiteralLengthSymbol { + panic(fmt.Errorf("llMax > maxLiteralLengthSymbol (%d)", llMax)) + } + + b.coders.mlEnc.HistogramFinished(mlMax, maxCount(mlH[:mlMax+1])) + b.coders.ofEnc.HistogramFinished(ofMax, maxCount(ofH[:ofMax+1])) + b.coders.llEnc.HistogramFinished(llMax, maxCount(llH[:llMax+1])) +} diff --git a/vendor/github.com/klauspost/compress/zstd/blocktype_string.go b/vendor/github.com/klauspost/compress/zstd/blocktype_string.go new file mode 100644 index 0000000..01a01e4 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/blocktype_string.go @@ -0,0 +1,85 @@ +// Code generated by "stringer -type=blockType,literalsBlockType,seqCompMode,tableIndex"; DO NOT EDIT. + +package zstd + +import "strconv" + +func _() { + // An "invalid array index" compiler error signifies that the constant values have changed. + // Re-run the stringer command to generate them again. + var x [1]struct{} + _ = x[blockTypeRaw-0] + _ = x[blockTypeRLE-1] + _ = x[blockTypeCompressed-2] + _ = x[blockTypeReserved-3] +} + +const _blockType_name = "blockTypeRawblockTypeRLEblockTypeCompressedblockTypeReserved" + +var _blockType_index = [...]uint8{0, 12, 24, 43, 60} + +func (i blockType) String() string { + if i >= blockType(len(_blockType_index)-1) { + return "blockType(" + strconv.FormatInt(int64(i), 10) + ")" + } + return _blockType_name[_blockType_index[i]:_blockType_index[i+1]] +} +func _() { + // An "invalid array index" compiler error signifies that the constant values have changed. + // Re-run the stringer command to generate them again. + var x [1]struct{} + _ = x[literalsBlockRaw-0] + _ = x[literalsBlockRLE-1] + _ = x[literalsBlockCompressed-2] + _ = x[literalsBlockTreeless-3] +} + +const _literalsBlockType_name = "literalsBlockRawliteralsBlockRLEliteralsBlockCompressedliteralsBlockTreeless" + +var _literalsBlockType_index = [...]uint8{0, 16, 32, 55, 76} + +func (i literalsBlockType) String() string { + if i >= literalsBlockType(len(_literalsBlockType_index)-1) { + return "literalsBlockType(" + strconv.FormatInt(int64(i), 10) + ")" + } + return _literalsBlockType_name[_literalsBlockType_index[i]:_literalsBlockType_index[i+1]] +} +func _() { + // An "invalid array index" compiler error signifies that the constant values have changed. + // Re-run the stringer command to generate them again. + var x [1]struct{} + _ = x[compModePredefined-0] + _ = x[compModeRLE-1] + _ = x[compModeFSE-2] + _ = x[compModeRepeat-3] +} + +const _seqCompMode_name = "compModePredefinedcompModeRLEcompModeFSEcompModeRepeat" + +var _seqCompMode_index = [...]uint8{0, 18, 29, 40, 54} + +func (i seqCompMode) String() string { + if i >= seqCompMode(len(_seqCompMode_index)-1) { + return "seqCompMode(" + strconv.FormatInt(int64(i), 10) + ")" + } + return _seqCompMode_name[_seqCompMode_index[i]:_seqCompMode_index[i+1]] +} +func _() { + // An "invalid array index" compiler error signifies that the constant values have changed. + // Re-run the stringer command to generate them again. + var x [1]struct{} + _ = x[tableLiteralLengths-0] + _ = x[tableOffsets-1] + _ = x[tableMatchLengths-2] +} + +const _tableIndex_name = "tableLiteralLengthstableOffsetstableMatchLengths" + +var _tableIndex_index = [...]uint8{0, 19, 31, 48} + +func (i tableIndex) String() string { + if i >= tableIndex(len(_tableIndex_index)-1) { + return "tableIndex(" + strconv.FormatInt(int64(i), 10) + ")" + } + return _tableIndex_name[_tableIndex_index[i]:_tableIndex_index[i+1]] +} diff --git a/vendor/github.com/klauspost/compress/zstd/bytebuf.go b/vendor/github.com/klauspost/compress/zstd/bytebuf.go new file mode 100644 index 0000000..b80191e --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/bytebuf.go @@ -0,0 +1,133 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "fmt" + "io" + "io/ioutil" +) + +type byteBuffer interface { + // Read up to 8 bytes. + // Returns io.ErrUnexpectedEOF if this cannot be satisfied. + readSmall(n int) ([]byte, error) + + // Read >8 bytes. + // MAY use the destination slice. + readBig(n int, dst []byte) ([]byte, error) + + // Read a single byte. + readByte() (byte, error) + + // Skip n bytes. + skipN(n int) error +} + +// in-memory buffer +type byteBuf []byte + +func (b *byteBuf) readSmall(n int) ([]byte, error) { + if debugAsserts && n > 8 { + panic(fmt.Errorf("small read > 8 (%d). use readBig", n)) + } + bb := *b + if len(bb) < n { + return nil, io.ErrUnexpectedEOF + } + r := bb[:n] + *b = bb[n:] + return r, nil +} + +func (b *byteBuf) readBig(n int, dst []byte) ([]byte, error) { + bb := *b + if len(bb) < n { + return nil, io.ErrUnexpectedEOF + } + r := bb[:n] + *b = bb[n:] + return r, nil +} + +func (b *byteBuf) remain() []byte { + return *b +} + +func (b *byteBuf) readByte() (byte, error) { + bb := *b + if len(bb) < 1 { + return 0, nil + } + r := bb[0] + *b = bb[1:] + return r, nil +} + +func (b *byteBuf) skipN(n int) error { + bb := *b + if len(bb) < n { + return io.ErrUnexpectedEOF + } + *b = bb[n:] + return nil +} + +// wrapper around a reader. +type readerWrapper struct { + r io.Reader + tmp [8]byte +} + +func (r *readerWrapper) readSmall(n int) ([]byte, error) { + if debugAsserts && n > 8 { + panic(fmt.Errorf("small read > 8 (%d). use readBig", n)) + } + n2, err := io.ReadFull(r.r, r.tmp[:n]) + // We only really care about the actual bytes read. + if err != nil { + if err == io.EOF { + return nil, io.ErrUnexpectedEOF + } + if debugDecoder { + println("readSmall: got", n2, "want", n, "err", err) + } + return nil, err + } + return r.tmp[:n], nil +} + +func (r *readerWrapper) readBig(n int, dst []byte) ([]byte, error) { + if cap(dst) < n { + dst = make([]byte, n) + } + n2, err := io.ReadFull(r.r, dst[:n]) + if err == io.EOF && n > 0 { + err = io.ErrUnexpectedEOF + } + return dst[:n2], err +} + +func (r *readerWrapper) readByte() (byte, error) { + n2, err := r.r.Read(r.tmp[:1]) + if err != nil { + if err == io.EOF { + err = io.ErrUnexpectedEOF + } + return 0, err + } + if n2 != 1 { + return 0, io.ErrUnexpectedEOF + } + return r.tmp[0], nil +} + +func (r *readerWrapper) skipN(n int) error { + n2, err := io.CopyN(ioutil.Discard, r.r, int64(n)) + if n2 != int64(n) { + err = io.ErrUnexpectedEOF + } + return err +} diff --git a/vendor/github.com/klauspost/compress/zstd/bytereader.go b/vendor/github.com/klauspost/compress/zstd/bytereader.go new file mode 100644 index 0000000..2c4fca1 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/bytereader.go @@ -0,0 +1,88 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +// byteReader provides a byte reader that reads +// little endian values from a byte stream. +// The input stream is manually advanced. +// The reader performs no bounds checks. +type byteReader struct { + b []byte + off int +} + +// init will initialize the reader and set the input. +func (b *byteReader) init(in []byte) { + b.b = in + b.off = 0 +} + +// advance the stream b n bytes. +func (b *byteReader) advance(n uint) { + b.off += int(n) +} + +// overread returns whether we have advanced too far. +func (b *byteReader) overread() bool { + return b.off > len(b.b) +} + +// Int32 returns a little endian int32 starting at current offset. +func (b byteReader) Int32() int32 { + b2 := b.b[b.off:] + b2 = b2[:4] + v3 := int32(b2[3]) + v2 := int32(b2[2]) + v1 := int32(b2[1]) + v0 := int32(b2[0]) + return v0 | (v1 << 8) | (v2 << 16) | (v3 << 24) +} + +// Uint8 returns the next byte +func (b *byteReader) Uint8() uint8 { + v := b.b[b.off] + return v +} + +// Uint32 returns a little endian uint32 starting at current offset. +func (b byteReader) Uint32() uint32 { + if r := b.remain(); r < 4 { + // Very rare + v := uint32(0) + for i := 1; i <= r; i++ { + v = (v << 8) | uint32(b.b[len(b.b)-i]) + } + return v + } + b2 := b.b[b.off:] + b2 = b2[:4] + v3 := uint32(b2[3]) + v2 := uint32(b2[2]) + v1 := uint32(b2[1]) + v0 := uint32(b2[0]) + return v0 | (v1 << 8) | (v2 << 16) | (v3 << 24) +} + +// Uint32NC returns a little endian uint32 starting at current offset. +// The caller must be sure if there are at least 4 bytes left. +func (b byteReader) Uint32NC() uint32 { + b2 := b.b[b.off:] + b2 = b2[:4] + v3 := uint32(b2[3]) + v2 := uint32(b2[2]) + v1 := uint32(b2[1]) + v0 := uint32(b2[0]) + return v0 | (v1 << 8) | (v2 << 16) | (v3 << 24) +} + +// unread returns the unread portion of the input. +func (b byteReader) unread() []byte { + return b.b[b.off:] +} + +// remain will return the number of bytes remaining. +func (b byteReader) remain() int { + return len(b.b) - b.off +} diff --git a/vendor/github.com/klauspost/compress/zstd/decodeheader.go b/vendor/github.com/klauspost/compress/zstd/decodeheader.go new file mode 100644 index 0000000..5022e71 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/decodeheader.go @@ -0,0 +1,230 @@ +// Copyright 2020+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. + +package zstd + +import ( + "bytes" + "encoding/binary" + "errors" + "io" +) + +// HeaderMaxSize is the maximum size of a Frame and Block Header. +// If less is sent to Header.Decode it *may* still contain enough information. +const HeaderMaxSize = 14 + 3 + +// Header contains information about the first frame and block within that. +type Header struct { + // SingleSegment specifies whether the data is to be decompressed into a + // single contiguous memory segment. + // It implies that WindowSize is invalid and that FrameContentSize is valid. + SingleSegment bool + + // WindowSize is the window of data to keep while decoding. + // Will only be set if SingleSegment is false. + WindowSize uint64 + + // Dictionary ID. + // If 0, no dictionary. + DictionaryID uint32 + + // HasFCS specifies whether FrameContentSize has a valid value. + HasFCS bool + + // FrameContentSize is the expected uncompressed size of the entire frame. + FrameContentSize uint64 + + // Skippable will be true if the frame is meant to be skipped. + // This implies that FirstBlock.OK is false. + Skippable bool + + // SkippableID is the user-specific ID for the skippable frame. + // Valid values are between 0 to 15, inclusive. + SkippableID int + + // SkippableSize is the length of the user data to skip following + // the header. + SkippableSize uint32 + + // HeaderSize is the raw size of the frame header. + // + // For normal frames, it includes the size of the magic number and + // the size of the header (per section 3.1.1.1). + // It does not include the size for any data blocks (section 3.1.1.2) nor + // the size for the trailing content checksum. + // + // For skippable frames, this counts the size of the magic number + // along with the size of the size field of the payload. + // It does not include the size of the skippable payload itself. + // The total frame size is the HeaderSize plus the SkippableSize. + HeaderSize int + + // First block information. + FirstBlock struct { + // OK will be set if first block could be decoded. + OK bool + + // Is this the last block of a frame? + Last bool + + // Is the data compressed? + // If true CompressedSize will be populated. + // Unfortunately DecompressedSize cannot be determined + // without decoding the blocks. + Compressed bool + + // DecompressedSize is the expected decompressed size of the block. + // Will be 0 if it cannot be determined. + DecompressedSize int + + // CompressedSize of the data in the block. + // Does not include the block header. + // Will be equal to DecompressedSize if not Compressed. + CompressedSize int + } + + // If set there is a checksum present for the block content. + // The checksum field at the end is always 4 bytes long. + HasCheckSum bool +} + +// Decode the header from the beginning of the stream. +// This will decode the frame header and the first block header if enough bytes are provided. +// It is recommended to provide at least HeaderMaxSize bytes. +// If the frame header cannot be read an error will be returned. +// If there isn't enough input, io.ErrUnexpectedEOF is returned. +// The FirstBlock.OK will indicate if enough information was available to decode the first block header. +func (h *Header) Decode(in []byte) error { + *h = Header{} + if len(in) < 4 { + return io.ErrUnexpectedEOF + } + h.HeaderSize += 4 + b, in := in[:4], in[4:] + if !bytes.Equal(b, frameMagic) { + if !bytes.Equal(b[1:4], skippableFrameMagic) || b[0]&0xf0 != 0x50 { + return ErrMagicMismatch + } + if len(in) < 4 { + return io.ErrUnexpectedEOF + } + h.HeaderSize += 4 + h.Skippable = true + h.SkippableID = int(b[0] & 0xf) + h.SkippableSize = binary.LittleEndian.Uint32(in) + return nil + } + + // Read Window_Descriptor + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#window_descriptor + if len(in) < 1 { + return io.ErrUnexpectedEOF + } + fhd, in := in[0], in[1:] + h.HeaderSize++ + h.SingleSegment = fhd&(1<<5) != 0 + h.HasCheckSum = fhd&(1<<2) != 0 + if fhd&(1<<3) != 0 { + return errors.New("reserved bit set on frame header") + } + + if !h.SingleSegment { + if len(in) < 1 { + return io.ErrUnexpectedEOF + } + var wd byte + wd, in = in[0], in[1:] + h.HeaderSize++ + windowLog := 10 + (wd >> 3) + windowBase := uint64(1) << windowLog + windowAdd := (windowBase / 8) * uint64(wd&0x7) + h.WindowSize = windowBase + windowAdd + } + + // Read Dictionary_ID + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#dictionary_id + if size := fhd & 3; size != 0 { + if size == 3 { + size = 4 + } + if len(in) < int(size) { + return io.ErrUnexpectedEOF + } + b, in = in[:size], in[size:] + h.HeaderSize += int(size) + switch size { + case 1: + h.DictionaryID = uint32(b[0]) + case 2: + h.DictionaryID = uint32(b[0]) | (uint32(b[1]) << 8) + case 4: + h.DictionaryID = uint32(b[0]) | (uint32(b[1]) << 8) | (uint32(b[2]) << 16) | (uint32(b[3]) << 24) + } + } + + // Read Frame_Content_Size + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#frame_content_size + var fcsSize int + v := fhd >> 6 + switch v { + case 0: + if h.SingleSegment { + fcsSize = 1 + } + default: + fcsSize = 1 << v + } + + if fcsSize > 0 { + h.HasFCS = true + if len(in) < fcsSize { + return io.ErrUnexpectedEOF + } + b, in = in[:fcsSize], in[fcsSize:] + h.HeaderSize += int(fcsSize) + switch fcsSize { + case 1: + h.FrameContentSize = uint64(b[0]) + case 2: + // When FCS_Field_Size is 2, the offset of 256 is added. + h.FrameContentSize = uint64(b[0]) | (uint64(b[1]) << 8) + 256 + case 4: + h.FrameContentSize = uint64(b[0]) | (uint64(b[1]) << 8) | (uint64(b[2]) << 16) | (uint64(b[3]) << 24) + case 8: + d1 := uint32(b[0]) | (uint32(b[1]) << 8) | (uint32(b[2]) << 16) | (uint32(b[3]) << 24) + d2 := uint32(b[4]) | (uint32(b[5]) << 8) | (uint32(b[6]) << 16) | (uint32(b[7]) << 24) + h.FrameContentSize = uint64(d1) | (uint64(d2) << 32) + } + } + + // Frame Header done, we will not fail from now on. + if len(in) < 3 { + return nil + } + tmp := in[:3] + bh := uint32(tmp[0]) | (uint32(tmp[1]) << 8) | (uint32(tmp[2]) << 16) + h.FirstBlock.Last = bh&1 != 0 + blockType := blockType((bh >> 1) & 3) + // find size. + cSize := int(bh >> 3) + switch blockType { + case blockTypeReserved: + return nil + case blockTypeRLE: + h.FirstBlock.Compressed = true + h.FirstBlock.DecompressedSize = cSize + h.FirstBlock.CompressedSize = 1 + case blockTypeCompressed: + h.FirstBlock.Compressed = true + h.FirstBlock.CompressedSize = cSize + case blockTypeRaw: + h.FirstBlock.DecompressedSize = cSize + h.FirstBlock.CompressedSize = cSize + default: + panic("Invalid block type") + } + + h.FirstBlock.OK = true + return nil +} diff --git a/vendor/github.com/klauspost/compress/zstd/decoder.go b/vendor/github.com/klauspost/compress/zstd/decoder.go new file mode 100644 index 0000000..9fcdaac --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/decoder.go @@ -0,0 +1,924 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "bytes" + "context" + "encoding/binary" + "io" + "sync" + + "github.com/klauspost/compress/zstd/internal/xxhash" +) + +// Decoder provides decoding of zstandard streams. +// The decoder has been designed to operate without allocations after a warmup. +// This means that you should store the decoder for best performance. +// To re-use a stream decoder, use the Reset(r io.Reader) error to switch to another stream. +// A decoder can safely be re-used even if the previous stream failed. +// To release the resources, you must call the Close() function on a decoder. +type Decoder struct { + o decoderOptions + + // Unreferenced decoders, ready for use. + decoders chan *blockDec + + // Current read position used for Reader functionality. + current decoderState + + // sync stream decoding + syncStream struct { + decodedFrame uint64 + br readerWrapper + enabled bool + inFrame bool + } + + frame *frameDec + + // Custom dictionaries. + // Always uses copies. + dicts map[uint32]dict + + // streamWg is the waitgroup for all streams + streamWg sync.WaitGroup +} + +// decoderState is used for maintaining state when the decoder +// is used for streaming. +type decoderState struct { + // current block being written to stream. + decodeOutput + + // output in order to be written to stream. + output chan decodeOutput + + // cancel remaining output. + cancel context.CancelFunc + + // crc of current frame + crc *xxhash.Digest + + flushed bool +} + +var ( + // Check the interfaces we want to support. + _ = io.WriterTo(&Decoder{}) + _ = io.Reader(&Decoder{}) +) + +// NewReader creates a new decoder. +// A nil Reader can be provided in which case Reset can be used to start a decode. +// +// A Decoder can be used in two modes: +// +// 1) As a stream, or +// 2) For stateless decoding using DecodeAll. +// +// Only a single stream can be decoded concurrently, but the same decoder +// can run multiple concurrent stateless decodes. It is even possible to +// use stateless decodes while a stream is being decoded. +// +// The Reset function can be used to initiate a new stream, which is will considerably +// reduce the allocations normally caused by NewReader. +func NewReader(r io.Reader, opts ...DOption) (*Decoder, error) { + initPredefined() + var d Decoder + d.o.setDefault() + for _, o := range opts { + err := o(&d.o) + if err != nil { + return nil, err + } + } + d.current.crc = xxhash.New() + d.current.flushed = true + + if r == nil { + d.current.err = ErrDecoderNilInput + } + + // Transfer option dicts. + d.dicts = make(map[uint32]dict, len(d.o.dicts)) + for _, dc := range d.o.dicts { + d.dicts[dc.id] = dc + } + d.o.dicts = nil + + // Create decoders + d.decoders = make(chan *blockDec, d.o.concurrent) + for i := 0; i < d.o.concurrent; i++ { + dec := newBlockDec(d.o.lowMem) + dec.localFrame = newFrameDec(d.o) + d.decoders <- dec + } + + if r == nil { + return &d, nil + } + return &d, d.Reset(r) +} + +// Read bytes from the decompressed stream into p. +// Returns the number of bytes written and any error that occurred. +// When the stream is done, io.EOF will be returned. +func (d *Decoder) Read(p []byte) (int, error) { + var n int + for { + if len(d.current.b) > 0 { + filled := copy(p, d.current.b) + p = p[filled:] + d.current.b = d.current.b[filled:] + n += filled + } + if len(p) == 0 { + break + } + if len(d.current.b) == 0 { + // We have an error and no more data + if d.current.err != nil { + break + } + if !d.nextBlock(n == 0) { + return n, d.current.err + } + } + } + if len(d.current.b) > 0 { + if debugDecoder { + println("returning", n, "still bytes left:", len(d.current.b)) + } + // Only return error at end of block + return n, nil + } + if d.current.err != nil { + d.drainOutput() + } + if debugDecoder { + println("returning", n, d.current.err, len(d.decoders)) + } + return n, d.current.err +} + +// Reset will reset the decoder the supplied stream after the current has finished processing. +// Note that this functionality cannot be used after Close has been called. +// Reset can be called with a nil reader to release references to the previous reader. +// After being called with a nil reader, no other operations than Reset or DecodeAll or Close +// should be used. +func (d *Decoder) Reset(r io.Reader) error { + if d.current.err == ErrDecoderClosed { + return d.current.err + } + + d.drainOutput() + + d.syncStream.br.r = nil + if r == nil { + d.current.err = ErrDecoderNilInput + if len(d.current.b) > 0 { + d.current.b = d.current.b[:0] + } + d.current.flushed = true + return nil + } + + // If bytes buffer and < 5MB, do sync decoding anyway. + if bb, ok := r.(byter); ok && bb.Len() < 5<<20 { + bb2 := bb + if debugDecoder { + println("*bytes.Buffer detected, doing sync decode, len:", bb.Len()) + } + b := bb2.Bytes() + var dst []byte + if cap(d.current.b) > 0 { + dst = d.current.b + } + + dst, err := d.DecodeAll(b, dst[:0]) + if err == nil { + err = io.EOF + } + d.current.b = dst + d.current.err = err + d.current.flushed = true + if debugDecoder { + println("sync decode to", len(dst), "bytes, err:", err) + } + return nil + } + // Remove current block. + d.stashDecoder() + d.current.decodeOutput = decodeOutput{} + d.current.err = nil + d.current.flushed = false + d.current.d = nil + + // Ensure no-one else is still running... + d.streamWg.Wait() + if d.frame == nil { + d.frame = newFrameDec(d.o) + } + + if d.o.concurrent == 1 { + return d.startSyncDecoder(r) + } + + d.current.output = make(chan decodeOutput, d.o.concurrent) + ctx, cancel := context.WithCancel(context.Background()) + d.current.cancel = cancel + d.streamWg.Add(1) + go d.startStreamDecoder(ctx, r, d.current.output) + + return nil +} + +// drainOutput will drain the output until errEndOfStream is sent. +func (d *Decoder) drainOutput() { + if d.current.cancel != nil { + if debugDecoder { + println("cancelling current") + } + d.current.cancel() + d.current.cancel = nil + } + if d.current.d != nil { + if debugDecoder { + printf("re-adding current decoder %p, decoders: %d", d.current.d, len(d.decoders)) + } + d.decoders <- d.current.d + d.current.d = nil + d.current.b = nil + } + if d.current.output == nil || d.current.flushed { + println("current already flushed") + return + } + for v := range d.current.output { + if v.d != nil { + if debugDecoder { + printf("re-adding decoder %p", v.d) + } + d.decoders <- v.d + } + } + d.current.output = nil + d.current.flushed = true +} + +// WriteTo writes data to w until there's no more data to write or when an error occurs. +// The return value n is the number of bytes written. +// Any error encountered during the write is also returned. +func (d *Decoder) WriteTo(w io.Writer) (int64, error) { + var n int64 + for { + if len(d.current.b) > 0 { + n2, err2 := w.Write(d.current.b) + n += int64(n2) + if err2 != nil && (d.current.err == nil || d.current.err == io.EOF) { + d.current.err = err2 + } else if n2 != len(d.current.b) { + d.current.err = io.ErrShortWrite + } + } + if d.current.err != nil { + break + } + d.nextBlock(true) + } + err := d.current.err + if err != nil { + d.drainOutput() + } + if err == io.EOF { + err = nil + } + return n, err +} + +// DecodeAll allows stateless decoding of a blob of bytes. +// Output will be appended to dst, so if the destination size is known +// you can pre-allocate the destination slice to avoid allocations. +// DecodeAll can be used concurrently. +// The Decoder concurrency limits will be respected. +func (d *Decoder) DecodeAll(input, dst []byte) ([]byte, error) { + if d.decoders == nil { + return dst, ErrDecoderClosed + } + + // Grab a block decoder and frame decoder. + block := <-d.decoders + frame := block.localFrame + defer func() { + if debugDecoder { + printf("re-adding decoder: %p", block) + } + frame.rawInput = nil + frame.bBuf = nil + if frame.history.decoders.br != nil { + frame.history.decoders.br.in = nil + } + d.decoders <- block + }() + frame.bBuf = input + + for { + frame.history.reset() + err := frame.reset(&frame.bBuf) + if err != nil { + if err == io.EOF { + if debugDecoder { + println("frame reset return EOF") + } + return dst, nil + } + return dst, err + } + if frame.DictionaryID != nil { + dict, ok := d.dicts[*frame.DictionaryID] + if !ok { + return nil, ErrUnknownDictionary + } + if debugDecoder { + println("setting dict", frame.DictionaryID) + } + frame.history.setDict(&dict) + } + + if frame.FrameContentSize != fcsUnknown && frame.FrameContentSize > d.o.maxDecodedSize-uint64(len(dst)) { + return dst, ErrDecoderSizeExceeded + } + if frame.FrameContentSize < 1<<30 { + // Never preallocate more than 1 GB up front. + if cap(dst)-len(dst) < int(frame.FrameContentSize) { + dst2 := make([]byte, len(dst), len(dst)+int(frame.FrameContentSize)) + copy(dst2, dst) + dst = dst2 + } + } + if cap(dst) == 0 { + // Allocate len(input) * 2 by default if nothing is provided + // and we didn't get frame content size. + size := len(input) * 2 + // Cap to 1 MB. + if size > 1<<20 { + size = 1 << 20 + } + if uint64(size) > d.o.maxDecodedSize { + size = int(d.o.maxDecodedSize) + } + dst = make([]byte, 0, size) + } + + dst, err = frame.runDecoder(dst, block) + if err != nil { + return dst, err + } + if len(frame.bBuf) == 0 { + if debugDecoder { + println("frame dbuf empty") + } + break + } + } + return dst, nil +} + +// nextBlock returns the next block. +// If an error occurs d.err will be set. +// Optionally the function can block for new output. +// If non-blocking mode is used the returned boolean will be false +// if no data was available without blocking. +func (d *Decoder) nextBlock(blocking bool) (ok bool) { + if d.current.err != nil { + // Keep error state. + return false + } + d.current.b = d.current.b[:0] + + // SYNC: + if d.syncStream.enabled { + if !blocking { + return false + } + ok = d.nextBlockSync() + if !ok { + d.stashDecoder() + } + return ok + } + + //ASYNC: + d.stashDecoder() + if blocking { + d.current.decodeOutput, ok = <-d.current.output + } else { + select { + case d.current.decodeOutput, ok = <-d.current.output: + default: + return false + } + } + if !ok { + // This should not happen, so signal error state... + d.current.err = io.ErrUnexpectedEOF + return false + } + next := d.current.decodeOutput + if next.d != nil && next.d.async.newHist != nil { + d.current.crc.Reset() + } + if debugDecoder { + var tmp [4]byte + binary.LittleEndian.PutUint32(tmp[:], uint32(xxhash.Sum64(next.b))) + println("got", len(d.current.b), "bytes, error:", d.current.err, "data crc:", tmp) + } + + if len(next.b) > 0 { + n, err := d.current.crc.Write(next.b) + if err == nil { + if n != len(next.b) { + d.current.err = io.ErrShortWrite + } + } + } + if next.err == nil && next.d != nil && len(next.d.checkCRC) != 0 { + got := d.current.crc.Sum64() + var tmp [4]byte + binary.LittleEndian.PutUint32(tmp[:], uint32(got)) + if !bytes.Equal(tmp[:], next.d.checkCRC) && !ignoreCRC { + if debugDecoder { + println("CRC Check Failed:", tmp[:], " (got) !=", next.d.checkCRC, "(on stream)") + } + d.current.err = ErrCRCMismatch + } else { + if debugDecoder { + println("CRC ok", tmp[:]) + } + } + } + + return true +} + +func (d *Decoder) nextBlockSync() (ok bool) { + if d.current.d == nil { + d.current.d = <-d.decoders + } + for len(d.current.b) == 0 { + if !d.syncStream.inFrame { + d.frame.history.reset() + d.current.err = d.frame.reset(&d.syncStream.br) + if d.current.err != nil { + return false + } + if d.frame.DictionaryID != nil { + dict, ok := d.dicts[*d.frame.DictionaryID] + if !ok { + d.current.err = ErrUnknownDictionary + return false + } else { + d.frame.history.setDict(&dict) + } + } + if d.frame.WindowSize > d.o.maxDecodedSize || d.frame.WindowSize > d.o.maxWindowSize { + d.current.err = ErrDecoderSizeExceeded + return false + } + + d.syncStream.decodedFrame = 0 + d.syncStream.inFrame = true + } + d.current.err = d.frame.next(d.current.d) + if d.current.err != nil { + return false + } + d.frame.history.ensureBlock() + if debugDecoder { + println("History trimmed:", len(d.frame.history.b), "decoded already:", d.syncStream.decodedFrame) + } + histBefore := len(d.frame.history.b) + d.current.err = d.current.d.decodeBuf(&d.frame.history) + + if d.current.err != nil { + println("error after:", d.current.err) + return false + } + d.current.b = d.frame.history.b[histBefore:] + if debugDecoder { + println("history after:", len(d.frame.history.b)) + } + + // Check frame size (before CRC) + d.syncStream.decodedFrame += uint64(len(d.current.b)) + if d.syncStream.decodedFrame > d.frame.FrameContentSize { + if debugDecoder { + printf("DecodedFrame (%d) > FrameContentSize (%d)\n", d.syncStream.decodedFrame, d.frame.FrameContentSize) + } + d.current.err = ErrFrameSizeExceeded + return false + } + + // Check FCS + if d.current.d.Last && d.frame.FrameContentSize != fcsUnknown && d.syncStream.decodedFrame != d.frame.FrameContentSize { + if debugDecoder { + printf("DecodedFrame (%d) != FrameContentSize (%d)\n", d.syncStream.decodedFrame, d.frame.FrameContentSize) + } + d.current.err = ErrFrameSizeMismatch + return false + } + + // Update/Check CRC + if d.frame.HasCheckSum { + d.frame.crc.Write(d.current.b) + if d.current.d.Last { + d.current.err = d.frame.checkCRC() + if d.current.err != nil { + println("CRC error:", d.current.err) + return false + } + } + } + d.syncStream.inFrame = !d.current.d.Last + } + return true +} + +func (d *Decoder) stashDecoder() { + if d.current.d != nil { + if debugDecoder { + printf("re-adding current decoder %p", d.current.d) + } + d.decoders <- d.current.d + d.current.d = nil + } +} + +// Close will release all resources. +// It is NOT possible to reuse the decoder after this. +func (d *Decoder) Close() { + if d.current.err == ErrDecoderClosed { + return + } + d.drainOutput() + if d.current.cancel != nil { + d.current.cancel() + d.streamWg.Wait() + d.current.cancel = nil + } + if d.decoders != nil { + close(d.decoders) + for dec := range d.decoders { + dec.Close() + } + d.decoders = nil + } + if d.current.d != nil { + d.current.d.Close() + d.current.d = nil + } + d.current.err = ErrDecoderClosed +} + +// IOReadCloser returns the decoder as an io.ReadCloser for convenience. +// Any changes to the decoder will be reflected, so the returned ReadCloser +// can be reused along with the decoder. +// io.WriterTo is also supported by the returned ReadCloser. +func (d *Decoder) IOReadCloser() io.ReadCloser { + return closeWrapper{d: d} +} + +// closeWrapper wraps a function call as a closer. +type closeWrapper struct { + d *Decoder +} + +// WriteTo forwards WriteTo calls to the decoder. +func (c closeWrapper) WriteTo(w io.Writer) (n int64, err error) { + return c.d.WriteTo(w) +} + +// Read forwards read calls to the decoder. +func (c closeWrapper) Read(p []byte) (n int, err error) { + return c.d.Read(p) +} + +// Close closes the decoder. +func (c closeWrapper) Close() error { + c.d.Close() + return nil +} + +type decodeOutput struct { + d *blockDec + b []byte + err error +} + +func (d *Decoder) startSyncDecoder(r io.Reader) error { + d.frame.history.reset() + d.syncStream.br = readerWrapper{r: r} + d.syncStream.inFrame = false + d.syncStream.enabled = true + d.syncStream.decodedFrame = 0 + return nil +} + +// Create Decoder: +// ASYNC: +// Spawn 4 go routines. +// 0: Read frames and decode blocks. +// 1: Decode block and literals. Receives hufftree and seqdecs, returns seqdecs and huff tree. +// 2: Wait for recentOffsets if needed. Decode sequences, send recentOffsets. +// 3: Wait for stream history, execute sequences, send stream history. +func (d *Decoder) startStreamDecoder(ctx context.Context, r io.Reader, output chan decodeOutput) { + defer d.streamWg.Done() + br := readerWrapper{r: r} + + var seqPrepare = make(chan *blockDec, d.o.concurrent) + var seqDecode = make(chan *blockDec, d.o.concurrent) + var seqExecute = make(chan *blockDec, d.o.concurrent) + + // Async 1: Prepare blocks... + go func() { + var hist history + var hasErr bool + for block := range seqPrepare { + if hasErr { + if block != nil { + seqDecode <- block + } + continue + } + if block.async.newHist != nil { + if debugDecoder { + println("Async 1: new history") + } + hist.reset() + if block.async.newHist.dict != nil { + hist.setDict(block.async.newHist.dict) + } + } + if block.err != nil || block.Type != blockTypeCompressed { + hasErr = block.err != nil + seqDecode <- block + continue + } + + remain, err := block.decodeLiterals(block.data, &hist) + block.err = err + hasErr = block.err != nil + if err == nil { + block.async.literals = hist.decoders.literals + block.async.seqData = remain + } else if debugDecoder { + println("decodeLiterals error:", err) + } + seqDecode <- block + } + close(seqDecode) + }() + + // Async 2: Decode sequences... + go func() { + var hist history + var hasErr bool + + for block := range seqDecode { + if hasErr { + if block != nil { + seqExecute <- block + } + continue + } + if block.async.newHist != nil { + if debugDecoder { + println("Async 2: new history, recent:", block.async.newHist.recentOffsets) + } + hist.decoders = block.async.newHist.decoders + hist.recentOffsets = block.async.newHist.recentOffsets + hist.windowSize = block.async.newHist.windowSize + if block.async.newHist.dict != nil { + hist.setDict(block.async.newHist.dict) + } + } + if block.err != nil || block.Type != blockTypeCompressed { + hasErr = block.err != nil + seqExecute <- block + continue + } + + hist.decoders.literals = block.async.literals + block.err = block.prepareSequences(block.async.seqData, &hist) + if debugDecoder && block.err != nil { + println("prepareSequences returned:", block.err) + } + hasErr = block.err != nil + if block.err == nil { + block.err = block.decodeSequences(&hist) + if debugDecoder && block.err != nil { + println("decodeSequences returned:", block.err) + } + hasErr = block.err != nil + // block.async.sequence = hist.decoders.seq[:hist.decoders.nSeqs] + block.async.seqSize = hist.decoders.seqSize + } + seqExecute <- block + } + close(seqExecute) + }() + + var wg sync.WaitGroup + wg.Add(1) + + // Async 3: Execute sequences... + frameHistCache := d.frame.history.b + go func() { + var hist history + var decodedFrame uint64 + var fcs uint64 + var hasErr bool + for block := range seqExecute { + out := decodeOutput{err: block.err, d: block} + if block.err != nil || hasErr { + hasErr = true + output <- out + continue + } + if block.async.newHist != nil { + if debugDecoder { + println("Async 3: new history") + } + hist.windowSize = block.async.newHist.windowSize + hist.allocFrameBuffer = block.async.newHist.allocFrameBuffer + if block.async.newHist.dict != nil { + hist.setDict(block.async.newHist.dict) + } + + if cap(hist.b) < hist.allocFrameBuffer { + if cap(frameHistCache) >= hist.allocFrameBuffer { + hist.b = frameHistCache + } else { + hist.b = make([]byte, 0, hist.allocFrameBuffer) + println("Alloc history sized", hist.allocFrameBuffer) + } + } + hist.b = hist.b[:0] + fcs = block.async.fcs + decodedFrame = 0 + } + do := decodeOutput{err: block.err, d: block} + switch block.Type { + case blockTypeRLE: + if debugDecoder { + println("add rle block length:", block.RLESize) + } + + if cap(block.dst) < int(block.RLESize) { + if block.lowMem { + block.dst = make([]byte, block.RLESize) + } else { + block.dst = make([]byte, maxBlockSize) + } + } + block.dst = block.dst[:block.RLESize] + v := block.data[0] + for i := range block.dst { + block.dst[i] = v + } + hist.append(block.dst) + do.b = block.dst + case blockTypeRaw: + if debugDecoder { + println("add raw block length:", len(block.data)) + } + hist.append(block.data) + do.b = block.data + case blockTypeCompressed: + if debugDecoder { + println("execute with history length:", len(hist.b), "window:", hist.windowSize) + } + hist.decoders.seqSize = block.async.seqSize + hist.decoders.literals = block.async.literals + do.err = block.executeSequences(&hist) + hasErr = do.err != nil + if debugDecoder && hasErr { + println("executeSequences returned:", do.err) + } + do.b = block.dst + } + if !hasErr { + decodedFrame += uint64(len(do.b)) + if decodedFrame > fcs { + println("fcs exceeded", block.Last, fcs, decodedFrame) + do.err = ErrFrameSizeExceeded + hasErr = true + } else if block.Last && fcs != fcsUnknown && decodedFrame != fcs { + do.err = ErrFrameSizeMismatch + hasErr = true + } else { + if debugDecoder { + println("fcs ok", block.Last, fcs, decodedFrame) + } + } + } + output <- do + } + close(output) + frameHistCache = hist.b + wg.Done() + if debugDecoder { + println("decoder goroutines finished") + } + }() + +decodeStream: + for { + frame := d.frame + if debugDecoder { + println("New frame...") + } + var historySent bool + frame.history.reset() + err := frame.reset(&br) + if debugDecoder && err != nil { + println("Frame decoder returned", err) + } + if err == nil && frame.DictionaryID != nil { + dict, ok := d.dicts[*frame.DictionaryID] + if !ok { + err = ErrUnknownDictionary + } else { + frame.history.setDict(&dict) + } + } + if err == nil && d.frame.WindowSize > d.o.maxWindowSize { + err = ErrDecoderSizeExceeded + } + if err != nil { + select { + case <-ctx.Done(): + case dec := <-d.decoders: + dec.sendErr(err) + seqPrepare <- dec + } + break decodeStream + } + + // Go through all blocks of the frame. + for { + var dec *blockDec + select { + case <-ctx.Done(): + break decodeStream + case dec = <-d.decoders: + // Once we have a decoder, we MUST return it. + } + err := frame.next(dec) + if !historySent { + h := frame.history + if debugDecoder { + println("Alloc History:", h.allocFrameBuffer) + } + dec.async.newHist = &h + dec.async.fcs = frame.FrameContentSize + historySent = true + } else { + dec.async.newHist = nil + } + if debugDecoder && err != nil { + println("next block returned error:", err) + } + dec.err = err + dec.checkCRC = nil + if dec.Last && frame.HasCheckSum && err == nil { + crc, err := frame.rawInput.readSmall(4) + if err != nil { + println("CRC missing?", err) + dec.err = err + } + var tmp [4]byte + copy(tmp[:], crc) + dec.checkCRC = tmp[:] + if debugDecoder { + println("found crc to check:", dec.checkCRC) + } + } + err = dec.err + last := dec.Last + seqPrepare <- dec + if err != nil { + break decodeStream + } + if last { + break + } + } + } + close(seqPrepare) + wg.Wait() + d.frame.history.b = frameHistCache +} diff --git a/vendor/github.com/klauspost/compress/zstd/decoder_options.go b/vendor/github.com/klauspost/compress/zstd/decoder_options.go new file mode 100644 index 0000000..fd05c9b --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/decoder_options.go @@ -0,0 +1,114 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "errors" + "runtime" +) + +// DOption is an option for creating a decoder. +type DOption func(*decoderOptions) error + +// options retains accumulated state of multiple options. +type decoderOptions struct { + lowMem bool + concurrent int + maxDecodedSize uint64 + maxWindowSize uint64 + dicts []dict +} + +func (o *decoderOptions) setDefault() { + *o = decoderOptions{ + // use less ram: true for now, but may change. + lowMem: true, + concurrent: runtime.GOMAXPROCS(0), + maxWindowSize: MaxWindowSize, + } + if o.concurrent > 4 { + o.concurrent = 4 + } + o.maxDecodedSize = 1 << 63 +} + +// WithDecoderLowmem will set whether to use a lower amount of memory, +// but possibly have to allocate more while running. +func WithDecoderLowmem(b bool) DOption { + return func(o *decoderOptions) error { o.lowMem = b; return nil } +} + +// WithDecoderConcurrency sets the number of created decoders. +// When decoding block with DecodeAll, this will limit the number +// of possible concurrently running decodes. +// When decoding streams, this will limit the number of +// inflight blocks. +// When decoding streams and setting maximum to 1, +// no async decoding will be done. +// When a value of 0 is provided GOMAXPROCS will be used. +// By default this will be set to 4 or GOMAXPROCS, whatever is lower. +func WithDecoderConcurrency(n int) DOption { + return func(o *decoderOptions) error { + if n < 0 { + return errors.New("concurrency must be at least 1") + } + if n == 0 { + o.concurrent = runtime.GOMAXPROCS(0) + } else { + o.concurrent = n + } + return nil + } +} + +// WithDecoderMaxMemory allows to set a maximum decoded size for in-memory +// non-streaming operations or maximum window size for streaming operations. +// This can be used to control memory usage of potentially hostile content. +// Maximum and default is 1 << 63 bytes. +func WithDecoderMaxMemory(n uint64) DOption { + return func(o *decoderOptions) error { + if n == 0 { + return errors.New("WithDecoderMaxMemory must be at least 1") + } + if n > 1<<63 { + return errors.New("WithDecoderMaxmemory must be less than 1 << 63") + } + o.maxDecodedSize = n + return nil + } +} + +// WithDecoderDicts allows to register one or more dictionaries for the decoder. +// If several dictionaries with the same ID is provided the last one will be used. +func WithDecoderDicts(dicts ...[]byte) DOption { + return func(o *decoderOptions) error { + for _, b := range dicts { + d, err := loadDict(b) + if err != nil { + return err + } + o.dicts = append(o.dicts, *d) + } + return nil + } +} + +// WithDecoderMaxWindow allows to set a maximum window size for decodes. +// This allows rejecting packets that will cause big memory usage. +// The Decoder will likely allocate more memory based on the WithDecoderLowmem setting. +// If WithDecoderMaxMemory is set to a lower value, that will be used. +// Default is 512MB, Maximum is ~3.75 TB as per zstandard spec. +func WithDecoderMaxWindow(size uint64) DOption { + return func(o *decoderOptions) error { + if size < MinWindowSize { + return errors.New("WithMaxWindowSize must be at least 1KB, 1024 bytes") + } + if size > (1<<41)+7*(1<<38) { + return errors.New("WithMaxWindowSize must be less than (1<<41) + 7*(1<<38) ~ 3.75TB") + } + o.maxWindowSize = size + return nil + } +} diff --git a/vendor/github.com/klauspost/compress/zstd/dict.go b/vendor/github.com/klauspost/compress/zstd/dict.go new file mode 100644 index 0000000..a36ae83 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/dict.go @@ -0,0 +1,122 @@ +package zstd + +import ( + "bytes" + "encoding/binary" + "errors" + "fmt" + "io" + + "github.com/klauspost/compress/huff0" +) + +type dict struct { + id uint32 + + litEnc *huff0.Scratch + llDec, ofDec, mlDec sequenceDec + //llEnc, ofEnc, mlEnc []*fseEncoder + offsets [3]int + content []byte +} + +var dictMagic = [4]byte{0x37, 0xa4, 0x30, 0xec} + +// ID returns the dictionary id or 0 if d is nil. +func (d *dict) ID() uint32 { + if d == nil { + return 0 + } + return d.id +} + +// DictContentSize returns the dictionary content size or 0 if d is nil. +func (d *dict) DictContentSize() int { + if d == nil { + return 0 + } + return len(d.content) +} + +// Load a dictionary as described in +// https://github.com/facebook/zstd/blob/master/doc/zstd_compression_format.md#dictionary-format +func loadDict(b []byte) (*dict, error) { + // Check static field size. + if len(b) <= 8+(3*4) { + return nil, io.ErrUnexpectedEOF + } + d := dict{ + llDec: sequenceDec{fse: &fseDecoder{}}, + ofDec: sequenceDec{fse: &fseDecoder{}}, + mlDec: sequenceDec{fse: &fseDecoder{}}, + } + if !bytes.Equal(b[:4], dictMagic[:]) { + return nil, ErrMagicMismatch + } + d.id = binary.LittleEndian.Uint32(b[4:8]) + if d.id == 0 { + return nil, errors.New("dictionaries cannot have ID 0") + } + + // Read literal table + var err error + d.litEnc, b, err = huff0.ReadTable(b[8:], nil) + if err != nil { + return nil, err + } + d.litEnc.Reuse = huff0.ReusePolicyMust + + br := byteReader{ + b: b, + off: 0, + } + readDec := func(i tableIndex, dec *fseDecoder) error { + if err := dec.readNCount(&br, uint16(maxTableSymbol[i])); err != nil { + return err + } + if br.overread() { + return io.ErrUnexpectedEOF + } + err = dec.transform(symbolTableX[i]) + if err != nil { + println("Transform table error:", err) + return err + } + if debugDecoder || debugEncoder { + println("Read table ok", "symbolLen:", dec.symbolLen) + } + // Set decoders as predefined so they aren't reused. + dec.preDefined = true + return nil + } + + if err := readDec(tableOffsets, d.ofDec.fse); err != nil { + return nil, err + } + if err := readDec(tableMatchLengths, d.mlDec.fse); err != nil { + return nil, err + } + if err := readDec(tableLiteralLengths, d.llDec.fse); err != nil { + return nil, err + } + if br.remain() < 12 { + return nil, io.ErrUnexpectedEOF + } + + d.offsets[0] = int(br.Uint32()) + br.advance(4) + d.offsets[1] = int(br.Uint32()) + br.advance(4) + d.offsets[2] = int(br.Uint32()) + br.advance(4) + if d.offsets[0] <= 0 || d.offsets[1] <= 0 || d.offsets[2] <= 0 { + return nil, errors.New("invalid offset in dictionary") + } + d.content = make([]byte, br.remain()) + copy(d.content, br.unread()) + if d.offsets[0] > len(d.content) || d.offsets[1] > len(d.content) || d.offsets[2] > len(d.content) { + return nil, fmt.Errorf("initial offset bigger than dictionary content size %d, offsets: %v", len(d.content), d.offsets) + } + + return &d, nil +} diff --git a/vendor/github.com/klauspost/compress/zstd/enc_base.go b/vendor/github.com/klauspost/compress/zstd/enc_base.go new file mode 100644 index 0000000..15ae8ee --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/enc_base.go @@ -0,0 +1,188 @@ +package zstd + +import ( + "fmt" + "math/bits" + + "github.com/klauspost/compress/zstd/internal/xxhash" +) + +const ( + dictShardBits = 6 +) + +type fastBase struct { + // cur is the offset at the start of hist + cur int32 + // maximum offset. Should be at least 2x block size. + maxMatchOff int32 + hist []byte + crc *xxhash.Digest + tmp [8]byte + blk *blockEnc + lastDictID uint32 + lowMem bool +} + +// CRC returns the underlying CRC writer. +func (e *fastBase) CRC() *xxhash.Digest { + return e.crc +} + +// AppendCRC will append the CRC to the destination slice and return it. +func (e *fastBase) AppendCRC(dst []byte) []byte { + crc := e.crc.Sum(e.tmp[:0]) + dst = append(dst, crc[7], crc[6], crc[5], crc[4]) + return dst +} + +// WindowSize returns the window size of the encoder, +// or a window size small enough to contain the input size, if > 0. +func (e *fastBase) WindowSize(size int64) int32 { + if size > 0 && size < int64(e.maxMatchOff) { + b := int32(1) << uint(bits.Len(uint(size))) + // Keep minimum window. + if b < 1024 { + b = 1024 + } + return b + } + return e.maxMatchOff +} + +// Block returns the current block. +func (e *fastBase) Block() *blockEnc { + return e.blk +} + +func (e *fastBase) addBlock(src []byte) int32 { + if debugAsserts && e.cur > bufferReset { + panic(fmt.Sprintf("ecur (%d) > buffer reset (%d)", e.cur, bufferReset)) + } + // check if we have space already + if len(e.hist)+len(src) > cap(e.hist) { + if cap(e.hist) == 0 { + e.ensureHist(len(src)) + } else { + if cap(e.hist) < int(e.maxMatchOff+maxCompressedBlockSize) { + panic(fmt.Errorf("unexpected buffer cap %d, want at least %d with window %d", cap(e.hist), e.maxMatchOff+maxCompressedBlockSize, e.maxMatchOff)) + } + // Move down + offset := int32(len(e.hist)) - e.maxMatchOff + copy(e.hist[0:e.maxMatchOff], e.hist[offset:]) + e.cur += offset + e.hist = e.hist[:e.maxMatchOff] + } + } + s := int32(len(e.hist)) + e.hist = append(e.hist, src...) + return s +} + +// ensureHist will ensure that history can keep at least this many bytes. +func (e *fastBase) ensureHist(n int) { + if cap(e.hist) >= n { + return + } + l := e.maxMatchOff + if (e.lowMem && e.maxMatchOff > maxCompressedBlockSize) || e.maxMatchOff <= maxCompressedBlockSize { + l += maxCompressedBlockSize + } else { + l += e.maxMatchOff + } + // Make it at least 1MB. + if l < 1<<20 && !e.lowMem { + l = 1 << 20 + } + // Make it at least the requested size. + if l < int32(n) { + l = int32(n) + } + e.hist = make([]byte, 0, l) +} + +// useBlock will replace the block with the provided one, +// but transfer recent offsets from the previous. +func (e *fastBase) UseBlock(enc *blockEnc) { + enc.reset(e.blk) + e.blk = enc +} + +func (e *fastBase) matchlen(s, t int32, src []byte) int32 { + if debugAsserts { + if s < 0 { + err := fmt.Sprintf("s (%d) < 0", s) + panic(err) + } + if t < 0 { + err := fmt.Sprintf("s (%d) < 0", s) + panic(err) + } + if s-t > e.maxMatchOff { + err := fmt.Sprintf("s (%d) - t (%d) > maxMatchOff (%d)", s, t, e.maxMatchOff) + panic(err) + } + if len(src)-int(s) > maxCompressedBlockSize { + panic(fmt.Sprintf("len(src)-s (%d) > maxCompressedBlockSize (%d)", len(src)-int(s), maxCompressedBlockSize)) + } + } + a := src[s:] + b := src[t:] + b = b[:len(a)] + end := int32((len(a) >> 3) << 3) + for i := int32(0); i < end; i += 8 { + if diff := load6432(a, i) ^ load6432(b, i); diff != 0 { + return i + int32(bits.TrailingZeros64(diff)>>3) + } + } + + a = a[end:] + b = b[end:] + for i := range a { + if a[i] != b[i] { + return int32(i) + end + } + } + return int32(len(a)) + end +} + +// Reset the encoding table. +func (e *fastBase) resetBase(d *dict, singleBlock bool) { + if e.blk == nil { + e.blk = &blockEnc{lowMem: e.lowMem} + e.blk.init() + } else { + e.blk.reset(nil) + } + e.blk.initNewEncode() + if e.crc == nil { + e.crc = xxhash.New() + } else { + e.crc.Reset() + } + if d != nil { + low := e.lowMem + if singleBlock { + e.lowMem = true + } + e.ensureHist(d.DictContentSize() + maxCompressedBlockSize) + e.lowMem = low + } + + // We offset current position so everything will be out of reach. + // If above reset line, history will be purged. + if e.cur < bufferReset { + e.cur += e.maxMatchOff + int32(len(e.hist)) + } + e.hist = e.hist[:0] + if d != nil { + // Set offsets (currently not used) + for i, off := range d.offsets { + e.blk.recentOffsets[i] = uint32(off) + e.blk.prevRecentOffsets[i] = e.blk.recentOffsets[i] + } + // Transfer litenc. + e.blk.dictLitEnc = d.litEnc + e.hist = append(e.hist, d.content...) + } +} diff --git a/vendor/github.com/klauspost/compress/zstd/enc_best.go b/vendor/github.com/klauspost/compress/zstd/enc_best.go new file mode 100644 index 0000000..96028ec --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/enc_best.go @@ -0,0 +1,558 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "bytes" + "fmt" + + "github.com/klauspost/compress" +) + +const ( + bestLongTableBits = 22 // Bits used in the long match table + bestLongTableSize = 1 << bestLongTableBits // Size of the table + bestLongLen = 8 // Bytes used for table hash + + // Note: Increasing the short table bits or making the hash shorter + // can actually lead to compression degradation since it will 'steal' more from the + // long match table and match offsets are quite big. + // This greatly depends on the type of input. + bestShortTableBits = 18 // Bits used in the short match table + bestShortTableSize = 1 << bestShortTableBits // Size of the table + bestShortLen = 4 // Bytes used for table hash + +) + +type match struct { + offset int32 + s int32 + length int32 + rep int32 + est int32 +} + +const highScore = 25000 + +// estBits will estimate output bits from predefined tables. +func (m *match) estBits(bitsPerByte int32) { + mlc := mlCode(uint32(m.length - zstdMinMatch)) + var ofc uint8 + if m.rep < 0 { + ofc = ofCode(uint32(m.s-m.offset) + 3) + } else { + ofc = ofCode(uint32(m.rep)) + } + // Cost, excluding + ofTT, mlTT := fsePredefEnc[tableOffsets].ct.symbolTT[ofc], fsePredefEnc[tableMatchLengths].ct.symbolTT[mlc] + + // Add cost of match encoding... + m.est = int32(ofTT.outBits + mlTT.outBits) + m.est += int32(ofTT.deltaNbBits>>16 + mlTT.deltaNbBits>>16) + // Subtract savings compared to literal encoding... + m.est -= (m.length * bitsPerByte) >> 10 + if m.est > 0 { + // Unlikely gain.. + m.length = 0 + m.est = highScore + } +} + +// bestFastEncoder uses 2 tables, one for short matches (5 bytes) and one for long matches. +// The long match table contains the previous entry with the same hash, +// effectively making it a "chain" of length 2. +// When we find a long match we choose between the two values and select the longest. +// When we find a short match, after checking the long, we check if we can find a long at n+1 +// and that it is longer (lazy matching). +type bestFastEncoder struct { + fastBase + table [bestShortTableSize]prevEntry + longTable [bestLongTableSize]prevEntry + dictTable []prevEntry + dictLongTable []prevEntry +} + +// Encode improves compression... +func (e *bestFastEncoder) Encode(blk *blockEnc, src []byte) { + const ( + // Input margin is the number of bytes we read (8) + // and the maximum we will read ahead (2) + inputMargin = 8 + 4 + minNonLiteralBlockSize = 16 + ) + + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = prevEntry{} + } + for i := range e.longTable[:] { + e.longTable[i] = prevEntry{} + } + e.cur = e.maxMatchOff + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + v2 := e.table[i].prev + if v < minOff { + v = 0 + v2 = 0 + } else { + v = v - e.cur + e.maxMatchOff + if v2 < minOff { + v2 = 0 + } else { + v2 = v2 - e.cur + e.maxMatchOff + } + } + e.table[i] = prevEntry{ + offset: v, + prev: v2, + } + } + for i := range e.longTable[:] { + v := e.longTable[i].offset + v2 := e.longTable[i].prev + if v < minOff { + v = 0 + v2 = 0 + } else { + v = v - e.cur + e.maxMatchOff + if v2 < minOff { + v2 = 0 + } else { + v2 = v2 - e.cur + e.maxMatchOff + } + } + e.longTable[i] = prevEntry{ + offset: v, + prev: v2, + } + } + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Use this to estimate literal cost. + // Scaled by 10 bits. + bitsPerByte := int32((compress.ShannonEntropyBits(src) * 1024) / len(src)) + // Huffman can never go < 1 bit/byte + if bitsPerByte < 1024 { + bitsPerByte = 1024 + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + const kSearchStrength = 10 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + offset3 := int32(blk.recentOffsets[2]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + _ = addLiterals + + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + // We allow the encoder to optionally turn off repeat offsets across blocks + canRepeat := len(blk.sequences) > 2 + + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + bestOf := func(a, b match) match { + if a.est+(a.s-b.s)*bitsPerByte>>10 < b.est+(b.s-a.s)*bitsPerByte>>10 { + return a + } + return b + } + const goodEnough = 100 + + nextHashL := hashLen(cv, bestLongTableBits, bestLongLen) + nextHashS := hashLen(cv, bestShortTableBits, bestShortLen) + candidateL := e.longTable[nextHashL] + candidateS := e.table[nextHashS] + + matchAt := func(offset int32, s int32, first uint32, rep int32) match { + if s-offset >= e.maxMatchOff || load3232(src, offset) != first { + return match{s: s, est: highScore} + } + if debugAsserts { + if !bytes.Equal(src[s:s+4], src[offset:offset+4]) { + panic(fmt.Sprintf("first match mismatch: %v != %v, first: %08x", src[s:s+4], src[offset:offset+4], first)) + } + } + m := match{offset: offset, s: s, length: 4 + e.matchlen(s+4, offset+4, src), rep: rep} + m.estBits(bitsPerByte) + return m + } + + best := bestOf(matchAt(candidateL.offset-e.cur, s, uint32(cv), -1), matchAt(candidateL.prev-e.cur, s, uint32(cv), -1)) + best = bestOf(best, matchAt(candidateS.offset-e.cur, s, uint32(cv), -1)) + best = bestOf(best, matchAt(candidateS.prev-e.cur, s, uint32(cv), -1)) + + if canRepeat && best.length < goodEnough { + cv32 := uint32(cv >> 8) + spp := s + 1 + best = bestOf(best, matchAt(spp-offset1, spp, cv32, 1)) + best = bestOf(best, matchAt(spp-offset2, spp, cv32, 2)) + best = bestOf(best, matchAt(spp-offset3, spp, cv32, 3)) + if best.length > 0 { + cv32 = uint32(cv >> 24) + spp += 2 + best = bestOf(best, matchAt(spp-offset1, spp, cv32, 1)) + best = bestOf(best, matchAt(spp-offset2, spp, cv32, 2)) + best = bestOf(best, matchAt(spp-offset3, spp, cv32, 3)) + } + } + // Load next and check... + e.longTable[nextHashL] = prevEntry{offset: s + e.cur, prev: candidateL.offset} + e.table[nextHashS] = prevEntry{offset: s + e.cur, prev: candidateS.offset} + + // Look far ahead, unless we have a really long match already... + if best.length < goodEnough { + // No match found, move forward on input, no need to check forward... + if best.length < 4 { + s += 1 + (s-nextEmit)>>(kSearchStrength-1) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + continue + } + + s++ + candidateS = e.table[hashLen(cv>>8, bestShortTableBits, bestShortLen)] + cv = load6432(src, s) + cv2 := load6432(src, s+1) + candidateL = e.longTable[hashLen(cv, bestLongTableBits, bestLongLen)] + candidateL2 := e.longTable[hashLen(cv2, bestLongTableBits, bestLongLen)] + + // Short at s+1 + best = bestOf(best, matchAt(candidateS.offset-e.cur, s, uint32(cv), -1)) + // Long at s+1, s+2 + best = bestOf(best, matchAt(candidateL.offset-e.cur, s, uint32(cv), -1)) + best = bestOf(best, matchAt(candidateL.prev-e.cur, s, uint32(cv), -1)) + best = bestOf(best, matchAt(candidateL2.offset-e.cur, s+1, uint32(cv2), -1)) + best = bestOf(best, matchAt(candidateL2.prev-e.cur, s+1, uint32(cv2), -1)) + if false { + // Short at s+3. + // Too often worse... + best = bestOf(best, matchAt(e.table[hashLen(cv2>>8, bestShortTableBits, bestShortLen)].offset-e.cur, s+2, uint32(cv2>>8), -1)) + } + // See if we can find a better match by checking where the current best ends. + // Use that offset to see if we can find a better full match. + if sAt := best.s + best.length; sAt < sLimit { + nextHashL := hashLen(load6432(src, sAt), bestLongTableBits, bestLongLen) + candidateEnd := e.longTable[nextHashL] + if pos := candidateEnd.offset - e.cur - best.length; pos >= 0 { + bestEnd := bestOf(best, matchAt(pos, best.s, load3232(src, best.s), -1)) + if pos := candidateEnd.prev - e.cur - best.length; pos >= 0 { + bestEnd = bestOf(bestEnd, matchAt(pos, best.s, load3232(src, best.s), -1)) + } + best = bestEnd + } + } + } + + if debugAsserts { + if !bytes.Equal(src[best.s:best.s+best.length], src[best.offset:best.offset+best.length]) { + panic(fmt.Sprintf("match mismatch: %v != %v", src[best.s:best.s+best.length], src[best.offset:best.offset+best.length])) + } + } + + // We have a match, we can store the forward value + if best.rep > 0 { + s = best.s + var seq seq + seq.matchLen = uint32(best.length - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := best.s + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + repIndex := best.offset + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = uint32(best.rep) + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Index match start+1 (long) -> s - 1 + index0 := s + s = best.s + best.length + + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, best.length) + + } + break encodeLoop + } + // Index skipped... + off := index0 + e.cur + for index0 < s-1 { + cv0 := load6432(src, index0) + h0 := hashLen(cv0, bestLongTableBits, bestLongLen) + h1 := hashLen(cv0, bestShortTableBits, bestShortLen) + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.table[h1] = prevEntry{offset: off, prev: e.table[h1].offset} + off++ + index0++ + } + switch best.rep { + case 2: + offset1, offset2 = offset2, offset1 + case 3: + offset1, offset2, offset3 = offset3, offset1, offset2 + } + cv = load6432(src, s) + continue + } + + // A 4-byte match has been found. Update recent offsets. + // We'll later see if more than 4 bytes. + s = best.s + t := best.offset + offset1, offset2, offset3 = s-t, offset1, offset2 + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the n-byte match as long as possible. + l := best.length + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + + // Index match start+1 (long) -> s - 1 + index0 := s - l + 1 + // every entry + for index0 < s-1 { + cv0 := load6432(src, index0) + h0 := hashLen(cv0, bestLongTableBits, bestLongLen) + h1 := hashLen(cv0, bestShortTableBits, bestShortLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.table[h1] = prevEntry{offset: off, prev: e.table[h1].offset} + index0++ + } + + cv = load6432(src, s) + if !canRepeat { + continue + } + + // Check offset 2 + for { + o2 := s - offset2 + if load3232(src, o2) != uint32(cv) { + // Do regular search + break + } + + // Store this, since we have it. + nextHashS := hashLen(cv, bestShortTableBits, bestShortLen) + nextHashL := hashLen(cv, bestLongTableBits, bestLongLen) + + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + e.longTable[nextHashL] = prevEntry{offset: s + e.cur, prev: e.longTable[nextHashL].offset} + e.table[nextHashS] = prevEntry{offset: s + e.cur, prev: e.table[nextHashS].offset} + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + // Finished + break encodeLoop + } + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + blk.recentOffsets[2] = uint32(offset3) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } +} + +// EncodeNoHist will encode a block with no history and no following blocks. +// Most notable difference is that src will not be copied for history and +// we do not need to check for max match length. +func (e *bestFastEncoder) EncodeNoHist(blk *blockEnc, src []byte) { + e.ensureHist(len(src)) + e.Encode(blk, src) +} + +// Reset will reset and set a dictionary if not nil +func (e *bestFastEncoder) Reset(d *dict, singleBlock bool) { + e.resetBase(d, singleBlock) + if d == nil { + return + } + // Init or copy dict table + if len(e.dictTable) != len(e.table) || d.id != e.lastDictID { + if len(e.dictTable) != len(e.table) { + e.dictTable = make([]prevEntry, len(e.table)) + } + end := int32(len(d.content)) - 8 + e.maxMatchOff + for i := e.maxMatchOff; i < end; i += 4 { + const hashLog = bestShortTableBits + + cv := load6432(d.content, i-e.maxMatchOff) + nextHash := hashLen(cv, hashLog, bestShortLen) // 0 -> 4 + nextHash1 := hashLen(cv>>8, hashLog, bestShortLen) // 1 -> 5 + nextHash2 := hashLen(cv>>16, hashLog, bestShortLen) // 2 -> 6 + nextHash3 := hashLen(cv>>24, hashLog, bestShortLen) // 3 -> 7 + e.dictTable[nextHash] = prevEntry{ + prev: e.dictTable[nextHash].offset, + offset: i, + } + e.dictTable[nextHash1] = prevEntry{ + prev: e.dictTable[nextHash1].offset, + offset: i + 1, + } + e.dictTable[nextHash2] = prevEntry{ + prev: e.dictTable[nextHash2].offset, + offset: i + 2, + } + e.dictTable[nextHash3] = prevEntry{ + prev: e.dictTable[nextHash3].offset, + offset: i + 3, + } + } + e.lastDictID = d.id + } + + // Init or copy dict table + if len(e.dictLongTable) != len(e.longTable) || d.id != e.lastDictID { + if len(e.dictLongTable) != len(e.longTable) { + e.dictLongTable = make([]prevEntry, len(e.longTable)) + } + if len(d.content) >= 8 { + cv := load6432(d.content, 0) + h := hashLen(cv, bestLongTableBits, bestLongLen) + e.dictLongTable[h] = prevEntry{ + offset: e.maxMatchOff, + prev: e.dictLongTable[h].offset, + } + + end := int32(len(d.content)) - 8 + e.maxMatchOff + off := 8 // First to read + for i := e.maxMatchOff + 1; i < end; i++ { + cv = cv>>8 | (uint64(d.content[off]) << 56) + h := hashLen(cv, bestLongTableBits, bestLongLen) + e.dictLongTable[h] = prevEntry{ + offset: i, + prev: e.dictLongTable[h].offset, + } + off++ + } + } + e.lastDictID = d.id + } + // Reset table to initial state + copy(e.longTable[:], e.dictLongTable) + + e.cur = e.maxMatchOff + // Reset table to initial state + copy(e.table[:], e.dictTable) +} diff --git a/vendor/github.com/klauspost/compress/zstd/enc_better.go b/vendor/github.com/klauspost/compress/zstd/enc_better.go new file mode 100644 index 0000000..602c05e --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/enc_better.go @@ -0,0 +1,1237 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import "fmt" + +const ( + betterLongTableBits = 19 // Bits used in the long match table + betterLongTableSize = 1 << betterLongTableBits // Size of the table + betterLongLen = 8 // Bytes used for table hash + + // Note: Increasing the short table bits or making the hash shorter + // can actually lead to compression degradation since it will 'steal' more from the + // long match table and match offsets are quite big. + // This greatly depends on the type of input. + betterShortTableBits = 13 // Bits used in the short match table + betterShortTableSize = 1 << betterShortTableBits // Size of the table + betterShortLen = 5 // Bytes used for table hash + + betterLongTableShardCnt = 1 << (betterLongTableBits - dictShardBits) // Number of shards in the table + betterLongTableShardSize = betterLongTableSize / betterLongTableShardCnt // Size of an individual shard + + betterShortTableShardCnt = 1 << (betterShortTableBits - dictShardBits) // Number of shards in the table + betterShortTableShardSize = betterShortTableSize / betterShortTableShardCnt // Size of an individual shard +) + +type prevEntry struct { + offset int32 + prev int32 +} + +// betterFastEncoder uses 2 tables, one for short matches (5 bytes) and one for long matches. +// The long match table contains the previous entry with the same hash, +// effectively making it a "chain" of length 2. +// When we find a long match we choose between the two values and select the longest. +// When we find a short match, after checking the long, we check if we can find a long at n+1 +// and that it is longer (lazy matching). +type betterFastEncoder struct { + fastBase + table [betterShortTableSize]tableEntry + longTable [betterLongTableSize]prevEntry +} + +type betterFastEncoderDict struct { + betterFastEncoder + dictTable []tableEntry + dictLongTable []prevEntry + shortTableShardDirty [betterShortTableShardCnt]bool + longTableShardDirty [betterLongTableShardCnt]bool + allDirty bool +} + +// Encode improves compression... +func (e *betterFastEncoder) Encode(blk *blockEnc, src []byte) { + const ( + // Input margin is the number of bytes we read (8) + // and the maximum we will read ahead (2) + inputMargin = 8 + 2 + minNonLiteralBlockSize = 16 + ) + + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + for i := range e.longTable[:] { + e.longTable[i] = prevEntry{} + } + e.cur = e.maxMatchOff + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.table[i].offset = v + } + for i := range e.longTable[:] { + v := e.longTable[i].offset + v2 := e.longTable[i].prev + if v < minOff { + v = 0 + v2 = 0 + } else { + v = v - e.cur + e.maxMatchOff + if v2 < minOff { + v2 = 0 + } else { + v2 = v2 - e.cur + e.maxMatchOff + } + } + e.longTable[i] = prevEntry{ + offset: v, + prev: v2, + } + } + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 1. + const stepSize = 1 + + const kSearchStrength = 9 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + var t int32 + // We allow the encoder to optionally turn off repeat offsets across blocks + canRepeat := len(blk.sequences) > 2 + var matched int32 + + for { + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + nextHashS := hashLen(cv, betterShortTableBits, betterShortLen) + nextHashL := hashLen(cv, betterLongTableBits, betterLongLen) + candidateL := e.longTable[nextHashL] + candidateS := e.table[nextHashS] + + const repOff = 1 + repIndex := s - offset1 + repOff + off := s + e.cur + e.longTable[nextHashL] = prevEntry{offset: off, prev: candidateL.offset} + e.table[nextHashS] = tableEntry{offset: off, val: uint32(cv)} + + if canRepeat { + if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) { + // Consider history as well. + var seq seq + lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src) + + seq.matchLen = uint32(lenght - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Index match start+1 (long) -> s - 1 + index0 := s + repOff + s += lenght + repOff + + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, lenght) + + } + break encodeLoop + } + // Index skipped... + for index0 < s-1 { + cv0 := load6432(src, index0) + cv1 := cv0 >> 8 + h0 := hashLen(cv0, betterLongTableBits, betterLongLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.table[hashLen(cv1, betterShortTableBits, betterShortLen)] = tableEntry{offset: off + 1, val: uint32(cv1)} + index0 += 2 + } + cv = load6432(src, s) + continue + } + const repOff2 = 1 + + // We deviate from the reference encoder and also check offset 2. + // Still slower and not much better, so disabled. + // repIndex = s - offset2 + repOff2 + if false && repIndex >= 0 && load6432(src, repIndex) == load6432(src, s+repOff) { + // Consider history as well. + var seq seq + lenght := 8 + e.matchlen(s+8+repOff2, repIndex+8, src) + + seq.matchLen = uint32(lenght - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff2 + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 2 + seq.offset = 2 + if debugSequences { + println("repeat sequence 2", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + index0 := s + repOff2 + s += lenght + repOff2 + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, lenght) + + } + break encodeLoop + } + + // Index skipped... + for index0 < s-1 { + cv0 := load6432(src, index0) + cv1 := cv0 >> 8 + h0 := hashLen(cv0, betterLongTableBits, betterLongLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.table[hashLen(cv1, betterShortTableBits, betterShortLen)] = tableEntry{offset: off + 1, val: uint32(cv1)} + index0 += 2 + } + cv = load6432(src, s) + // Swap offsets + offset1, offset2 = offset2, offset1 + continue + } + } + // Find the offsets of our two matches. + coffsetL := candidateL.offset - e.cur + coffsetLP := candidateL.prev - e.cur + + // Check if we have a long match. + if s-coffsetL < e.maxMatchOff && cv == load6432(src, coffsetL) { + // Found a long match, at least 8 bytes. + matched = e.matchlen(s+8, coffsetL+8, src) + 8 + t = coffsetL + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + + if s-coffsetLP < e.maxMatchOff && cv == load6432(src, coffsetLP) { + // Found a long match, at least 8 bytes. + prevMatch := e.matchlen(s+8, coffsetLP+8, src) + 8 + if prevMatch > matched { + matched = prevMatch + t = coffsetLP + } + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + } + break + } + + // Check if we have a long match on prev. + if s-coffsetLP < e.maxMatchOff && cv == load6432(src, coffsetLP) { + // Found a long match, at least 8 bytes. + matched = e.matchlen(s+8, coffsetLP+8, src) + 8 + t = coffsetLP + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + break + } + + coffsetS := candidateS.offset - e.cur + + // Check if we have a short match. + if s-coffsetS < e.maxMatchOff && uint32(cv) == candidateS.val { + // found a regular match + matched = e.matchlen(s+4, coffsetS+4, src) + 4 + + // See if we can find a long match at s+1 + const checkAt = 1 + cv := load6432(src, s+checkAt) + nextHashL = hashLen(cv, betterLongTableBits, betterLongLen) + candidateL = e.longTable[nextHashL] + coffsetL = candidateL.offset - e.cur + + // We can store it, since we have at least a 4 byte match. + e.longTable[nextHashL] = prevEntry{offset: s + checkAt + e.cur, prev: candidateL.offset} + if s-coffsetL < e.maxMatchOff && cv == load6432(src, coffsetL) { + // Found a long match, at least 8 bytes. + matchedNext := e.matchlen(s+8+checkAt, coffsetL+8, src) + 8 + if matchedNext > matched { + t = coffsetL + s += checkAt + matched = matchedNext + if debugMatches { + println("long match (after short)") + } + break + } + } + + // Check prev long... + coffsetL = candidateL.prev - e.cur + if s-coffsetL < e.maxMatchOff && cv == load6432(src, coffsetL) { + // Found a long match, at least 8 bytes. + matchedNext := e.matchlen(s+8+checkAt, coffsetL+8, src) + 8 + if matchedNext > matched { + t = coffsetL + s += checkAt + matched = matchedNext + if debugMatches { + println("prev long match (after short)") + } + break + } + } + t = coffsetS + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + if debugMatches { + println("short match") + } + break + } + + // No match found, move forward in input. + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + + // Try to find a better match by searching for a long match at the end of the current best match + if s+matched < sLimit { + nextHashL := hashLen(load6432(src, s+matched), betterLongTableBits, betterLongLen) + cv := load3232(src, s) + candidateL := e.longTable[nextHashL] + coffsetL := candidateL.offset - e.cur - matched + if coffsetL >= 0 && coffsetL < s && s-coffsetL < e.maxMatchOff && cv == load3232(src, coffsetL) { + // Found a long match, at least 4 bytes. + matchedNext := e.matchlen(s+4, coffsetL+4, src) + 4 + if matchedNext > matched { + t = coffsetL + matched = matchedNext + if debugMatches { + println("long match at end-of-match") + } + } + } + + // Check prev long... + if true { + coffsetL = candidateL.prev - e.cur - matched + if coffsetL >= 0 && coffsetL < s && s-coffsetL < e.maxMatchOff && cv == load3232(src, coffsetL) { + // Found a long match, at least 4 bytes. + matchedNext := e.matchlen(s+4, coffsetL+4, src) + 4 + if matchedNext > matched { + t = coffsetL + matched = matchedNext + if debugMatches { + println("prev long match at end-of-match") + } + } + } + } + } + // A match has been found. Update recent offsets. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && canRepeat && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the n-byte match as long as possible. + l := matched + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + + // Index match start+1 (long) -> s - 1 + index0 := s - l + 1 + for index0 < s-1 { + cv0 := load6432(src, index0) + cv1 := cv0 >> 8 + h0 := hashLen(cv0, betterLongTableBits, betterLongLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.table[hashLen(cv1, betterShortTableBits, betterShortLen)] = tableEntry{offset: off + 1, val: uint32(cv1)} + index0 += 2 + } + + cv = load6432(src, s) + if !canRepeat { + continue + } + + // Check offset 2 + for { + o2 := s - offset2 + if load3232(src, o2) != uint32(cv) { + // Do regular search + break + } + + // Store this, since we have it. + nextHashS := hashLen(cv, betterShortTableBits, betterShortLen) + nextHashL := hashLen(cv, betterLongTableBits, betterLongLen) + + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + e.longTable[nextHashL] = prevEntry{offset: s + e.cur, prev: e.longTable[nextHashL].offset} + e.table[nextHashS] = tableEntry{offset: s + e.cur, val: uint32(cv)} + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + // Finished + break encodeLoop + } + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } +} + +// EncodeNoHist will encode a block with no history and no following blocks. +// Most notable difference is that src will not be copied for history and +// we do not need to check for max match length. +func (e *betterFastEncoder) EncodeNoHist(blk *blockEnc, src []byte) { + e.ensureHist(len(src)) + e.Encode(blk, src) +} + +// Encode improves compression... +func (e *betterFastEncoderDict) Encode(blk *blockEnc, src []byte) { + const ( + // Input margin is the number of bytes we read (8) + // and the maximum we will read ahead (2) + inputMargin = 8 + 2 + minNonLiteralBlockSize = 16 + ) + + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + for i := range e.longTable[:] { + e.longTable[i] = prevEntry{} + } + e.cur = e.maxMatchOff + e.allDirty = true + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.table[i].offset = v + } + for i := range e.longTable[:] { + v := e.longTable[i].offset + v2 := e.longTable[i].prev + if v < minOff { + v = 0 + v2 = 0 + } else { + v = v - e.cur + e.maxMatchOff + if v2 < minOff { + v2 = 0 + } else { + v2 = v2 - e.cur + e.maxMatchOff + } + } + e.longTable[i] = prevEntry{ + offset: v, + prev: v2, + } + } + e.allDirty = true + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 1. + const stepSize = 1 + + const kSearchStrength = 9 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + var t int32 + // We allow the encoder to optionally turn off repeat offsets across blocks + canRepeat := len(blk.sequences) > 2 + var matched int32 + + for { + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + nextHashS := hashLen(cv, betterShortTableBits, betterShortLen) + nextHashL := hashLen(cv, betterLongTableBits, betterLongLen) + candidateL := e.longTable[nextHashL] + candidateS := e.table[nextHashS] + + const repOff = 1 + repIndex := s - offset1 + repOff + off := s + e.cur + e.longTable[nextHashL] = prevEntry{offset: off, prev: candidateL.offset} + e.markLongShardDirty(nextHashL) + e.table[nextHashS] = tableEntry{offset: off, val: uint32(cv)} + e.markShortShardDirty(nextHashS) + + if canRepeat { + if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) { + // Consider history as well. + var seq seq + lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src) + + seq.matchLen = uint32(lenght - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Index match start+1 (long) -> s - 1 + index0 := s + repOff + s += lenght + repOff + + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, lenght) + + } + break encodeLoop + } + // Index skipped... + for index0 < s-1 { + cv0 := load6432(src, index0) + cv1 := cv0 >> 8 + h0 := hashLen(cv0, betterLongTableBits, betterLongLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.markLongShardDirty(h0) + h1 := hashLen(cv1, betterShortTableBits, betterShortLen) + e.table[h1] = tableEntry{offset: off + 1, val: uint32(cv1)} + e.markShortShardDirty(h1) + index0 += 2 + } + cv = load6432(src, s) + continue + } + const repOff2 = 1 + + // We deviate from the reference encoder and also check offset 2. + // Still slower and not much better, so disabled. + // repIndex = s - offset2 + repOff2 + if false && repIndex >= 0 && load6432(src, repIndex) == load6432(src, s+repOff) { + // Consider history as well. + var seq seq + lenght := 8 + e.matchlen(s+8+repOff2, repIndex+8, src) + + seq.matchLen = uint32(lenght - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff2 + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 2 + seq.offset = 2 + if debugSequences { + println("repeat sequence 2", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + index0 := s + repOff2 + s += lenght + repOff2 + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, lenght) + + } + break encodeLoop + } + + // Index skipped... + for index0 < s-1 { + cv0 := load6432(src, index0) + cv1 := cv0 >> 8 + h0 := hashLen(cv0, betterLongTableBits, betterLongLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.markLongShardDirty(h0) + h1 := hashLen(cv1, betterShortTableBits, betterShortLen) + e.table[h1] = tableEntry{offset: off + 1, val: uint32(cv1)} + e.markShortShardDirty(h1) + index0 += 2 + } + cv = load6432(src, s) + // Swap offsets + offset1, offset2 = offset2, offset1 + continue + } + } + // Find the offsets of our two matches. + coffsetL := candidateL.offset - e.cur + coffsetLP := candidateL.prev - e.cur + + // Check if we have a long match. + if s-coffsetL < e.maxMatchOff && cv == load6432(src, coffsetL) { + // Found a long match, at least 8 bytes. + matched = e.matchlen(s+8, coffsetL+8, src) + 8 + t = coffsetL + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + + if s-coffsetLP < e.maxMatchOff && cv == load6432(src, coffsetLP) { + // Found a long match, at least 8 bytes. + prevMatch := e.matchlen(s+8, coffsetLP+8, src) + 8 + if prevMatch > matched { + matched = prevMatch + t = coffsetLP + } + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + } + break + } + + // Check if we have a long match on prev. + if s-coffsetLP < e.maxMatchOff && cv == load6432(src, coffsetLP) { + // Found a long match, at least 8 bytes. + matched = e.matchlen(s+8, coffsetLP+8, src) + 8 + t = coffsetLP + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + break + } + + coffsetS := candidateS.offset - e.cur + + // Check if we have a short match. + if s-coffsetS < e.maxMatchOff && uint32(cv) == candidateS.val { + // found a regular match + matched = e.matchlen(s+4, coffsetS+4, src) + 4 + + // See if we can find a long match at s+1 + const checkAt = 1 + cv := load6432(src, s+checkAt) + nextHashL = hashLen(cv, betterLongTableBits, betterLongLen) + candidateL = e.longTable[nextHashL] + coffsetL = candidateL.offset - e.cur + + // We can store it, since we have at least a 4 byte match. + e.longTable[nextHashL] = prevEntry{offset: s + checkAt + e.cur, prev: candidateL.offset} + e.markLongShardDirty(nextHashL) + if s-coffsetL < e.maxMatchOff && cv == load6432(src, coffsetL) { + // Found a long match, at least 8 bytes. + matchedNext := e.matchlen(s+8+checkAt, coffsetL+8, src) + 8 + if matchedNext > matched { + t = coffsetL + s += checkAt + matched = matchedNext + if debugMatches { + println("long match (after short)") + } + break + } + } + + // Check prev long... + coffsetL = candidateL.prev - e.cur + if s-coffsetL < e.maxMatchOff && cv == load6432(src, coffsetL) { + // Found a long match, at least 8 bytes. + matchedNext := e.matchlen(s+8+checkAt, coffsetL+8, src) + 8 + if matchedNext > matched { + t = coffsetL + s += checkAt + matched = matchedNext + if debugMatches { + println("prev long match (after short)") + } + break + } + } + t = coffsetS + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + if debugMatches { + println("short match") + } + break + } + + // No match found, move forward in input. + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + // Try to find a better match by searching for a long match at the end of the current best match + if s+matched < sLimit { + nextHashL := hashLen(load6432(src, s+matched), betterLongTableBits, betterLongLen) + cv := load3232(src, s) + candidateL := e.longTable[nextHashL] + coffsetL := candidateL.offset - e.cur - matched + if coffsetL >= 0 && coffsetL < s && s-coffsetL < e.maxMatchOff && cv == load3232(src, coffsetL) { + // Found a long match, at least 4 bytes. + matchedNext := e.matchlen(s+4, coffsetL+4, src) + 4 + if matchedNext > matched { + t = coffsetL + matched = matchedNext + if debugMatches { + println("long match at end-of-match") + } + } + } + + // Check prev long... + if true { + coffsetL = candidateL.prev - e.cur - matched + if coffsetL >= 0 && coffsetL < s && s-coffsetL < e.maxMatchOff && cv == load3232(src, coffsetL) { + // Found a long match, at least 4 bytes. + matchedNext := e.matchlen(s+4, coffsetL+4, src) + 4 + if matchedNext > matched { + t = coffsetL + matched = matchedNext + if debugMatches { + println("prev long match at end-of-match") + } + } + } + } + } + // A match has been found. Update recent offsets. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && canRepeat && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the n-byte match as long as possible. + l := matched + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + + // Index match start+1 (long) -> s - 1 + index0 := s - l + 1 + for index0 < s-1 { + cv0 := load6432(src, index0) + cv1 := cv0 >> 8 + h0 := hashLen(cv0, betterLongTableBits, betterLongLen) + off := index0 + e.cur + e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset} + e.markLongShardDirty(h0) + h1 := hashLen(cv1, betterShortTableBits, betterShortLen) + e.table[h1] = tableEntry{offset: off + 1, val: uint32(cv1)} + e.markShortShardDirty(h1) + index0 += 2 + } + + cv = load6432(src, s) + if !canRepeat { + continue + } + + // Check offset 2 + for { + o2 := s - offset2 + if load3232(src, o2) != uint32(cv) { + // Do regular search + break + } + + // Store this, since we have it. + nextHashS := hashLen(cv, betterShortTableBits, betterShortLen) + nextHashL := hashLen(cv, betterLongTableBits, betterLongLen) + + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + e.longTable[nextHashL] = prevEntry{offset: s + e.cur, prev: e.longTable[nextHashL].offset} + e.markLongShardDirty(nextHashL) + e.table[nextHashS] = tableEntry{offset: s + e.cur, val: uint32(cv)} + e.markShortShardDirty(nextHashS) + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + // Finished + break encodeLoop + } + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } +} + +// ResetDict will reset and set a dictionary if not nil +func (e *betterFastEncoder) Reset(d *dict, singleBlock bool) { + e.resetBase(d, singleBlock) + if d != nil { + panic("betterFastEncoder: Reset with dict") + } +} + +// ResetDict will reset and set a dictionary if not nil +func (e *betterFastEncoderDict) Reset(d *dict, singleBlock bool) { + e.resetBase(d, singleBlock) + if d == nil { + return + } + // Init or copy dict table + if len(e.dictTable) != len(e.table) || d.id != e.lastDictID { + if len(e.dictTable) != len(e.table) { + e.dictTable = make([]tableEntry, len(e.table)) + } + end := int32(len(d.content)) - 8 + e.maxMatchOff + for i := e.maxMatchOff; i < end; i += 4 { + const hashLog = betterShortTableBits + + cv := load6432(d.content, i-e.maxMatchOff) + nextHash := hashLen(cv, hashLog, betterShortLen) // 0 -> 4 + nextHash1 := hashLen(cv>>8, hashLog, betterShortLen) // 1 -> 5 + nextHash2 := hashLen(cv>>16, hashLog, betterShortLen) // 2 -> 6 + nextHash3 := hashLen(cv>>24, hashLog, betterShortLen) // 3 -> 7 + e.dictTable[nextHash] = tableEntry{ + val: uint32(cv), + offset: i, + } + e.dictTable[nextHash1] = tableEntry{ + val: uint32(cv >> 8), + offset: i + 1, + } + e.dictTable[nextHash2] = tableEntry{ + val: uint32(cv >> 16), + offset: i + 2, + } + e.dictTable[nextHash3] = tableEntry{ + val: uint32(cv >> 24), + offset: i + 3, + } + } + e.lastDictID = d.id + e.allDirty = true + } + + // Init or copy dict table + if len(e.dictLongTable) != len(e.longTable) || d.id != e.lastDictID { + if len(e.dictLongTable) != len(e.longTable) { + e.dictLongTable = make([]prevEntry, len(e.longTable)) + } + if len(d.content) >= 8 { + cv := load6432(d.content, 0) + h := hashLen(cv, betterLongTableBits, betterLongLen) + e.dictLongTable[h] = prevEntry{ + offset: e.maxMatchOff, + prev: e.dictLongTable[h].offset, + } + + end := int32(len(d.content)) - 8 + e.maxMatchOff + off := 8 // First to read + for i := e.maxMatchOff + 1; i < end; i++ { + cv = cv>>8 | (uint64(d.content[off]) << 56) + h := hashLen(cv, betterLongTableBits, betterLongLen) + e.dictLongTable[h] = prevEntry{ + offset: i, + prev: e.dictLongTable[h].offset, + } + off++ + } + } + e.lastDictID = d.id + e.allDirty = true + } + + // Reset table to initial state + { + dirtyShardCnt := 0 + if !e.allDirty { + for i := range e.shortTableShardDirty { + if e.shortTableShardDirty[i] { + dirtyShardCnt++ + } + } + } + const shardCnt = betterShortTableShardCnt + const shardSize = betterShortTableShardSize + if e.allDirty || dirtyShardCnt > shardCnt*4/6 { + copy(e.table[:], e.dictTable) + for i := range e.shortTableShardDirty { + e.shortTableShardDirty[i] = false + } + } else { + for i := range e.shortTableShardDirty { + if !e.shortTableShardDirty[i] { + continue + } + + copy(e.table[i*shardSize:(i+1)*shardSize], e.dictTable[i*shardSize:(i+1)*shardSize]) + e.shortTableShardDirty[i] = false + } + } + } + { + dirtyShardCnt := 0 + if !e.allDirty { + for i := range e.shortTableShardDirty { + if e.shortTableShardDirty[i] { + dirtyShardCnt++ + } + } + } + const shardCnt = betterLongTableShardCnt + const shardSize = betterLongTableShardSize + if e.allDirty || dirtyShardCnt > shardCnt*4/6 { + copy(e.longTable[:], e.dictLongTable) + for i := range e.longTableShardDirty { + e.longTableShardDirty[i] = false + } + } else { + for i := range e.longTableShardDirty { + if !e.longTableShardDirty[i] { + continue + } + + copy(e.longTable[i*shardSize:(i+1)*shardSize], e.dictLongTable[i*shardSize:(i+1)*shardSize]) + e.longTableShardDirty[i] = false + } + } + } + e.cur = e.maxMatchOff + e.allDirty = false +} + +func (e *betterFastEncoderDict) markLongShardDirty(entryNum uint32) { + e.longTableShardDirty[entryNum/betterLongTableShardSize] = true +} + +func (e *betterFastEncoderDict) markShortShardDirty(entryNum uint32) { + e.shortTableShardDirty[entryNum/betterShortTableShardSize] = true +} diff --git a/vendor/github.com/klauspost/compress/zstd/enc_dfast.go b/vendor/github.com/klauspost/compress/zstd/enc_dfast.go new file mode 100644 index 0000000..d6b3104 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/enc_dfast.go @@ -0,0 +1,1124 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import "fmt" + +const ( + dFastLongTableBits = 17 // Bits used in the long match table + dFastLongTableSize = 1 << dFastLongTableBits // Size of the table + dFastLongTableMask = dFastLongTableSize - 1 // Mask for table indices. Redundant, but can eliminate bounds checks. + dFastLongLen = 8 // Bytes used for table hash + + dLongTableShardCnt = 1 << (dFastLongTableBits - dictShardBits) // Number of shards in the table + dLongTableShardSize = dFastLongTableSize / tableShardCnt // Size of an individual shard + + dFastShortTableBits = tableBits // Bits used in the short match table + dFastShortTableSize = 1 << dFastShortTableBits // Size of the table + dFastShortTableMask = dFastShortTableSize - 1 // Mask for table indices. Redundant, but can eliminate bounds checks. + dFastShortLen = 5 // Bytes used for table hash + +) + +type doubleFastEncoder struct { + fastEncoder + longTable [dFastLongTableSize]tableEntry +} + +type doubleFastEncoderDict struct { + fastEncoderDict + longTable [dFastLongTableSize]tableEntry + dictLongTable []tableEntry + longTableShardDirty [dLongTableShardCnt]bool +} + +// Encode mimmics functionality in zstd_dfast.c +func (e *doubleFastEncoder) Encode(blk *blockEnc, src []byte) { + const ( + // Input margin is the number of bytes we read (8) + // and the maximum we will read ahead (2) + inputMargin = 8 + 2 + minNonLiteralBlockSize = 16 + ) + + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + for i := range e.longTable[:] { + e.longTable[i] = tableEntry{} + } + e.cur = e.maxMatchOff + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.table[i].offset = v + } + for i := range e.longTable[:] { + v := e.longTable[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.longTable[i].offset = v + } + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 1. + const stepSize = 1 + + const kSearchStrength = 8 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + var t int32 + // We allow the encoder to optionally turn off repeat offsets across blocks + canRepeat := len(blk.sequences) > 2 + + for { + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + nextHashS := hashLen(cv, dFastShortTableBits, dFastShortLen) + nextHashL := hashLen(cv, dFastLongTableBits, dFastLongLen) + candidateL := e.longTable[nextHashL] + candidateS := e.table[nextHashS] + + const repOff = 1 + repIndex := s - offset1 + repOff + entry := tableEntry{offset: s + e.cur, val: uint32(cv)} + e.longTable[nextHashL] = entry + e.table[nextHashS] = entry + + if canRepeat { + if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) { + // Consider history as well. + var seq seq + lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src) + + seq.matchLen = uint32(lenght - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + s += lenght + repOff + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, lenght) + + } + break encodeLoop + } + cv = load6432(src, s) + continue + } + } + // Find the offsets of our two matches. + coffsetL := s - (candidateL.offset - e.cur) + coffsetS := s - (candidateS.offset - e.cur) + + // Check if we have a long match. + if coffsetL < e.maxMatchOff && uint32(cv) == candidateL.val { + // Found a long match, likely at least 8 bytes. + // Reference encoder checks all 8 bytes, we only check 4, + // but the likelihood of both the first 4 bytes and the hash matching should be enough. + t = candidateL.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + break + } + + // Check if we have a short match. + if coffsetS < e.maxMatchOff && uint32(cv) == candidateS.val { + // found a regular match + // See if we can find a long match at s+1 + const checkAt = 1 + cv := load6432(src, s+checkAt) + nextHashL = hashLen(cv, dFastLongTableBits, dFastLongLen) + candidateL = e.longTable[nextHashL] + coffsetL = s - (candidateL.offset - e.cur) + checkAt + + // We can store it, since we have at least a 4 byte match. + e.longTable[nextHashL] = tableEntry{offset: s + checkAt + e.cur, val: uint32(cv)} + if coffsetL < e.maxMatchOff && uint32(cv) == candidateL.val { + // Found a long match, likely at least 8 bytes. + // Reference encoder checks all 8 bytes, we only check 4, + // but the likelihood of both the first 4 bytes and the hash matching should be enough. + t = candidateL.offset - e.cur + s += checkAt + if debugMatches { + println("long match (after short)") + } + break + } + + t = candidateS.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + if debugMatches { + println("short match") + } + break + } + + // No match found, move forward in input. + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + + // A 4-byte match has been found. Update recent offsets. + // We'll later see if more than 4 bytes. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && canRepeat && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the 4-byte match as long as possible. + l := e.matchlen(s+4, t+4, src) + 4 + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + + // Index match start+1 (long) and start+2 (short) + index0 := s - l + 1 + // Index match end-2 (long) and end-1 (short) + index1 := s - 2 + + cv0 := load6432(src, index0) + cv1 := load6432(src, index1) + te0 := tableEntry{offset: index0 + e.cur, val: uint32(cv0)} + te1 := tableEntry{offset: index1 + e.cur, val: uint32(cv1)} + e.longTable[hashLen(cv0, dFastLongTableBits, dFastLongLen)] = te0 + e.longTable[hashLen(cv1, dFastLongTableBits, dFastLongLen)] = te1 + cv0 >>= 8 + cv1 >>= 8 + te0.offset++ + te1.offset++ + te0.val = uint32(cv0) + te1.val = uint32(cv1) + e.table[hashLen(cv0, dFastShortTableBits, dFastShortLen)] = te0 + e.table[hashLen(cv1, dFastShortTableBits, dFastShortLen)] = te1 + + cv = load6432(src, s) + + if !canRepeat { + continue + } + + // Check offset 2 + for { + o2 := s - offset2 + if load3232(src, o2) != uint32(cv) { + // Do regular search + break + } + + // Store this, since we have it. + nextHashS := hashLen(cv, dFastShortTableBits, dFastShortLen) + nextHashL := hashLen(cv, dFastLongTableBits, dFastLongLen) + + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + entry := tableEntry{offset: s + e.cur, val: uint32(cv)} + e.longTable[nextHashL] = entry + e.table[nextHashS] = entry + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + // Finished + break encodeLoop + } + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } +} + +// EncodeNoHist will encode a block with no history and no following blocks. +// Most notable difference is that src will not be copied for history and +// we do not need to check for max match length. +func (e *doubleFastEncoder) EncodeNoHist(blk *blockEnc, src []byte) { + const ( + // Input margin is the number of bytes we read (8) + // and the maximum we will read ahead (2) + inputMargin = 8 + 2 + minNonLiteralBlockSize = 16 + ) + + // Protect against e.cur wraparound. + if e.cur >= bufferReset { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + for i := range e.longTable[:] { + e.longTable[i] = tableEntry{} + } + e.cur = e.maxMatchOff + } + + s := int32(0) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 1. + const stepSize = 1 + + const kSearchStrength = 8 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + var t int32 + for { + + nextHashS := hashLen(cv, dFastShortTableBits, dFastShortLen) + nextHashL := hashLen(cv, dFastLongTableBits, dFastLongLen) + candidateL := e.longTable[nextHashL] + candidateS := e.table[nextHashS] + + const repOff = 1 + repIndex := s - offset1 + repOff + entry := tableEntry{offset: s + e.cur, val: uint32(cv)} + e.longTable[nextHashL] = entry + e.table[nextHashS] = entry + + if len(blk.sequences) > 2 { + if load3232(src, repIndex) == uint32(cv>>(repOff*8)) { + // Consider history as well. + var seq seq + //length := 4 + e.matchlen(s+4+repOff, repIndex+4, src) + length := 4 + int32(matchLen(src[s+4+repOff:], src[repIndex+4:])) + + seq.matchLen = uint32(length - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + s += length + repOff + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, length) + + } + break encodeLoop + } + cv = load6432(src, s) + continue + } + } + // Find the offsets of our two matches. + coffsetL := s - (candidateL.offset - e.cur) + coffsetS := s - (candidateS.offset - e.cur) + + // Check if we have a long match. + if coffsetL < e.maxMatchOff && uint32(cv) == candidateL.val { + // Found a long match, likely at least 8 bytes. + // Reference encoder checks all 8 bytes, we only check 4, + // but the likelihood of both the first 4 bytes and the hash matching should be enough. + t = candidateL.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d). cur: %d", s, t, e.cur)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + break + } + + // Check if we have a short match. + if coffsetS < e.maxMatchOff && uint32(cv) == candidateS.val { + // found a regular match + // See if we can find a long match at s+1 + const checkAt = 1 + cv := load6432(src, s+checkAt) + nextHashL = hashLen(cv, dFastLongTableBits, dFastLongLen) + candidateL = e.longTable[nextHashL] + coffsetL = s - (candidateL.offset - e.cur) + checkAt + + // We can store it, since we have at least a 4 byte match. + e.longTable[nextHashL] = tableEntry{offset: s + checkAt + e.cur, val: uint32(cv)} + if coffsetL < e.maxMatchOff && uint32(cv) == candidateL.val { + // Found a long match, likely at least 8 bytes. + // Reference encoder checks all 8 bytes, we only check 4, + // but the likelihood of both the first 4 bytes and the hash matching should be enough. + t = candidateL.offset - e.cur + s += checkAt + if debugMatches { + println("long match (after short)") + } + break + } + + t = candidateS.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + if debugMatches { + println("short match") + } + break + } + + // No match found, move forward in input. + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + + // A 4-byte match has been found. Update recent offsets. + // We'll later see if more than 4 bytes. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + // Extend the 4-byte match as long as possible. + //l := e.matchlen(s+4, t+4, src) + 4 + l := int32(matchLen(src[s+4:], src[t+4:])) + 4 + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] { + s-- + t-- + l++ + } + + // Write our sequence + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + + // Index match start+1 (long) and start+2 (short) + index0 := s - l + 1 + // Index match end-2 (long) and end-1 (short) + index1 := s - 2 + + cv0 := load6432(src, index0) + cv1 := load6432(src, index1) + te0 := tableEntry{offset: index0 + e.cur, val: uint32(cv0)} + te1 := tableEntry{offset: index1 + e.cur, val: uint32(cv1)} + e.longTable[hashLen(cv0, dFastLongTableBits, dFastLongLen)] = te0 + e.longTable[hashLen(cv1, dFastLongTableBits, dFastLongLen)] = te1 + cv0 >>= 8 + cv1 >>= 8 + te0.offset++ + te1.offset++ + te0.val = uint32(cv0) + te1.val = uint32(cv1) + e.table[hashLen(cv0, dFastShortTableBits, dFastShortLen)] = te0 + e.table[hashLen(cv1, dFastShortTableBits, dFastShortLen)] = te1 + + cv = load6432(src, s) + + if len(blk.sequences) <= 2 { + continue + } + + // Check offset 2 + for { + o2 := s - offset2 + if load3232(src, o2) != uint32(cv) { + // Do regular search + break + } + + // Store this, since we have it. + nextHashS := hashLen(cv1>>8, dFastShortTableBits, dFastShortLen) + nextHashL := hashLen(cv, dFastLongTableBits, dFastLongLen) + + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + //l := 4 + e.matchlen(s+4, o2+4, src) + l := 4 + int32(matchLen(src[s+4:], src[o2+4:])) + + entry := tableEntry{offset: s + e.cur, val: uint32(cv)} + e.longTable[nextHashL] = entry + e.table[nextHashS] = entry + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + // Finished + break encodeLoop + } + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } + + // We do not store history, so we must offset e.cur to avoid false matches for next user. + if e.cur < bufferReset { + e.cur += int32(len(src)) + } +} + +// Encode will encode the content, with a dictionary if initialized for it. +func (e *doubleFastEncoderDict) Encode(blk *blockEnc, src []byte) { + const ( + // Input margin is the number of bytes we read (8) + // and the maximum we will read ahead (2) + inputMargin = 8 + 2 + minNonLiteralBlockSize = 16 + ) + + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + for i := range e.longTable[:] { + e.longTable[i] = tableEntry{} + } + e.markAllShardsDirty() + e.cur = e.maxMatchOff + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.table[i].offset = v + } + for i := range e.longTable[:] { + v := e.longTable[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.longTable[i].offset = v + } + e.markAllShardsDirty() + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 1. + const stepSize = 1 + + const kSearchStrength = 8 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + var t int32 + // We allow the encoder to optionally turn off repeat offsets across blocks + canRepeat := len(blk.sequences) > 2 + + for { + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + nextHashS := hashLen(cv, dFastShortTableBits, dFastShortLen) + nextHashL := hashLen(cv, dFastLongTableBits, dFastLongLen) + candidateL := e.longTable[nextHashL] + candidateS := e.table[nextHashS] + + const repOff = 1 + repIndex := s - offset1 + repOff + entry := tableEntry{offset: s + e.cur, val: uint32(cv)} + e.longTable[nextHashL] = entry + e.markLongShardDirty(nextHashL) + e.table[nextHashS] = entry + e.markShardDirty(nextHashS) + + if canRepeat { + if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) { + // Consider history as well. + var seq seq + lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src) + + seq.matchLen = uint32(lenght - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + repOff + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + s += lenght + repOff + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, lenght) + + } + break encodeLoop + } + cv = load6432(src, s) + continue + } + } + // Find the offsets of our two matches. + coffsetL := s - (candidateL.offset - e.cur) + coffsetS := s - (candidateS.offset - e.cur) + + // Check if we have a long match. + if coffsetL < e.maxMatchOff && uint32(cv) == candidateL.val { + // Found a long match, likely at least 8 bytes. + // Reference encoder checks all 8 bytes, we only check 4, + // but the likelihood of both the first 4 bytes and the hash matching should be enough. + t = candidateL.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugMatches { + println("long match") + } + break + } + + // Check if we have a short match. + if coffsetS < e.maxMatchOff && uint32(cv) == candidateS.val { + // found a regular match + // See if we can find a long match at s+1 + const checkAt = 1 + cv := load6432(src, s+checkAt) + nextHashL = hashLen(cv, dFastLongTableBits, dFastLongLen) + candidateL = e.longTable[nextHashL] + coffsetL = s - (candidateL.offset - e.cur) + checkAt + + // We can store it, since we have at least a 4 byte match. + e.longTable[nextHashL] = tableEntry{offset: s + checkAt + e.cur, val: uint32(cv)} + e.markLongShardDirty(nextHashL) + if coffsetL < e.maxMatchOff && uint32(cv) == candidateL.val { + // Found a long match, likely at least 8 bytes. + // Reference encoder checks all 8 bytes, we only check 4, + // but the likelihood of both the first 4 bytes and the hash matching should be enough. + t = candidateL.offset - e.cur + s += checkAt + if debugMatches { + println("long match (after short)") + } + break + } + + t = candidateS.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + if debugMatches { + println("short match") + } + break + } + + // No match found, move forward in input. + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + + // A 4-byte match has been found. Update recent offsets. + // We'll later see if more than 4 bytes. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && canRepeat && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the 4-byte match as long as possible. + l := e.matchlen(s+4, t+4, src) + 4 + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + + // Index match start+1 (long) and start+2 (short) + index0 := s - l + 1 + // Index match end-2 (long) and end-1 (short) + index1 := s - 2 + + cv0 := load6432(src, index0) + cv1 := load6432(src, index1) + te0 := tableEntry{offset: index0 + e.cur, val: uint32(cv0)} + te1 := tableEntry{offset: index1 + e.cur, val: uint32(cv1)} + longHash1 := hashLen(cv0, dFastLongTableBits, dFastLongLen) + longHash2 := hashLen(cv0, dFastLongTableBits, dFastLongLen) + e.longTable[longHash1] = te0 + e.longTable[longHash2] = te1 + e.markLongShardDirty(longHash1) + e.markLongShardDirty(longHash2) + cv0 >>= 8 + cv1 >>= 8 + te0.offset++ + te1.offset++ + te0.val = uint32(cv0) + te1.val = uint32(cv1) + hashVal1 := hashLen(cv0, dFastShortTableBits, dFastShortLen) + hashVal2 := hashLen(cv1, dFastShortTableBits, dFastShortLen) + e.table[hashVal1] = te0 + e.markShardDirty(hashVal1) + e.table[hashVal2] = te1 + e.markShardDirty(hashVal2) + + cv = load6432(src, s) + + if !canRepeat { + continue + } + + // Check offset 2 + for { + o2 := s - offset2 + if load3232(src, o2) != uint32(cv) { + // Do regular search + break + } + + // Store this, since we have it. + nextHashS := hashLen(cv, dFastShortTableBits, dFastShortLen) + nextHashL := hashLen(cv, dFastLongTableBits, dFastLongLen) + + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + entry := tableEntry{offset: s + e.cur, val: uint32(cv)} + e.longTable[nextHashL] = entry + e.markLongShardDirty(nextHashL) + e.table[nextHashS] = entry + e.markShardDirty(nextHashS) + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + // Finished + break encodeLoop + } + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } + // If we encoded more than 64K mark all dirty. + if len(src) > 64<<10 { + e.markAllShardsDirty() + } +} + +// ResetDict will reset and set a dictionary if not nil +func (e *doubleFastEncoder) Reset(d *dict, singleBlock bool) { + e.fastEncoder.Reset(d, singleBlock) + if d != nil { + panic("doubleFastEncoder: Reset with dict not supported") + } +} + +// ResetDict will reset and set a dictionary if not nil +func (e *doubleFastEncoderDict) Reset(d *dict, singleBlock bool) { + allDirty := e.allDirty + e.fastEncoderDict.Reset(d, singleBlock) + if d == nil { + return + } + + // Init or copy dict table + if len(e.dictLongTable) != len(e.longTable) || d.id != e.lastDictID { + if len(e.dictLongTable) != len(e.longTable) { + e.dictLongTable = make([]tableEntry, len(e.longTable)) + } + if len(d.content) >= 8 { + cv := load6432(d.content, 0) + e.dictLongTable[hashLen(cv, dFastLongTableBits, dFastLongLen)] = tableEntry{ + val: uint32(cv), + offset: e.maxMatchOff, + } + end := int32(len(d.content)) - 8 + e.maxMatchOff + for i := e.maxMatchOff + 1; i < end; i++ { + cv = cv>>8 | (uint64(d.content[i-e.maxMatchOff+7]) << 56) + e.dictLongTable[hashLen(cv, dFastLongTableBits, dFastLongLen)] = tableEntry{ + val: uint32(cv), + offset: i, + } + } + } + e.lastDictID = d.id + e.allDirty = true + } + // Reset table to initial state + e.cur = e.maxMatchOff + + dirtyShardCnt := 0 + if !allDirty { + for i := range e.longTableShardDirty { + if e.longTableShardDirty[i] { + dirtyShardCnt++ + } + } + } + + if allDirty || dirtyShardCnt > dLongTableShardCnt/2 { + copy(e.longTable[:], e.dictLongTable) + for i := range e.longTableShardDirty { + e.longTableShardDirty[i] = false + } + return + } + for i := range e.longTableShardDirty { + if !e.longTableShardDirty[i] { + continue + } + + copy(e.longTable[i*dLongTableShardSize:(i+1)*dLongTableShardSize], e.dictLongTable[i*dLongTableShardSize:(i+1)*dLongTableShardSize]) + e.longTableShardDirty[i] = false + } +} + +func (e *doubleFastEncoderDict) markLongShardDirty(entryNum uint32) { + e.longTableShardDirty[entryNum/dLongTableShardSize] = true +} diff --git a/vendor/github.com/klauspost/compress/zstd/enc_fast.go b/vendor/github.com/klauspost/compress/zstd/enc_fast.go new file mode 100644 index 0000000..f51ab52 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/enc_fast.go @@ -0,0 +1,898 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "fmt" +) + +const ( + tableBits = 15 // Bits used in the table + tableSize = 1 << tableBits // Size of the table + tableShardCnt = 1 << (tableBits - dictShardBits) // Number of shards in the table + tableShardSize = tableSize / tableShardCnt // Size of an individual shard + tableFastHashLen = 6 + tableMask = tableSize - 1 // Mask for table indices. Redundant, but can eliminate bounds checks. + maxMatchLength = 131074 +) + +type tableEntry struct { + val uint32 + offset int32 +} + +type fastEncoder struct { + fastBase + table [tableSize]tableEntry +} + +type fastEncoderDict struct { + fastEncoder + dictTable []tableEntry + tableShardDirty [tableShardCnt]bool + allDirty bool +} + +// Encode mimmics functionality in zstd_fast.c +func (e *fastEncoder) Encode(blk *blockEnc, src []byte) { + const ( + inputMargin = 8 + minNonLiteralBlockSize = 1 + 1 + inputMargin + ) + + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + e.cur = e.maxMatchOff + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.table[i].offset = v + } + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 2. + const stepSize = 2 + + // TEMPLATE + const hashLog = tableBits + // seems global, but would be nice to tweak. + const kSearchStrength = 6 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + // t will contain the match offset when we find one. + // When existing the search loop, we have already checked 4 bytes. + var t int32 + + // We will not use repeat offsets across blocks. + // By not using them for the first 3 matches + canRepeat := len(blk.sequences) > 2 + + for { + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + nextHash := hashLen(cv, hashLog, tableFastHashLen) + nextHash2 := hashLen(cv>>8, hashLog, tableFastHashLen) + candidate := e.table[nextHash] + candidate2 := e.table[nextHash2] + repIndex := s - offset1 + 2 + + e.table[nextHash] = tableEntry{offset: s + e.cur, val: uint32(cv)} + e.table[nextHash2] = tableEntry{offset: s + e.cur + 1, val: uint32(cv >> 8)} + + if canRepeat && repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>16) { + // Consider history as well. + var seq seq + var length int32 + length = 4 + e.matchlen(s+6, repIndex+4, src) + seq.matchLen = uint32(length - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + 2 + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + sMin := s - e.maxMatchOff + if sMin < 0 { + sMin = 0 + } + for repIndex > sMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + s += length + 2 + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, length) + + } + break encodeLoop + } + cv = load6432(src, s) + continue + } + coffset0 := s - (candidate.offset - e.cur) + coffset1 := s - (candidate2.offset - e.cur) + 1 + if coffset0 < e.maxMatchOff && uint32(cv) == candidate.val { + // found a regular match + t = candidate.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + break + } + + if coffset1 < e.maxMatchOff && uint32(cv>>8) == candidate2.val { + // found a regular match + t = candidate2.offset - e.cur + s++ + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + break + } + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + // A 4-byte match has been found. We'll later see if more than 4 bytes. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && canRepeat && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the 4-byte match as long as possible. + l := e.matchlen(s+4, t+4, src) + 4 + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence. + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + // Don't use repeat offsets + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + + // Check offset 2 + if o2 := s - offset2; canRepeat && load3232(src, o2) == uint32(cv) { + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + // Store this, since we have it. + nextHash := hashLen(cv, hashLog, tableFastHashLen) + e.table[nextHash] = tableEntry{offset: s + e.cur, val: uint32(cv)} + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + break encodeLoop + } + // Prepare next loop. + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } +} + +// EncodeNoHist will encode a block with no history and no following blocks. +// Most notable difference is that src will not be copied for history and +// we do not need to check for max match length. +func (e *fastEncoder) EncodeNoHist(blk *blockEnc, src []byte) { + const ( + inputMargin = 8 + minNonLiteralBlockSize = 1 + 1 + inputMargin + ) + if debugEncoder { + if len(src) > maxBlockSize { + panic("src too big") + } + } + + // Protect against e.cur wraparound. + if e.cur >= bufferReset { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + e.cur = e.maxMatchOff + } + + s := int32(0) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 2. + const stepSize = 2 + + // TEMPLATE + const hashLog = tableBits + // seems global, but would be nice to tweak. + const kSearchStrength = 6 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + // t will contain the match offset when we find one. + // When existing the search loop, we have already checked 4 bytes. + var t int32 + + // We will not use repeat offsets across blocks. + // By not using them for the first 3 matches + + for { + nextHash := hashLen(cv, hashLog, tableFastHashLen) + nextHash2 := hashLen(cv>>8, hashLog, tableFastHashLen) + candidate := e.table[nextHash] + candidate2 := e.table[nextHash2] + repIndex := s - offset1 + 2 + + e.table[nextHash] = tableEntry{offset: s + e.cur, val: uint32(cv)} + e.table[nextHash2] = tableEntry{offset: s + e.cur + 1, val: uint32(cv >> 8)} + + if len(blk.sequences) > 2 && load3232(src, repIndex) == uint32(cv>>16) { + // Consider history as well. + var seq seq + length := 4 + e.matchlen(s+6, repIndex+4, src) + + seq.matchLen = uint32(length - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + 2 + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + sMin := s - e.maxMatchOff + if sMin < 0 { + sMin = 0 + } + for repIndex > sMin && start > startLimit && src[repIndex-1] == src[start-1] { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + s += length + 2 + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, length) + + } + break encodeLoop + } + cv = load6432(src, s) + continue + } + coffset0 := s - (candidate.offset - e.cur) + coffset1 := s - (candidate2.offset - e.cur) + 1 + if coffset0 < e.maxMatchOff && uint32(cv) == candidate.val { + // found a regular match + t = candidate.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic(fmt.Sprintf("t (%d) < 0, candidate.offset: %d, e.cur: %d, coffset0: %d, e.maxMatchOff: %d", t, candidate.offset, e.cur, coffset0, e.maxMatchOff)) + } + break + } + + if coffset1 < e.maxMatchOff && uint32(cv>>8) == candidate2.val { + // found a regular match + t = candidate2.offset - e.cur + s++ + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + break + } + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + // A 4-byte match has been found. We'll later see if more than 4 bytes. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && t < 0 { + panic(fmt.Sprintf("t (%d) < 0 ", t)) + } + // Extend the 4-byte match as long as possible. + l := e.matchlen(s+4, t+4, src) + 4 + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] { + s-- + t-- + l++ + } + + // Write our sequence. + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + // Don't use repeat offsets + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + + // Check offset 2 + if o2 := s - offset2; len(blk.sequences) > 2 && load3232(src, o2) == uint32(cv) { + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + // Store this, since we have it. + nextHash := hashLen(cv, hashLog, tableFastHashLen) + e.table[nextHash] = tableEntry{offset: s + e.cur, val: uint32(cv)} + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + break encodeLoop + } + // Prepare next loop. + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } + // We do not store history, so we must offset e.cur to avoid false matches for next user. + if e.cur < bufferReset { + e.cur += int32(len(src)) + } +} + +// Encode will encode the content, with a dictionary if initialized for it. +func (e *fastEncoderDict) Encode(blk *blockEnc, src []byte) { + const ( + inputMargin = 8 + minNonLiteralBlockSize = 1 + 1 + inputMargin + ) + if e.allDirty || len(src) > 32<<10 { + e.fastEncoder.Encode(blk, src) + e.allDirty = true + return + } + // Protect against e.cur wraparound. + for e.cur >= bufferReset { + if len(e.hist) == 0 { + for i := range e.table[:] { + e.table[i] = tableEntry{} + } + e.cur = e.maxMatchOff + break + } + // Shift down everything in the table that isn't already too far away. + minOff := e.cur + int32(len(e.hist)) - e.maxMatchOff + for i := range e.table[:] { + v := e.table[i].offset + if v < minOff { + v = 0 + } else { + v = v - e.cur + e.maxMatchOff + } + e.table[i].offset = v + } + e.cur = e.maxMatchOff + break + } + + s := e.addBlock(src) + blk.size = len(src) + if len(src) < minNonLiteralBlockSize { + blk.extraLits = len(src) + blk.literals = blk.literals[:len(src)] + copy(blk.literals, src) + return + } + + // Override src + src = e.hist + sLimit := int32(len(src)) - inputMargin + // stepSize is the number of bytes to skip on every main loop iteration. + // It should be >= 2. + const stepSize = 2 + + // TEMPLATE + const hashLog = tableBits + // seems global, but would be nice to tweak. + const kSearchStrength = 7 + + // nextEmit is where in src the next emitLiteral should start from. + nextEmit := s + cv := load6432(src, s) + + // Relative offsets + offset1 := int32(blk.recentOffsets[0]) + offset2 := int32(blk.recentOffsets[1]) + + addLiterals := func(s *seq, until int32) { + if until == nextEmit { + return + } + blk.literals = append(blk.literals, src[nextEmit:until]...) + s.litLen = uint32(until - nextEmit) + } + if debugEncoder { + println("recent offsets:", blk.recentOffsets) + } + +encodeLoop: + for { + // t will contain the match offset when we find one. + // When existing the search loop, we have already checked 4 bytes. + var t int32 + + // We will not use repeat offsets across blocks. + // By not using them for the first 3 matches + canRepeat := len(blk.sequences) > 2 + + for { + if debugAsserts && canRepeat && offset1 == 0 { + panic("offset0 was 0") + } + + nextHash := hashLen(cv, hashLog, tableFastHashLen) + nextHash2 := hashLen(cv>>8, hashLog, tableFastHashLen) + candidate := e.table[nextHash] + candidate2 := e.table[nextHash2] + repIndex := s - offset1 + 2 + + e.table[nextHash] = tableEntry{offset: s + e.cur, val: uint32(cv)} + e.markShardDirty(nextHash) + e.table[nextHash2] = tableEntry{offset: s + e.cur + 1, val: uint32(cv >> 8)} + e.markShardDirty(nextHash2) + + if canRepeat && repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>16) { + // Consider history as well. + var seq seq + var length int32 + length = 4 + e.matchlen(s+6, repIndex+4, src) + + seq.matchLen = uint32(length - zstdMinMatch) + + // We might be able to match backwards. + // Extend as long as we can. + start := s + 2 + // We end the search early, so we don't risk 0 literals + // and have to do special offset treatment. + startLimit := nextEmit + 1 + + sMin := s - e.maxMatchOff + if sMin < 0 { + sMin = 0 + } + for repIndex > sMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch { + repIndex-- + start-- + seq.matchLen++ + } + addLiterals(&seq, start) + + // rep 0 + seq.offset = 1 + if debugSequences { + println("repeat sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + s += length + 2 + nextEmit = s + if s >= sLimit { + if debugEncoder { + println("repeat ended", s, length) + + } + break encodeLoop + } + cv = load6432(src, s) + continue + } + coffset0 := s - (candidate.offset - e.cur) + coffset1 := s - (candidate2.offset - e.cur) + 1 + if coffset0 < e.maxMatchOff && uint32(cv) == candidate.val { + // found a regular match + t = candidate.offset - e.cur + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + break + } + + if coffset1 < e.maxMatchOff && uint32(cv>>8) == candidate2.val { + // found a regular match + t = candidate2.offset - e.cur + s++ + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + if debugAsserts && s-t > e.maxMatchOff { + panic("s - t >e.maxMatchOff") + } + if debugAsserts && t < 0 { + panic("t<0") + } + break + } + s += stepSize + ((s - nextEmit) >> (kSearchStrength - 1)) + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + } + // A 4-byte match has been found. We'll later see if more than 4 bytes. + offset2 = offset1 + offset1 = s - t + + if debugAsserts && s <= t { + panic(fmt.Sprintf("s (%d) <= t (%d)", s, t)) + } + + if debugAsserts && canRepeat && int(offset1) > len(src) { + panic("invalid offset") + } + + // Extend the 4-byte match as long as possible. + l := e.matchlen(s+4, t+4, src) + 4 + + // Extend backwards + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { + s-- + t-- + l++ + } + + // Write our sequence. + var seq seq + seq.litLen = uint32(s - nextEmit) + seq.matchLen = uint32(l - zstdMinMatch) + if seq.litLen > 0 { + blk.literals = append(blk.literals, src[nextEmit:s]...) + } + // Don't use repeat offsets + seq.offset = uint32(s-t) + 3 + s += l + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + nextEmit = s + if s >= sLimit { + break encodeLoop + } + cv = load6432(src, s) + + // Check offset 2 + if o2 := s - offset2; canRepeat && load3232(src, o2) == uint32(cv) { + // We have at least 4 byte match. + // No need to check backwards. We come straight from a match + l := 4 + e.matchlen(s+4, o2+4, src) + + // Store this, since we have it. + nextHash := hashLen(cv, hashLog, tableFastHashLen) + e.table[nextHash] = tableEntry{offset: s + e.cur, val: uint32(cv)} + e.markShardDirty(nextHash) + seq.matchLen = uint32(l) - zstdMinMatch + seq.litLen = 0 + // Since litlen is always 0, this is offset 1. + seq.offset = 1 + s += l + nextEmit = s + if debugSequences { + println("sequence", seq, "next s:", s) + } + blk.sequences = append(blk.sequences, seq) + + // Swap offset 1 and 2. + offset1, offset2 = offset2, offset1 + if s >= sLimit { + break encodeLoop + } + // Prepare next loop. + cv = load6432(src, s) + } + } + + if int(nextEmit) < len(src) { + blk.literals = append(blk.literals, src[nextEmit:]...) + blk.extraLits = len(src) - int(nextEmit) + } + blk.recentOffsets[0] = uint32(offset1) + blk.recentOffsets[1] = uint32(offset2) + if debugEncoder { + println("returning, recent offsets:", blk.recentOffsets, "extra literals:", blk.extraLits) + } +} + +// ResetDict will reset and set a dictionary if not nil +func (e *fastEncoder) Reset(d *dict, singleBlock bool) { + e.resetBase(d, singleBlock) + if d != nil { + panic("fastEncoder: Reset with dict") + } +} + +// ResetDict will reset and set a dictionary if not nil +func (e *fastEncoderDict) Reset(d *dict, singleBlock bool) { + e.resetBase(d, singleBlock) + if d == nil { + return + } + + // Init or copy dict table + if len(e.dictTable) != len(e.table) || d.id != e.lastDictID { + if len(e.dictTable) != len(e.table) { + e.dictTable = make([]tableEntry, len(e.table)) + } + if true { + end := e.maxMatchOff + int32(len(d.content)) - 8 + for i := e.maxMatchOff; i < end; i += 3 { + const hashLog = tableBits + + cv := load6432(d.content, i-e.maxMatchOff) + nextHash := hashLen(cv, hashLog, tableFastHashLen) // 0 -> 5 + nextHash1 := hashLen(cv>>8, hashLog, tableFastHashLen) // 1 -> 6 + nextHash2 := hashLen(cv>>16, hashLog, tableFastHashLen) // 2 -> 7 + e.dictTable[nextHash] = tableEntry{ + val: uint32(cv), + offset: i, + } + e.dictTable[nextHash1] = tableEntry{ + val: uint32(cv >> 8), + offset: i + 1, + } + e.dictTable[nextHash2] = tableEntry{ + val: uint32(cv >> 16), + offset: i + 2, + } + } + } + e.lastDictID = d.id + e.allDirty = true + } + + e.cur = e.maxMatchOff + dirtyShardCnt := 0 + if !e.allDirty { + for i := range e.tableShardDirty { + if e.tableShardDirty[i] { + dirtyShardCnt++ + } + } + } + + const shardCnt = tableShardCnt + const shardSize = tableShardSize + if e.allDirty || dirtyShardCnt > shardCnt*4/6 { + copy(e.table[:], e.dictTable) + for i := range e.tableShardDirty { + e.tableShardDirty[i] = false + } + e.allDirty = false + return + } + for i := range e.tableShardDirty { + if !e.tableShardDirty[i] { + continue + } + + copy(e.table[i*shardSize:(i+1)*shardSize], e.dictTable[i*shardSize:(i+1)*shardSize]) + e.tableShardDirty[i] = false + } + e.allDirty = false +} + +func (e *fastEncoderDict) markAllShardsDirty() { + e.allDirty = true +} + +func (e *fastEncoderDict) markShardDirty(entryNum uint32) { + e.tableShardDirty[entryNum/tableShardSize] = true +} diff --git a/vendor/github.com/klauspost/compress/zstd/encoder.go b/vendor/github.com/klauspost/compress/zstd/encoder.go new file mode 100644 index 0000000..dcc987a --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/encoder.go @@ -0,0 +1,641 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "crypto/rand" + "fmt" + "io" + rdebug "runtime/debug" + "sync" + + "github.com/klauspost/compress/zstd/internal/xxhash" +) + +// Encoder provides encoding to Zstandard. +// An Encoder can be used for either compressing a stream via the +// io.WriteCloser interface supported by the Encoder or as multiple independent +// tasks via the EncodeAll function. +// Smaller encodes are encouraged to use the EncodeAll function. +// Use NewWriter to create a new instance. +type Encoder struct { + o encoderOptions + encoders chan encoder + state encoderState + init sync.Once +} + +type encoder interface { + Encode(blk *blockEnc, src []byte) + EncodeNoHist(blk *blockEnc, src []byte) + Block() *blockEnc + CRC() *xxhash.Digest + AppendCRC([]byte) []byte + WindowSize(size int64) int32 + UseBlock(*blockEnc) + Reset(d *dict, singleBlock bool) +} + +type encoderState struct { + w io.Writer + filling []byte + current []byte + previous []byte + encoder encoder + writing *blockEnc + err error + writeErr error + nWritten int64 + nInput int64 + frameContentSize int64 + headerWritten bool + eofWritten bool + fullFrameWritten bool + + // This waitgroup indicates an encode is running. + wg sync.WaitGroup + // This waitgroup indicates we have a block encoding/writing. + wWg sync.WaitGroup +} + +// NewWriter will create a new Zstandard encoder. +// If the encoder will be used for encoding blocks a nil writer can be used. +func NewWriter(w io.Writer, opts ...EOption) (*Encoder, error) { + initPredefined() + var e Encoder + e.o.setDefault() + for _, o := range opts { + err := o(&e.o) + if err != nil { + return nil, err + } + } + if w != nil { + e.Reset(w) + } + return &e, nil +} + +func (e *Encoder) initialize() { + if e.o.concurrent == 0 { + e.o.setDefault() + } + e.encoders = make(chan encoder, e.o.concurrent) + for i := 0; i < e.o.concurrent; i++ { + enc := e.o.encoder() + e.encoders <- enc + } +} + +// Reset will re-initialize the writer and new writes will encode to the supplied writer +// as a new, independent stream. +func (e *Encoder) Reset(w io.Writer) { + s := &e.state + s.wg.Wait() + s.wWg.Wait() + if cap(s.filling) == 0 { + s.filling = make([]byte, 0, e.o.blockSize) + } + if e.o.concurrent > 1 { + if cap(s.current) == 0 { + s.current = make([]byte, 0, e.o.blockSize) + } + if cap(s.previous) == 0 { + s.previous = make([]byte, 0, e.o.blockSize) + } + s.current = s.current[:0] + s.previous = s.previous[:0] + if s.writing == nil { + s.writing = &blockEnc{lowMem: e.o.lowMem} + s.writing.init() + } + s.writing.initNewEncode() + } + if s.encoder == nil { + s.encoder = e.o.encoder() + } + s.filling = s.filling[:0] + s.encoder.Reset(e.o.dict, false) + s.headerWritten = false + s.eofWritten = false + s.fullFrameWritten = false + s.w = w + s.err = nil + s.nWritten = 0 + s.nInput = 0 + s.writeErr = nil + s.frameContentSize = 0 +} + +// ResetContentSize will reset and set a content size for the next stream. +// If the bytes written does not match the size given an error will be returned +// when calling Close(). +// This is removed when Reset is called. +// Sizes <= 0 results in no content size set. +func (e *Encoder) ResetContentSize(w io.Writer, size int64) { + e.Reset(w) + if size >= 0 { + e.state.frameContentSize = size + } +} + +// Write data to the encoder. +// Input data will be buffered and as the buffer fills up +// content will be compressed and written to the output. +// When done writing, use Close to flush the remaining output +// and write CRC if requested. +func (e *Encoder) Write(p []byte) (n int, err error) { + s := &e.state + for len(p) > 0 { + if len(p)+len(s.filling) < e.o.blockSize { + if e.o.crc { + _, _ = s.encoder.CRC().Write(p) + } + s.filling = append(s.filling, p...) + return n + len(p), nil + } + add := p + if len(p)+len(s.filling) > e.o.blockSize { + add = add[:e.o.blockSize-len(s.filling)] + } + if e.o.crc { + _, _ = s.encoder.CRC().Write(add) + } + s.filling = append(s.filling, add...) + p = p[len(add):] + n += len(add) + if len(s.filling) < e.o.blockSize { + return n, nil + } + err := e.nextBlock(false) + if err != nil { + return n, err + } + if debugAsserts && len(s.filling) > 0 { + panic(len(s.filling)) + } + } + return n, nil +} + +// nextBlock will synchronize and start compressing input in e.state.filling. +// If an error has occurred during encoding it will be returned. +func (e *Encoder) nextBlock(final bool) error { + s := &e.state + // Wait for current block. + s.wg.Wait() + if s.err != nil { + return s.err + } + if len(s.filling) > e.o.blockSize { + return fmt.Errorf("block > maxStoreBlockSize") + } + if !s.headerWritten { + // If we have a single block encode, do a sync compression. + if final && len(s.filling) == 0 && !e.o.fullZero { + s.headerWritten = true + s.fullFrameWritten = true + s.eofWritten = true + return nil + } + if final && len(s.filling) > 0 { + s.current = e.EncodeAll(s.filling, s.current[:0]) + var n2 int + n2, s.err = s.w.Write(s.current) + if s.err != nil { + return s.err + } + s.nWritten += int64(n2) + s.nInput += int64(len(s.filling)) + s.current = s.current[:0] + s.filling = s.filling[:0] + s.headerWritten = true + s.fullFrameWritten = true + s.eofWritten = true + return nil + } + + var tmp [maxHeaderSize]byte + fh := frameHeader{ + ContentSize: uint64(s.frameContentSize), + WindowSize: uint32(s.encoder.WindowSize(s.frameContentSize)), + SingleSegment: false, + Checksum: e.o.crc, + DictID: e.o.dict.ID(), + } + + dst, err := fh.appendTo(tmp[:0]) + if err != nil { + return err + } + s.headerWritten = true + s.wWg.Wait() + var n2 int + n2, s.err = s.w.Write(dst) + if s.err != nil { + return s.err + } + s.nWritten += int64(n2) + } + if s.eofWritten { + // Ensure we only write it once. + final = false + } + + if len(s.filling) == 0 { + // Final block, but no data. + if final { + enc := s.encoder + blk := enc.Block() + blk.reset(nil) + blk.last = true + blk.encodeRaw(nil) + s.wWg.Wait() + _, s.err = s.w.Write(blk.output) + s.nWritten += int64(len(blk.output)) + s.eofWritten = true + } + return s.err + } + + // SYNC: + if e.o.concurrent == 1 { + src := s.filling + s.nInput += int64(len(s.filling)) + if debugEncoder { + println("Adding sync block,", len(src), "bytes, final:", final) + } + enc := s.encoder + blk := enc.Block() + blk.reset(nil) + enc.Encode(blk, src) + blk.last = final + if final { + s.eofWritten = true + } + + err := errIncompressible + // If we got the exact same number of literals as input, + // assume the literals cannot be compressed. + if len(src) != len(blk.literals) || len(src) != e.o.blockSize { + err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) + } + switch err { + case errIncompressible: + if debugEncoder { + println("Storing incompressible block as raw") + } + blk.encodeRaw(src) + // In fast mode, we do not transfer offsets, so we don't have to deal with changing the. + case nil: + default: + s.err = err + return err + } + _, s.err = s.w.Write(blk.output) + s.nWritten += int64(len(blk.output)) + s.filling = s.filling[:0] + return s.err + } + + // Move blocks forward. + s.filling, s.current, s.previous = s.previous[:0], s.filling, s.current + s.nInput += int64(len(s.current)) + s.wg.Add(1) + go func(src []byte) { + if debugEncoder { + println("Adding block,", len(src), "bytes, final:", final) + } + defer func() { + if r := recover(); r != nil { + s.err = fmt.Errorf("panic while encoding: %v", r) + rdebug.PrintStack() + } + s.wg.Done() + }() + enc := s.encoder + blk := enc.Block() + enc.Encode(blk, src) + blk.last = final + if final { + s.eofWritten = true + } + // Wait for pending writes. + s.wWg.Wait() + if s.writeErr != nil { + s.err = s.writeErr + return + } + // Transfer encoders from previous write block. + blk.swapEncoders(s.writing) + // Transfer recent offsets to next. + enc.UseBlock(s.writing) + s.writing = blk + s.wWg.Add(1) + go func() { + defer func() { + if r := recover(); r != nil { + s.writeErr = fmt.Errorf("panic while encoding/writing: %v", r) + rdebug.PrintStack() + } + s.wWg.Done() + }() + err := errIncompressible + // If we got the exact same number of literals as input, + // assume the literals cannot be compressed. + if len(src) != len(blk.literals) || len(src) != e.o.blockSize { + err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) + } + switch err { + case errIncompressible: + if debugEncoder { + println("Storing incompressible block as raw") + } + blk.encodeRaw(src) + // In fast mode, we do not transfer offsets, so we don't have to deal with changing the. + case nil: + default: + s.writeErr = err + return + } + _, s.writeErr = s.w.Write(blk.output) + s.nWritten += int64(len(blk.output)) + }() + }(s.current) + return nil +} + +// ReadFrom reads data from r until EOF or error. +// The return value n is the number of bytes read. +// Any error except io.EOF encountered during the read is also returned. +// +// The Copy function uses ReaderFrom if available. +func (e *Encoder) ReadFrom(r io.Reader) (n int64, err error) { + if debugEncoder { + println("Using ReadFrom") + } + + // Flush any current writes. + if len(e.state.filling) > 0 { + if err := e.nextBlock(false); err != nil { + return 0, err + } + } + e.state.filling = e.state.filling[:e.o.blockSize] + src := e.state.filling + for { + n2, err := r.Read(src) + if e.o.crc { + _, _ = e.state.encoder.CRC().Write(src[:n2]) + } + // src is now the unfilled part... + src = src[n2:] + n += int64(n2) + switch err { + case io.EOF: + e.state.filling = e.state.filling[:len(e.state.filling)-len(src)] + if debugEncoder { + println("ReadFrom: got EOF final block:", len(e.state.filling)) + } + return n, nil + case nil: + default: + if debugEncoder { + println("ReadFrom: got error:", err) + } + e.state.err = err + return n, err + } + if len(src) > 0 { + if debugEncoder { + println("ReadFrom: got space left in source:", len(src)) + } + continue + } + err = e.nextBlock(false) + if err != nil { + return n, err + } + e.state.filling = e.state.filling[:e.o.blockSize] + src = e.state.filling + } +} + +// Flush will send the currently written data to output +// and block until everything has been written. +// This should only be used on rare occasions where pushing the currently queued data is critical. +func (e *Encoder) Flush() error { + s := &e.state + if len(s.filling) > 0 { + err := e.nextBlock(false) + if err != nil { + return err + } + } + s.wg.Wait() + s.wWg.Wait() + if s.err != nil { + return s.err + } + return s.writeErr +} + +// Close will flush the final output and close the stream. +// The function will block until everything has been written. +// The Encoder can still be re-used after calling this. +func (e *Encoder) Close() error { + s := &e.state + if s.encoder == nil { + return nil + } + err := e.nextBlock(true) + if err != nil { + return err + } + if s.frameContentSize > 0 { + if s.nInput != s.frameContentSize { + return fmt.Errorf("frame content size %d given, but %d bytes was written", s.frameContentSize, s.nInput) + } + } + if e.state.fullFrameWritten { + return s.err + } + s.wg.Wait() + s.wWg.Wait() + + if s.err != nil { + return s.err + } + if s.writeErr != nil { + return s.writeErr + } + + // Write CRC + if e.o.crc && s.err == nil { + // heap alloc. + var tmp [4]byte + _, s.err = s.w.Write(s.encoder.AppendCRC(tmp[:0])) + s.nWritten += 4 + } + + // Add padding with content from crypto/rand.Reader + if s.err == nil && e.o.pad > 0 { + add := calcSkippableFrame(s.nWritten, int64(e.o.pad)) + frame, err := skippableFrame(s.filling[:0], add, rand.Reader) + if err != nil { + return err + } + _, s.err = s.w.Write(frame) + } + return s.err +} + +// EncodeAll will encode all input in src and append it to dst. +// This function can be called concurrently, but each call will only run on a single goroutine. +// If empty input is given, nothing is returned, unless WithZeroFrames is specified. +// Encoded blocks can be concatenated and the result will be the combined input stream. +// Data compressed with EncodeAll can be decoded with the Decoder, +// using either a stream or DecodeAll. +func (e *Encoder) EncodeAll(src, dst []byte) []byte { + if len(src) == 0 { + if e.o.fullZero { + // Add frame header. + fh := frameHeader{ + ContentSize: 0, + WindowSize: MinWindowSize, + SingleSegment: true, + // Adding a checksum would be a waste of space. + Checksum: false, + DictID: 0, + } + dst, _ = fh.appendTo(dst) + + // Write raw block as last one only. + var blk blockHeader + blk.setSize(0) + blk.setType(blockTypeRaw) + blk.setLast(true) + dst = blk.appendTo(dst) + } + return dst + } + e.init.Do(e.initialize) + enc := <-e.encoders + defer func() { + // Release encoder reference to last block. + // If a non-single block is needed the encoder will reset again. + e.encoders <- enc + }() + // Use single segments when above minimum window and below 1MB. + single := len(src) < 1<<20 && len(src) > MinWindowSize + if e.o.single != nil { + single = *e.o.single + } + fh := frameHeader{ + ContentSize: uint64(len(src)), + WindowSize: uint32(enc.WindowSize(int64(len(src)))), + SingleSegment: single, + Checksum: e.o.crc, + DictID: e.o.dict.ID(), + } + + // If less than 1MB, allocate a buffer up front. + if len(dst) == 0 && cap(dst) == 0 && len(src) < 1<<20 && !e.o.lowMem { + dst = make([]byte, 0, len(src)) + } + dst, err := fh.appendTo(dst) + if err != nil { + panic(err) + } + + // If we can do everything in one block, prefer that. + if len(src) <= maxCompressedBlockSize { + enc.Reset(e.o.dict, true) + // Slightly faster with no history and everything in one block. + if e.o.crc { + _, _ = enc.CRC().Write(src) + } + blk := enc.Block() + blk.last = true + if e.o.dict == nil { + enc.EncodeNoHist(blk, src) + } else { + enc.Encode(blk, src) + } + + // If we got the exact same number of literals as input, + // assume the literals cannot be compressed. + err := errIncompressible + oldout := blk.output + if len(blk.literals) != len(src) || len(src) != e.o.blockSize { + // Output directly to dst + blk.output = dst + err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) + } + + switch err { + case errIncompressible: + if debugEncoder { + println("Storing incompressible block as raw") + } + dst = blk.encodeRawTo(dst, src) + case nil: + dst = blk.output + default: + panic(err) + } + blk.output = oldout + } else { + enc.Reset(e.o.dict, false) + blk := enc.Block() + for len(src) > 0 { + todo := src + if len(todo) > e.o.blockSize { + todo = todo[:e.o.blockSize] + } + src = src[len(todo):] + if e.o.crc { + _, _ = enc.CRC().Write(todo) + } + blk.pushOffsets() + enc.Encode(blk, todo) + if len(src) == 0 { + blk.last = true + } + err := errIncompressible + // If we got the exact same number of literals as input, + // assume the literals cannot be compressed. + if len(blk.literals) != len(todo) || len(todo) != e.o.blockSize { + err = blk.encode(todo, e.o.noEntropy, !e.o.allLitEntropy) + } + + switch err { + case errIncompressible: + if debugEncoder { + println("Storing incompressible block as raw") + } + dst = blk.encodeRawTo(dst, todo) + blk.popOffsets() + case nil: + dst = append(dst, blk.output...) + default: + panic(err) + } + blk.reset(nil) + } + } + if e.o.crc { + dst = enc.AppendCRC(dst) + } + // Add padding with content from crypto/rand.Reader + if e.o.pad > 0 { + add := calcSkippableFrame(int64(len(dst)), int64(e.o.pad)) + dst, err = skippableFrame(dst, add, rand.Reader) + if err != nil { + panic(err) + } + } + return dst +} diff --git a/vendor/github.com/klauspost/compress/zstd/encoder_options.go b/vendor/github.com/klauspost/compress/zstd/encoder_options.go new file mode 100644 index 0000000..44d8dbd --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/encoder_options.go @@ -0,0 +1,317 @@ +package zstd + +import ( + "errors" + "fmt" + "runtime" + "strings" +) + +// EOption is an option for creating a encoder. +type EOption func(*encoderOptions) error + +// options retains accumulated state of multiple options. +type encoderOptions struct { + concurrent int + level EncoderLevel + single *bool + pad int + blockSize int + windowSize int + crc bool + fullZero bool + noEntropy bool + allLitEntropy bool + customWindow bool + customALEntropy bool + customBlockSize bool + lowMem bool + dict *dict +} + +func (o *encoderOptions) setDefault() { + *o = encoderOptions{ + concurrent: runtime.GOMAXPROCS(0), + crc: true, + single: nil, + blockSize: maxCompressedBlockSize, + windowSize: 8 << 20, + level: SpeedDefault, + allLitEntropy: true, + lowMem: false, + } +} + +// encoder returns an encoder with the selected options. +func (o encoderOptions) encoder() encoder { + switch o.level { + case SpeedFastest: + if o.dict != nil { + return &fastEncoderDict{fastEncoder: fastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}}} + } + return &fastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}} + + case SpeedDefault: + if o.dict != nil { + return &doubleFastEncoderDict{fastEncoderDict: fastEncoderDict{fastEncoder: fastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}}}} + } + return &doubleFastEncoder{fastEncoder: fastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}}} + case SpeedBetterCompression: + if o.dict != nil { + return &betterFastEncoderDict{betterFastEncoder: betterFastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}}} + } + return &betterFastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}} + case SpeedBestCompression: + return &bestFastEncoder{fastBase: fastBase{maxMatchOff: int32(o.windowSize), lowMem: o.lowMem}} + } + panic("unknown compression level") +} + +// WithEncoderCRC will add CRC value to output. +// Output will be 4 bytes larger. +func WithEncoderCRC(b bool) EOption { + return func(o *encoderOptions) error { o.crc = b; return nil } +} + +// WithEncoderConcurrency will set the concurrency, +// meaning the maximum number of encoders to run concurrently. +// The value supplied must be at least 1. +// For streams, setting a value of 1 will disable async compression. +// By default this will be set to GOMAXPROCS. +func WithEncoderConcurrency(n int) EOption { + return func(o *encoderOptions) error { + if n <= 0 { + return fmt.Errorf("concurrency must be at least 1") + } + o.concurrent = n + return nil + } +} + +// WithWindowSize will set the maximum allowed back-reference distance. +// The value must be a power of two between MinWindowSize and MaxWindowSize. +// A larger value will enable better compression but allocate more memory and, +// for above-default values, take considerably longer. +// The default value is determined by the compression level. +func WithWindowSize(n int) EOption { + return func(o *encoderOptions) error { + switch { + case n < MinWindowSize: + return fmt.Errorf("window size must be at least %d", MinWindowSize) + case n > MaxWindowSize: + return fmt.Errorf("window size must be at most %d", MaxWindowSize) + case (n & (n - 1)) != 0: + return errors.New("window size must be a power of 2") + } + + o.windowSize = n + o.customWindow = true + if o.blockSize > o.windowSize { + o.blockSize = o.windowSize + o.customBlockSize = true + } + return nil + } +} + +// WithEncoderPadding will add padding to all output so the size will be a multiple of n. +// This can be used to obfuscate the exact output size or make blocks of a certain size. +// The contents will be a skippable frame, so it will be invisible by the decoder. +// n must be > 0 and <= 1GB, 1<<30 bytes. +// The padded area will be filled with data from crypto/rand.Reader. +// If `EncodeAll` is used with data already in the destination, the total size will be multiple of this. +func WithEncoderPadding(n int) EOption { + return func(o *encoderOptions) error { + if n <= 0 { + return fmt.Errorf("padding must be at least 1") + } + // No need to waste our time. + if n == 1 { + o.pad = 0 + } + if n > 1<<30 { + return fmt.Errorf("padding must less than 1GB (1<<30 bytes) ") + } + o.pad = n + return nil + } +} + +// EncoderLevel predefines encoder compression levels. +// Only use the constants made available, since the actual mapping +// of these values are very likely to change and your compression could change +// unpredictably when upgrading the library. +type EncoderLevel int + +const ( + speedNotSet EncoderLevel = iota + + // SpeedFastest will choose the fastest reasonable compression. + // This is roughly equivalent to the fastest Zstandard mode. + SpeedFastest + + // SpeedDefault is the default "pretty fast" compression option. + // This is roughly equivalent to the default Zstandard mode (level 3). + SpeedDefault + + // SpeedBetterCompression will yield better compression than the default. + // Currently it is about zstd level 7-8 with ~ 2x-3x the default CPU usage. + // By using this, notice that CPU usage may go up in the future. + SpeedBetterCompression + + // SpeedBestCompression will choose the best available compression option. + // This will offer the best compression no matter the CPU cost. + SpeedBestCompression + + // speedLast should be kept as the last actual compression option. + // The is not for external usage, but is used to keep track of the valid options. + speedLast +) + +// EncoderLevelFromString will convert a string representation of an encoding level back +// to a compression level. The compare is not case sensitive. +// If the string wasn't recognized, (false, SpeedDefault) will be returned. +func EncoderLevelFromString(s string) (bool, EncoderLevel) { + for l := speedNotSet + 1; l < speedLast; l++ { + if strings.EqualFold(s, l.String()) { + return true, l + } + } + return false, SpeedDefault +} + +// EncoderLevelFromZstd will return an encoder level that closest matches the compression +// ratio of a specific zstd compression level. +// Many input values will provide the same compression level. +func EncoderLevelFromZstd(level int) EncoderLevel { + switch { + case level < 3: + return SpeedFastest + case level >= 3 && level < 6: + return SpeedDefault + case level >= 6 && level < 10: + return SpeedBetterCompression + default: + return SpeedBestCompression + } +} + +// String provides a string representation of the compression level. +func (e EncoderLevel) String() string { + switch e { + case SpeedFastest: + return "fastest" + case SpeedDefault: + return "default" + case SpeedBetterCompression: + return "better" + case SpeedBestCompression: + return "best" + default: + return "invalid" + } +} + +// WithEncoderLevel specifies a predefined compression level. +func WithEncoderLevel(l EncoderLevel) EOption { + return func(o *encoderOptions) error { + switch { + case l <= speedNotSet || l >= speedLast: + return fmt.Errorf("unknown encoder level") + } + o.level = l + if !o.customWindow { + switch o.level { + case SpeedFastest: + o.windowSize = 4 << 20 + if !o.customBlockSize { + o.blockSize = 1 << 16 + } + case SpeedDefault: + o.windowSize = 8 << 20 + case SpeedBetterCompression: + o.windowSize = 16 << 20 + case SpeedBestCompression: + o.windowSize = 32 << 20 + } + } + if !o.customALEntropy { + o.allLitEntropy = l > SpeedFastest + } + + return nil + } +} + +// WithZeroFrames will encode 0 length input as full frames. +// This can be needed for compatibility with zstandard usage, +// but is not needed for this package. +func WithZeroFrames(b bool) EOption { + return func(o *encoderOptions) error { + o.fullZero = b + return nil + } +} + +// WithAllLitEntropyCompression will apply entropy compression if no matches are found. +// Disabling this will skip incompressible data faster, but in cases with no matches but +// skewed character distribution compression is lost. +// Default value depends on the compression level selected. +func WithAllLitEntropyCompression(b bool) EOption { + return func(o *encoderOptions) error { + o.customALEntropy = true + o.allLitEntropy = b + return nil + } +} + +// WithNoEntropyCompression will always skip entropy compression of literals. +// This can be useful if content has matches, but unlikely to benefit from entropy +// compression. Usually the slight speed improvement is not worth enabling this. +func WithNoEntropyCompression(b bool) EOption { + return func(o *encoderOptions) error { + o.noEntropy = b + return nil + } +} + +// WithSingleSegment will set the "single segment" flag when EncodeAll is used. +// If this flag is set, data must be regenerated within a single continuous memory segment. +// In this case, Window_Descriptor byte is skipped, but Frame_Content_Size is necessarily present. +// As a consequence, the decoder must allocate a memory segment of size equal or larger than size of your content. +// In order to preserve the decoder from unreasonable memory requirements, +// a decoder is allowed to reject a compressed frame which requests a memory size beyond decoder's authorized range. +// For broader compatibility, decoders are recommended to support memory sizes of at least 8 MB. +// This is only a recommendation, each decoder is free to support higher or lower limits, depending on local limitations. +// If this is not specified, block encodes will automatically choose this based on the input size. +// This setting has no effect on streamed encodes. +func WithSingleSegment(b bool) EOption { + return func(o *encoderOptions) error { + o.single = &b + return nil + } +} + +// WithLowerEncoderMem will trade in some memory cases trade less memory usage for +// slower encoding speed. +// This will not change the window size which is the primary function for reducing +// memory usage. See WithWindowSize. +func WithLowerEncoderMem(b bool) EOption { + return func(o *encoderOptions) error { + o.lowMem = b + return nil + } +} + +// WithEncoderDict allows to register a dictionary that will be used for the encode. +// The encoder *may* choose to use no dictionary instead for certain payloads. +func WithEncoderDict(dict []byte) EOption { + return func(o *encoderOptions) error { + d, err := loadDict(dict) + if err != nil { + return err + } + o.dict = d + return nil + } +} diff --git a/vendor/github.com/klauspost/compress/zstd/framedec.go b/vendor/github.com/klauspost/compress/zstd/framedec.go new file mode 100644 index 0000000..11089d2 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/framedec.go @@ -0,0 +1,376 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "bytes" + "encoding/hex" + "errors" + "io" + + "github.com/klauspost/compress/zstd/internal/xxhash" +) + +type frameDec struct { + o decoderOptions + crc *xxhash.Digest + + WindowSize uint64 + + // Frame history passed between blocks + history history + + rawInput byteBuffer + + // Byte buffer that can be reused for small input blocks. + bBuf byteBuf + + FrameContentSize uint64 + + DictionaryID *uint32 + HasCheckSum bool + SingleSegment bool +} + +const ( + // MinWindowSize is the minimum Window Size, which is 1 KB. + MinWindowSize = 1 << 10 + + // MaxWindowSize is the maximum encoder window size + // and the default decoder maximum window size. + MaxWindowSize = 1 << 29 +) + +var ( + frameMagic = []byte{0x28, 0xb5, 0x2f, 0xfd} + skippableFrameMagic = []byte{0x2a, 0x4d, 0x18} +) + +func newFrameDec(o decoderOptions) *frameDec { + if o.maxWindowSize > o.maxDecodedSize { + o.maxWindowSize = o.maxDecodedSize + } + d := frameDec{ + o: o, + } + return &d +} + +// reset will read the frame header and prepare for block decoding. +// If nothing can be read from the input, io.EOF will be returned. +// Any other error indicated that the stream contained data, but +// there was a problem. +func (d *frameDec) reset(br byteBuffer) error { + d.HasCheckSum = false + d.WindowSize = 0 + var signature [4]byte + for { + var err error + // Check if we can read more... + b, err := br.readSmall(1) + switch err { + case io.EOF, io.ErrUnexpectedEOF: + return io.EOF + default: + return err + case nil: + signature[0] = b[0] + } + // Read the rest, don't allow io.ErrUnexpectedEOF + b, err = br.readSmall(3) + switch err { + case io.EOF: + return io.EOF + default: + return err + case nil: + copy(signature[1:], b) + } + + if !bytes.Equal(signature[1:4], skippableFrameMagic) || signature[0]&0xf0 != 0x50 { + if debugDecoder { + println("Not skippable", hex.EncodeToString(signature[:]), hex.EncodeToString(skippableFrameMagic)) + } + // Break if not skippable frame. + break + } + // Read size to skip + b, err = br.readSmall(4) + if err != nil { + if debugDecoder { + println("Reading Frame Size", err) + } + return err + } + n := uint32(b[0]) | (uint32(b[1]) << 8) | (uint32(b[2]) << 16) | (uint32(b[3]) << 24) + println("Skipping frame with", n, "bytes.") + err = br.skipN(int(n)) + if err != nil { + if debugDecoder { + println("Reading discarded frame", err) + } + return err + } + } + if !bytes.Equal(signature[:], frameMagic) { + if debugDecoder { + println("Got magic numbers: ", signature, "want:", frameMagic) + } + return ErrMagicMismatch + } + + // Read Frame_Header_Descriptor + fhd, err := br.readByte() + if err != nil { + if debugDecoder { + println("Reading Frame_Header_Descriptor", err) + } + return err + } + d.SingleSegment = fhd&(1<<5) != 0 + + if fhd&(1<<3) != 0 { + return errors.New("reserved bit set on frame header") + } + + // Read Window_Descriptor + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#window_descriptor + d.WindowSize = 0 + if !d.SingleSegment { + wd, err := br.readByte() + if err != nil { + if debugDecoder { + println("Reading Window_Descriptor", err) + } + return err + } + printf("raw: %x, mantissa: %d, exponent: %d\n", wd, wd&7, wd>>3) + windowLog := 10 + (wd >> 3) + windowBase := uint64(1) << windowLog + windowAdd := (windowBase / 8) * uint64(wd&0x7) + d.WindowSize = windowBase + windowAdd + } + + // Read Dictionary_ID + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#dictionary_id + d.DictionaryID = nil + if size := fhd & 3; size != 0 { + if size == 3 { + size = 4 + } + + b, err := br.readSmall(int(size)) + if err != nil { + println("Reading Dictionary_ID", err) + return err + } + var id uint32 + switch size { + case 1: + id = uint32(b[0]) + case 2: + id = uint32(b[0]) | (uint32(b[1]) << 8) + case 4: + id = uint32(b[0]) | (uint32(b[1]) << 8) | (uint32(b[2]) << 16) | (uint32(b[3]) << 24) + } + if debugDecoder { + println("Dict size", size, "ID:", id) + } + if id > 0 { + // ID 0 means "sorry, no dictionary anyway". + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#dictionary-format + d.DictionaryID = &id + } + } + + // Read Frame_Content_Size + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#frame_content_size + var fcsSize int + v := fhd >> 6 + switch v { + case 0: + if d.SingleSegment { + fcsSize = 1 + } + default: + fcsSize = 1 << v + } + d.FrameContentSize = fcsUnknown + if fcsSize > 0 { + b, err := br.readSmall(fcsSize) + if err != nil { + println("Reading Frame content", err) + return err + } + switch fcsSize { + case 1: + d.FrameContentSize = uint64(b[0]) + case 2: + // When FCS_Field_Size is 2, the offset of 256 is added. + d.FrameContentSize = uint64(b[0]) | (uint64(b[1]) << 8) + 256 + case 4: + d.FrameContentSize = uint64(b[0]) | (uint64(b[1]) << 8) | (uint64(b[2]) << 16) | (uint64(b[3]) << 24) + case 8: + d1 := uint32(b[0]) | (uint32(b[1]) << 8) | (uint32(b[2]) << 16) | (uint32(b[3]) << 24) + d2 := uint32(b[4]) | (uint32(b[5]) << 8) | (uint32(b[6]) << 16) | (uint32(b[7]) << 24) + d.FrameContentSize = uint64(d1) | (uint64(d2) << 32) + } + if debugDecoder { + println("Read FCS:", d.FrameContentSize) + } + } + + // Move this to shared. + d.HasCheckSum = fhd&(1<<2) != 0 + if d.HasCheckSum { + if d.crc == nil { + d.crc = xxhash.New() + } + d.crc.Reset() + } + + if d.WindowSize == 0 && d.SingleSegment { + // We may not need window in this case. + d.WindowSize = d.FrameContentSize + if d.WindowSize < MinWindowSize { + d.WindowSize = MinWindowSize + } + } + + if d.WindowSize > uint64(d.o.maxWindowSize) { + if debugDecoder { + printf("window size %d > max %d\n", d.WindowSize, d.o.maxWindowSize) + } + return ErrWindowSizeExceeded + } + // The minimum Window_Size is 1 KB. + if d.WindowSize < MinWindowSize { + if debugDecoder { + println("got window size: ", d.WindowSize) + } + return ErrWindowSizeTooSmall + } + d.history.windowSize = int(d.WindowSize) + if d.o.lowMem && d.history.windowSize < maxBlockSize { + d.history.allocFrameBuffer = d.history.windowSize * 2 + // TODO: Maybe use FrameContent size + } else { + d.history.allocFrameBuffer = d.history.windowSize + maxBlockSize + } + + if debugDecoder { + println("Frame: Dict:", d.DictionaryID, "FrameContentSize:", d.FrameContentSize, "singleseg:", d.SingleSegment, "window:", d.WindowSize, "crc:", d.HasCheckSum) + } + + // history contains input - maybe we do something + d.rawInput = br + return nil +} + +// next will start decoding the next block from stream. +func (d *frameDec) next(block *blockDec) error { + if debugDecoder { + println("decoding new block") + } + err := block.reset(d.rawInput, d.WindowSize) + if err != nil { + println("block error:", err) + // Signal the frame decoder we have a problem. + block.sendErr(err) + return err + } + return nil +} + +// checkCRC will check the checksum if the frame has one. +// Will return ErrCRCMismatch if crc check failed, otherwise nil. +func (d *frameDec) checkCRC() error { + if !d.HasCheckSum { + return nil + } + var tmp [4]byte + got := d.crc.Sum64() + // Flip to match file order. + tmp[0] = byte(got >> 0) + tmp[1] = byte(got >> 8) + tmp[2] = byte(got >> 16) + tmp[3] = byte(got >> 24) + + // We can overwrite upper tmp now + want, err := d.rawInput.readSmall(4) + if err != nil { + println("CRC missing?", err) + return err + } + + if !bytes.Equal(tmp[:], want) && !ignoreCRC { + if debugDecoder { + println("CRC Check Failed:", tmp[:], "!=", want) + } + return ErrCRCMismatch + } + if debugDecoder { + println("CRC ok", tmp[:]) + } + return nil +} + +// runDecoder will create a sync decoder that will decode a block of data. +func (d *frameDec) runDecoder(dst []byte, dec *blockDec) ([]byte, error) { + saved := d.history.b + + // We use the history for output to avoid copying it. + d.history.b = dst + d.history.ignoreBuffer = len(dst) + // Store input length, so we only check new data. + crcStart := len(dst) + var err error + for { + err = dec.reset(d.rawInput, d.WindowSize) + if err != nil { + break + } + if debugDecoder { + println("next block:", dec) + } + err = dec.decodeBuf(&d.history) + if err != nil { + break + } + if uint64(len(d.history.b)) > d.o.maxDecodedSize { + err = ErrDecoderSizeExceeded + break + } + if uint64(len(d.history.b)-crcStart) > d.FrameContentSize { + println("runDecoder: FrameContentSize exceeded", uint64(len(d.history.b)-crcStart), ">", d.FrameContentSize) + err = ErrFrameSizeExceeded + break + } + if dec.Last { + break + } + if debugDecoder { + println("runDecoder: FrameContentSize", uint64(len(d.history.b)-crcStart), "<=", d.FrameContentSize) + } + } + dst = d.history.b + if err == nil { + if d.FrameContentSize != fcsUnknown && uint64(len(d.history.b)-crcStart) != d.FrameContentSize { + err = ErrFrameSizeMismatch + } else if d.HasCheckSum { + var n int + n, err = d.crc.Write(dst[crcStart:]) + if err == nil { + if n != len(dst)-crcStart { + err = io.ErrShortWrite + } else { + err = d.checkCRC() + } + } + } + } + d.history.b = saved + return dst, err +} diff --git a/vendor/github.com/klauspost/compress/zstd/frameenc.go b/vendor/github.com/klauspost/compress/zstd/frameenc.go new file mode 100644 index 0000000..4ef7f5a --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/frameenc.go @@ -0,0 +1,137 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "encoding/binary" + "fmt" + "io" + "math" + "math/bits" +) + +type frameHeader struct { + ContentSize uint64 + WindowSize uint32 + SingleSegment bool + Checksum bool + DictID uint32 +} + +const maxHeaderSize = 14 + +func (f frameHeader) appendTo(dst []byte) ([]byte, error) { + dst = append(dst, frameMagic...) + var fhd uint8 + if f.Checksum { + fhd |= 1 << 2 + } + if f.SingleSegment { + fhd |= 1 << 5 + } + + var dictIDContent []byte + if f.DictID > 0 { + var tmp [4]byte + if f.DictID < 256 { + fhd |= 1 + tmp[0] = uint8(f.DictID) + dictIDContent = tmp[:1] + } else if f.DictID < 1<<16 { + fhd |= 2 + binary.LittleEndian.PutUint16(tmp[:2], uint16(f.DictID)) + dictIDContent = tmp[:2] + } else { + fhd |= 3 + binary.LittleEndian.PutUint32(tmp[:4], f.DictID) + dictIDContent = tmp[:4] + } + } + var fcs uint8 + if f.ContentSize >= 256 { + fcs++ + } + if f.ContentSize >= 65536+256 { + fcs++ + } + if f.ContentSize >= 0xffffffff { + fcs++ + } + + fhd |= fcs << 6 + + dst = append(dst, fhd) + if !f.SingleSegment { + const winLogMin = 10 + windowLog := (bits.Len32(f.WindowSize-1) - winLogMin) << 3 + dst = append(dst, uint8(windowLog)) + } + if f.DictID > 0 { + dst = append(dst, dictIDContent...) + } + switch fcs { + case 0: + if f.SingleSegment { + dst = append(dst, uint8(f.ContentSize)) + } + // Unless SingleSegment is set, framessizes < 256 are nto stored. + case 1: + f.ContentSize -= 256 + dst = append(dst, uint8(f.ContentSize), uint8(f.ContentSize>>8)) + case 2: + dst = append(dst, uint8(f.ContentSize), uint8(f.ContentSize>>8), uint8(f.ContentSize>>16), uint8(f.ContentSize>>24)) + case 3: + dst = append(dst, uint8(f.ContentSize), uint8(f.ContentSize>>8), uint8(f.ContentSize>>16), uint8(f.ContentSize>>24), + uint8(f.ContentSize>>32), uint8(f.ContentSize>>40), uint8(f.ContentSize>>48), uint8(f.ContentSize>>56)) + default: + panic("invalid fcs") + } + return dst, nil +} + +const skippableFrameHeader = 4 + 4 + +// calcSkippableFrame will return a total size to be added for written +// to be divisible by multiple. +// The value will always be > skippableFrameHeader. +// The function will panic if written < 0 or wantMultiple <= 0. +func calcSkippableFrame(written, wantMultiple int64) int { + if wantMultiple <= 0 { + panic("wantMultiple <= 0") + } + if written < 0 { + panic("written < 0") + } + leftOver := written % wantMultiple + if leftOver == 0 { + return 0 + } + toAdd := wantMultiple - leftOver + for toAdd < skippableFrameHeader { + toAdd += wantMultiple + } + return int(toAdd) +} + +// skippableFrame will add a skippable frame with a total size of bytes. +// total should be >= skippableFrameHeader and < math.MaxUint32. +func skippableFrame(dst []byte, total int, r io.Reader) ([]byte, error) { + if total == 0 { + return dst, nil + } + if total < skippableFrameHeader { + return dst, fmt.Errorf("requested skippable frame (%d) < 8", total) + } + if int64(total) > math.MaxUint32 { + return dst, fmt.Errorf("requested skippable frame (%d) > max uint32", total) + } + dst = append(dst, 0x50, 0x2a, 0x4d, 0x18) + f := uint32(total - skippableFrameHeader) + dst = append(dst, uint8(f), uint8(f>>8), uint8(f>>16), uint8(f>>24)) + start := len(dst) + dst = append(dst, make([]byte, f)...) + _, err := io.ReadFull(r, dst[start:]) + return dst, err +} diff --git a/vendor/github.com/klauspost/compress/zstd/fse_decoder.go b/vendor/github.com/klauspost/compress/zstd/fse_decoder.go new file mode 100644 index 0000000..bb3d4fd --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/fse_decoder.go @@ -0,0 +1,385 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "errors" + "fmt" +) + +const ( + tablelogAbsoluteMax = 9 +) + +const ( + /*!MEMORY_USAGE : + * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.) + * Increasing memory usage improves compression ratio + * Reduced memory usage can improve speed, due to cache effect + * Recommended max value is 14, for 16KB, which nicely fits into Intel x86 L1 cache */ + maxMemoryUsage = tablelogAbsoluteMax + 2 + + maxTableLog = maxMemoryUsage - 2 + maxTablesize = 1 << maxTableLog + maxTableMask = (1 << maxTableLog) - 1 + minTablelog = 5 + maxSymbolValue = 255 +) + +// fseDecoder provides temporary storage for compression and decompression. +type fseDecoder struct { + dt [maxTablesize]decSymbol // Decompression table. + symbolLen uint16 // Length of active part of the symbol table. + actualTableLog uint8 // Selected tablelog. + maxBits uint8 // Maximum number of additional bits + + // used for table creation to avoid allocations. + stateTable [256]uint16 + norm [maxSymbolValue + 1]int16 + preDefined bool +} + +// tableStep returns the next table index. +func tableStep(tableSize uint32) uint32 { + return (tableSize >> 1) + (tableSize >> 3) + 3 +} + +// readNCount will read the symbol distribution so decoding tables can be constructed. +func (s *fseDecoder) readNCount(b *byteReader, maxSymbol uint16) error { + var ( + charnum uint16 + previous0 bool + ) + if b.remain() < 4 { + return errors.New("input too small") + } + bitStream := b.Uint32NC() + nbBits := uint((bitStream & 0xF) + minTablelog) // extract tableLog + if nbBits > tablelogAbsoluteMax { + println("Invalid tablelog:", nbBits) + return errors.New("tableLog too large") + } + bitStream >>= 4 + bitCount := uint(4) + + s.actualTableLog = uint8(nbBits) + remaining := int32((1 << nbBits) + 1) + threshold := int32(1 << nbBits) + gotTotal := int32(0) + nbBits++ + + for remaining > 1 && charnum <= maxSymbol { + if previous0 { + //println("prev0") + n0 := charnum + for (bitStream & 0xFFFF) == 0xFFFF { + //println("24 x 0") + n0 += 24 + if r := b.remain(); r > 5 { + b.advance(2) + // The check above should make sure we can read 32 bits + bitStream = b.Uint32NC() >> bitCount + } else { + // end of bit stream + bitStream >>= 16 + bitCount += 16 + } + } + //printf("bitstream: %d, 0b%b", bitStream&3, bitStream) + for (bitStream & 3) == 3 { + n0 += 3 + bitStream >>= 2 + bitCount += 2 + } + n0 += uint16(bitStream & 3) + bitCount += 2 + + if n0 > maxSymbolValue { + return errors.New("maxSymbolValue too small") + } + //println("inserting ", n0-charnum, "zeroes from idx", charnum, "ending before", n0) + for charnum < n0 { + s.norm[uint8(charnum)] = 0 + charnum++ + } + + if r := b.remain(); r >= 7 || r-int(bitCount>>3) >= 4 { + b.advance(bitCount >> 3) + bitCount &= 7 + // The check above should make sure we can read 32 bits + bitStream = b.Uint32NC() >> bitCount + } else { + bitStream >>= 2 + } + } + + max := (2*threshold - 1) - remaining + var count int32 + + if int32(bitStream)&(threshold-1) < max { + count = int32(bitStream) & (threshold - 1) + if debugAsserts && nbBits < 1 { + panic("nbBits underflow") + } + bitCount += nbBits - 1 + } else { + count = int32(bitStream) & (2*threshold - 1) + if count >= threshold { + count -= max + } + bitCount += nbBits + } + + // extra accuracy + count-- + if count < 0 { + // -1 means +1 + remaining += count + gotTotal -= count + } else { + remaining -= count + gotTotal += count + } + s.norm[charnum&0xff] = int16(count) + charnum++ + previous0 = count == 0 + for remaining < threshold { + nbBits-- + threshold >>= 1 + } + + if r := b.remain(); r >= 7 || r-int(bitCount>>3) >= 4 { + b.advance(bitCount >> 3) + bitCount &= 7 + // The check above should make sure we can read 32 bits + bitStream = b.Uint32NC() >> (bitCount & 31) + } else { + bitCount -= (uint)(8 * (len(b.b) - 4 - b.off)) + b.off = len(b.b) - 4 + bitStream = b.Uint32() >> (bitCount & 31) + } + } + s.symbolLen = charnum + if s.symbolLen <= 1 { + return fmt.Errorf("symbolLen (%d) too small", s.symbolLen) + } + if s.symbolLen > maxSymbolValue+1 { + return fmt.Errorf("symbolLen (%d) too big", s.symbolLen) + } + if remaining != 1 { + return fmt.Errorf("corruption detected (remaining %d != 1)", remaining) + } + if bitCount > 32 { + return fmt.Errorf("corruption detected (bitCount %d > 32)", bitCount) + } + if gotTotal != 1<> 3) + // println(s.norm[:s.symbolLen], s.symbolLen) + return s.buildDtable() +} + +// decSymbol contains information about a state entry, +// Including the state offset base, the output symbol and +// the number of bits to read for the low part of the destination state. +// Using a composite uint64 is faster than a struct with separate members. +type decSymbol uint64 + +func newDecSymbol(nbits, addBits uint8, newState uint16, baseline uint32) decSymbol { + return decSymbol(nbits) | (decSymbol(addBits) << 8) | (decSymbol(newState) << 16) | (decSymbol(baseline) << 32) +} + +func (d decSymbol) nbBits() uint8 { + return uint8(d) +} + +func (d decSymbol) addBits() uint8 { + return uint8(d >> 8) +} + +func (d decSymbol) newState() uint16 { + return uint16(d >> 16) +} + +func (d decSymbol) baseline() uint32 { + return uint32(d >> 32) +} + +func (d decSymbol) baselineInt() int { + return int(d >> 32) +} + +func (d *decSymbol) set(nbits, addBits uint8, newState uint16, baseline uint32) { + *d = decSymbol(nbits) | (decSymbol(addBits) << 8) | (decSymbol(newState) << 16) | (decSymbol(baseline) << 32) +} + +func (d *decSymbol) setNBits(nBits uint8) { + const mask = 0xffffffffffffff00 + *d = (*d & mask) | decSymbol(nBits) +} + +func (d *decSymbol) setAddBits(addBits uint8) { + const mask = 0xffffffffffff00ff + *d = (*d & mask) | (decSymbol(addBits) << 8) +} + +func (d *decSymbol) setNewState(state uint16) { + const mask = 0xffffffff0000ffff + *d = (*d & mask) | decSymbol(state)<<16 +} + +func (d *decSymbol) setBaseline(baseline uint32) { + const mask = 0xffffffff + *d = (*d & mask) | decSymbol(baseline)<<32 +} + +func (d *decSymbol) setExt(addBits uint8, baseline uint32) { + const mask = 0xffff00ff + *d = (*d & mask) | (decSymbol(addBits) << 8) | (decSymbol(baseline) << 32) +} + +// decSymbolValue returns the transformed decSymbol for the given symbol. +func decSymbolValue(symb uint8, t []baseOffset) (decSymbol, error) { + if int(symb) >= len(t) { + return 0, fmt.Errorf("rle symbol %d >= max %d", symb, len(t)) + } + lu := t[symb] + return newDecSymbol(0, lu.addBits, 0, lu.baseLine), nil +} + +// setRLE will set the decoder til RLE mode. +func (s *fseDecoder) setRLE(symbol decSymbol) { + s.actualTableLog = 0 + s.maxBits = symbol.addBits() + s.dt[0] = symbol +} + +// buildDtable will build the decoding table. +func (s *fseDecoder) buildDtable() error { + tableSize := uint32(1 << s.actualTableLog) + highThreshold := tableSize - 1 + symbolNext := s.stateTable[:256] + + // Init, lay down lowprob symbols + { + for i, v := range s.norm[:s.symbolLen] { + if v == -1 { + s.dt[highThreshold].setAddBits(uint8(i)) + highThreshold-- + symbolNext[i] = 1 + } else { + symbolNext[i] = uint16(v) + } + } + } + // Spread symbols + { + tableMask := tableSize - 1 + step := tableStep(tableSize) + position := uint32(0) + for ss, v := range s.norm[:s.symbolLen] { + for i := 0; i < int(v); i++ { + s.dt[position].setAddBits(uint8(ss)) + position = (position + step) & tableMask + for position > highThreshold { + // lowprob area + position = (position + step) & tableMask + } + } + } + if position != 0 { + // position must reach all cells once, otherwise normalizedCounter is incorrect + return errors.New("corrupted input (position != 0)") + } + } + + // Build Decoding table + { + tableSize := uint16(1 << s.actualTableLog) + for u, v := range s.dt[:tableSize] { + symbol := v.addBits() + nextState := symbolNext[symbol] + symbolNext[symbol] = nextState + 1 + nBits := s.actualTableLog - byte(highBits(uint32(nextState))) + s.dt[u&maxTableMask].setNBits(nBits) + newState := (nextState << nBits) - tableSize + if newState > tableSize { + return fmt.Errorf("newState (%d) outside table size (%d)", newState, tableSize) + } + if newState == uint16(u) && nBits == 0 { + // Seems weird that this is possible with nbits > 0. + return fmt.Errorf("newState (%d) == oldState (%d) and no bits", newState, u) + } + s.dt[u&maxTableMask].setNewState(newState) + } + } + return nil +} + +// transform will transform the decoder table into a table usable for +// decoding without having to apply the transformation while decoding. +// The state will contain the base value and the number of bits to read. +func (s *fseDecoder) transform(t []baseOffset) error { + tableSize := uint16(1 << s.actualTableLog) + s.maxBits = 0 + for i, v := range s.dt[:tableSize] { + add := v.addBits() + if int(add) >= len(t) { + return fmt.Errorf("invalid decoding table entry %d, symbol %d >= max (%d)", i, v.addBits(), len(t)) + } + lu := t[add] + if lu.addBits > s.maxBits { + s.maxBits = lu.addBits + } + v.setExt(lu.addBits, lu.baseLine) + s.dt[i] = v + } + return nil +} + +type fseState struct { + dt []decSymbol + state decSymbol +} + +// Initialize and decodeAsync first state and symbol. +func (s *fseState) init(br *bitReader, tableLog uint8, dt []decSymbol) { + s.dt = dt + br.fill() + s.state = dt[br.getBits(tableLog)] +} + +// next returns the current symbol and sets the next state. +// At least tablelog bits must be available in the bit reader. +func (s *fseState) next(br *bitReader) { + lowBits := uint16(br.getBits(s.state.nbBits())) + s.state = s.dt[s.state.newState()+lowBits] +} + +// finished returns true if all bits have been read from the bitstream +// and the next state would require reading bits from the input. +func (s *fseState) finished(br *bitReader) bool { + return br.finished() && s.state.nbBits() > 0 +} + +// final returns the current state symbol without decoding the next. +func (s *fseState) final() (int, uint8) { + return s.state.baselineInt(), s.state.addBits() +} + +// final returns the current state symbol without decoding the next. +func (s decSymbol) final() (int, uint8) { + return s.baselineInt(), s.addBits() +} + +// nextFast returns the next symbol and sets the next state. +// This can only be used if no symbols are 0 bits. +// At least tablelog bits must be available in the bit reader. +func (s *fseState) nextFast(br *bitReader) (uint32, uint8) { + lowBits := br.get16BitsFast(s.state.nbBits()) + s.state = s.dt[s.state.newState()+lowBits] + return s.state.baseline(), s.state.addBits() +} diff --git a/vendor/github.com/klauspost/compress/zstd/fse_encoder.go b/vendor/github.com/klauspost/compress/zstd/fse_encoder.go new file mode 100644 index 0000000..5442061 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/fse_encoder.go @@ -0,0 +1,724 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "errors" + "fmt" + "math" +) + +const ( + // For encoding we only support up to + maxEncTableLog = 8 + maxEncTablesize = 1 << maxTableLog + maxEncTableMask = (1 << maxTableLog) - 1 + minEncTablelog = 5 + maxEncSymbolValue = maxMatchLengthSymbol +) + +// Scratch provides temporary storage for compression and decompression. +type fseEncoder struct { + symbolLen uint16 // Length of active part of the symbol table. + actualTableLog uint8 // Selected tablelog. + ct cTable // Compression tables. + maxCount int // count of the most probable symbol + zeroBits bool // no bits has prob > 50%. + clearCount bool // clear count + useRLE bool // This encoder is for RLE + preDefined bool // This encoder is predefined. + reUsed bool // Set to know when the encoder has been reused. + rleVal uint8 // RLE Symbol + maxBits uint8 // Maximum output bits after transform. + + // TODO: Technically zstd should be fine with 64 bytes. + count [256]uint32 + norm [256]int16 +} + +// cTable contains tables used for compression. +type cTable struct { + tableSymbol []byte + stateTable []uint16 + symbolTT []symbolTransform +} + +// symbolTransform contains the state transform for a symbol. +type symbolTransform struct { + deltaNbBits uint32 + deltaFindState int16 + outBits uint8 +} + +// String prints values as a human readable string. +func (s symbolTransform) String() string { + return fmt.Sprintf("{deltabits: %08x, findstate:%d outbits:%d}", s.deltaNbBits, s.deltaFindState, s.outBits) +} + +// Histogram allows to populate the histogram and skip that step in the compression, +// It otherwise allows to inspect the histogram when compression is done. +// To indicate that you have populated the histogram call HistogramFinished +// with the value of the highest populated symbol, as well as the number of entries +// in the most populated entry. These are accepted at face value. +func (s *fseEncoder) Histogram() *[256]uint32 { + return &s.count +} + +// HistogramFinished can be called to indicate that the histogram has been populated. +// maxSymbol is the index of the highest set symbol of the next data segment. +// maxCount is the number of entries in the most populated entry. +// These are accepted at face value. +func (s *fseEncoder) HistogramFinished(maxSymbol uint8, maxCount int) { + s.maxCount = maxCount + s.symbolLen = uint16(maxSymbol) + 1 + s.clearCount = maxCount != 0 +} + +// prepare will prepare and allocate scratch tables used for both compression and decompression. +func (s *fseEncoder) prepare() (*fseEncoder, error) { + if s == nil { + s = &fseEncoder{} + } + s.useRLE = false + if s.clearCount && s.maxCount == 0 { + for i := range s.count { + s.count[i] = 0 + } + s.clearCount = false + } + return s, nil +} + +// allocCtable will allocate tables needed for compression. +// If existing tables a re big enough, they are simply re-used. +func (s *fseEncoder) allocCtable() { + tableSize := 1 << s.actualTableLog + // get tableSymbol that is big enough. + if cap(s.ct.tableSymbol) < tableSize { + s.ct.tableSymbol = make([]byte, tableSize) + } + s.ct.tableSymbol = s.ct.tableSymbol[:tableSize] + + ctSize := tableSize + if cap(s.ct.stateTable) < ctSize { + s.ct.stateTable = make([]uint16, ctSize) + } + s.ct.stateTable = s.ct.stateTable[:ctSize] + + if cap(s.ct.symbolTT) < 256 { + s.ct.symbolTT = make([]symbolTransform, 256) + } + s.ct.symbolTT = s.ct.symbolTT[:256] +} + +// buildCTable will populate the compression table so it is ready to be used. +func (s *fseEncoder) buildCTable() error { + tableSize := uint32(1 << s.actualTableLog) + highThreshold := tableSize - 1 + var cumul [256]int16 + + s.allocCtable() + tableSymbol := s.ct.tableSymbol[:tableSize] + // symbol start positions + { + cumul[0] = 0 + for ui, v := range s.norm[:s.symbolLen-1] { + u := byte(ui) // one less than reference + if v == -1 { + // Low proba symbol + cumul[u+1] = cumul[u] + 1 + tableSymbol[highThreshold] = u + highThreshold-- + } else { + cumul[u+1] = cumul[u] + v + } + } + // Encode last symbol separately to avoid overflowing u + u := int(s.symbolLen - 1) + v := s.norm[s.symbolLen-1] + if v == -1 { + // Low proba symbol + cumul[u+1] = cumul[u] + 1 + tableSymbol[highThreshold] = byte(u) + highThreshold-- + } else { + cumul[u+1] = cumul[u] + v + } + if uint32(cumul[s.symbolLen]) != tableSize { + return fmt.Errorf("internal error: expected cumul[s.symbolLen] (%d) == tableSize (%d)", cumul[s.symbolLen], tableSize) + } + cumul[s.symbolLen] = int16(tableSize) + 1 + } + // Spread symbols + s.zeroBits = false + { + step := tableStep(tableSize) + tableMask := tableSize - 1 + var position uint32 + // if any symbol > largeLimit, we may have 0 bits output. + largeLimit := int16(1 << (s.actualTableLog - 1)) + for ui, v := range s.norm[:s.symbolLen] { + symbol := byte(ui) + if v > largeLimit { + s.zeroBits = true + } + for nbOccurrences := int16(0); nbOccurrences < v; nbOccurrences++ { + tableSymbol[position] = symbol + position = (position + step) & tableMask + for position > highThreshold { + position = (position + step) & tableMask + } /* Low proba area */ + } + } + + // Check if we have gone through all positions + if position != 0 { + return errors.New("position!=0") + } + } + + // Build table + table := s.ct.stateTable + { + tsi := int(tableSize) + for u, v := range tableSymbol { + // TableU16 : sorted by symbol order; gives next state value + table[cumul[v]] = uint16(tsi + u) + cumul[v]++ + } + } + + // Build Symbol Transformation Table + { + total := int16(0) + symbolTT := s.ct.symbolTT[:s.symbolLen] + tableLog := s.actualTableLog + tl := (uint32(tableLog) << 16) - (1 << tableLog) + for i, v := range s.norm[:s.symbolLen] { + switch v { + case 0: + case -1, 1: + symbolTT[i].deltaNbBits = tl + symbolTT[i].deltaFindState = total - 1 + total++ + default: + maxBitsOut := uint32(tableLog) - highBit(uint32(v-1)) + minStatePlus := uint32(v) << maxBitsOut + symbolTT[i].deltaNbBits = (maxBitsOut << 16) - minStatePlus + symbolTT[i].deltaFindState = total - v + total += v + } + } + if total != int16(tableSize) { + return fmt.Errorf("total mismatch %d (got) != %d (want)", total, tableSize) + } + } + return nil +} + +var rtbTable = [...]uint32{0, 473195, 504333, 520860, 550000, 700000, 750000, 830000} + +func (s *fseEncoder) setRLE(val byte) { + s.allocCtable() + s.actualTableLog = 0 + s.ct.stateTable = s.ct.stateTable[:1] + s.ct.symbolTT[val] = symbolTransform{ + deltaFindState: 0, + deltaNbBits: 0, + } + if debugEncoder { + println("setRLE: val", val, "symbolTT", s.ct.symbolTT[val]) + } + s.rleVal = val + s.useRLE = true +} + +// setBits will set output bits for the transform. +// if nil is provided, the number of bits is equal to the index. +func (s *fseEncoder) setBits(transform []byte) { + if s.reUsed || s.preDefined { + return + } + if s.useRLE { + if transform == nil { + s.ct.symbolTT[s.rleVal].outBits = s.rleVal + s.maxBits = s.rleVal + return + } + s.maxBits = transform[s.rleVal] + s.ct.symbolTT[s.rleVal].outBits = s.maxBits + return + } + if transform == nil { + for i := range s.ct.symbolTT[:s.symbolLen] { + s.ct.symbolTT[i].outBits = uint8(i) + } + s.maxBits = uint8(s.symbolLen - 1) + return + } + s.maxBits = 0 + for i, v := range transform[:s.symbolLen] { + s.ct.symbolTT[i].outBits = v + if v > s.maxBits { + // We could assume bits always going up, but we play safe. + s.maxBits = v + } + } +} + +// normalizeCount will normalize the count of the symbols so +// the total is equal to the table size. +// If successful, compression tables will also be made ready. +func (s *fseEncoder) normalizeCount(length int) error { + if s.reUsed { + return nil + } + s.optimalTableLog(length) + var ( + tableLog = s.actualTableLog + scale = 62 - uint64(tableLog) + step = (1 << 62) / uint64(length) + vStep = uint64(1) << (scale - 20) + stillToDistribute = int16(1 << tableLog) + largest int + largestP int16 + lowThreshold = (uint32)(length >> tableLog) + ) + if s.maxCount == length { + s.useRLE = true + return nil + } + s.useRLE = false + for i, cnt := range s.count[:s.symbolLen] { + // already handled + // if (count[s] == s.length) return 0; /* rle special case */ + + if cnt == 0 { + s.norm[i] = 0 + continue + } + if cnt <= lowThreshold { + s.norm[i] = -1 + stillToDistribute-- + } else { + proba := (int16)((uint64(cnt) * step) >> scale) + if proba < 8 { + restToBeat := vStep * uint64(rtbTable[proba]) + v := uint64(cnt)*step - (uint64(proba) << scale) + if v > restToBeat { + proba++ + } + } + if proba > largestP { + largestP = proba + largest = i + } + s.norm[i] = proba + stillToDistribute -= proba + } + } + + if -stillToDistribute >= (s.norm[largest] >> 1) { + // corner case, need another normalization method + err := s.normalizeCount2(length) + if err != nil { + return err + } + if debugAsserts { + err = s.validateNorm() + if err != nil { + return err + } + } + return s.buildCTable() + } + s.norm[largest] += stillToDistribute + if debugAsserts { + err := s.validateNorm() + if err != nil { + return err + } + } + return s.buildCTable() +} + +// Secondary normalization method. +// To be used when primary method fails. +func (s *fseEncoder) normalizeCount2(length int) error { + const notYetAssigned = -2 + var ( + distributed uint32 + total = uint32(length) + tableLog = s.actualTableLog + lowThreshold = total >> tableLog + lowOne = (total * 3) >> (tableLog + 1) + ) + for i, cnt := range s.count[:s.symbolLen] { + if cnt == 0 { + s.norm[i] = 0 + continue + } + if cnt <= lowThreshold { + s.norm[i] = -1 + distributed++ + total -= cnt + continue + } + if cnt <= lowOne { + s.norm[i] = 1 + distributed++ + total -= cnt + continue + } + s.norm[i] = notYetAssigned + } + toDistribute := (1 << tableLog) - distributed + + if (total / toDistribute) > lowOne { + // risk of rounding to zero + lowOne = (total * 3) / (toDistribute * 2) + for i, cnt := range s.count[:s.symbolLen] { + if (s.norm[i] == notYetAssigned) && (cnt <= lowOne) { + s.norm[i] = 1 + distributed++ + total -= cnt + continue + } + } + toDistribute = (1 << tableLog) - distributed + } + if distributed == uint32(s.symbolLen)+1 { + // all values are pretty poor; + // probably incompressible data (should have already been detected); + // find max, then give all remaining points to max + var maxV int + var maxC uint32 + for i, cnt := range s.count[:s.symbolLen] { + if cnt > maxC { + maxV = i + maxC = cnt + } + } + s.norm[maxV] += int16(toDistribute) + return nil + } + + if total == 0 { + // all of the symbols were low enough for the lowOne or lowThreshold + for i := uint32(0); toDistribute > 0; i = (i + 1) % (uint32(s.symbolLen)) { + if s.norm[i] > 0 { + toDistribute-- + s.norm[i]++ + } + } + return nil + } + + var ( + vStepLog = 62 - uint64(tableLog) + mid = uint64((1 << (vStepLog - 1)) - 1) + rStep = (((1 << vStepLog) * uint64(toDistribute)) + mid) / uint64(total) // scale on remaining + tmpTotal = mid + ) + for i, cnt := range s.count[:s.symbolLen] { + if s.norm[i] == notYetAssigned { + var ( + end = tmpTotal + uint64(cnt)*rStep + sStart = uint32(tmpTotal >> vStepLog) + sEnd = uint32(end >> vStepLog) + weight = sEnd - sStart + ) + if weight < 1 { + return errors.New("weight < 1") + } + s.norm[i] = int16(weight) + tmpTotal = end + } + } + return nil +} + +// optimalTableLog calculates and sets the optimal tableLog in s.actualTableLog +func (s *fseEncoder) optimalTableLog(length int) { + tableLog := uint8(maxEncTableLog) + minBitsSrc := highBit(uint32(length)) + 1 + minBitsSymbols := highBit(uint32(s.symbolLen-1)) + 2 + minBits := uint8(minBitsSymbols) + if minBitsSrc < minBitsSymbols { + minBits = uint8(minBitsSrc) + } + + maxBitsSrc := uint8(highBit(uint32(length-1))) - 2 + if maxBitsSrc < tableLog { + // Accuracy can be reduced + tableLog = maxBitsSrc + } + if minBits > tableLog { + tableLog = minBits + } + // Need a minimum to safely represent all symbol values + if tableLog < minEncTablelog { + tableLog = minEncTablelog + } + if tableLog > maxEncTableLog { + tableLog = maxEncTableLog + } + s.actualTableLog = tableLog +} + +// validateNorm validates the normalized histogram table. +func (s *fseEncoder) validateNorm() (err error) { + var total int + for _, v := range s.norm[:s.symbolLen] { + if v >= 0 { + total += int(v) + } else { + total -= int(v) + } + } + defer func() { + if err == nil { + return + } + fmt.Printf("selected TableLog: %d, Symbol length: %d\n", s.actualTableLog, s.symbolLen) + for i, v := range s.norm[:s.symbolLen] { + fmt.Printf("%3d: %5d -> %4d \n", i, s.count[i], v) + } + }() + if total != (1 << s.actualTableLog) { + return fmt.Errorf("warning: Total == %d != %d", total, 1<> 3) + 3 + 2 + + // Write Table Size + bitStream = uint32(tableLog - minEncTablelog) + bitCount = uint(4) + remaining = int16(tableSize + 1) /* +1 for extra accuracy */ + threshold = int16(tableSize) + nbBits = uint(tableLog + 1) + outP = len(out) + ) + if cap(out) < outP+maxHeaderSize { + out = append(out, make([]byte, maxHeaderSize*3)...) + out = out[:len(out)-maxHeaderSize*3] + } + out = out[:outP+maxHeaderSize] + + // stops at 1 + for remaining > 1 { + if previous0 { + start := charnum + for s.norm[charnum] == 0 { + charnum++ + } + for charnum >= start+24 { + start += 24 + bitStream += uint32(0xFFFF) << bitCount + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += 2 + bitStream >>= 16 + } + for charnum >= start+3 { + start += 3 + bitStream += 3 << bitCount + bitCount += 2 + } + bitStream += uint32(charnum-start) << bitCount + bitCount += 2 + if bitCount > 16 { + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += 2 + bitStream >>= 16 + bitCount -= 16 + } + } + + count := s.norm[charnum] + charnum++ + max := (2*threshold - 1) - remaining + if count < 0 { + remaining += count + } else { + remaining -= count + } + count++ // +1 for extra accuracy + if count >= threshold { + count += max // [0..max[ [max..threshold[ (...) [threshold+max 2*threshold[ + } + bitStream += uint32(count) << bitCount + bitCount += nbBits + if count < max { + bitCount-- + } + + previous0 = count == 1 + if remaining < 1 { + return nil, errors.New("internal error: remaining < 1") + } + for remaining < threshold { + nbBits-- + threshold >>= 1 + } + + if bitCount > 16 { + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += 2 + bitStream >>= 16 + bitCount -= 16 + } + } + + if outP+2 > len(out) { + return nil, fmt.Errorf("internal error: %d > %d, maxheader: %d, sl: %d, tl: %d, normcount: %v", outP+2, len(out), maxHeaderSize, s.symbolLen, int(tableLog), s.norm[:s.symbolLen]) + } + out[outP] = byte(bitStream) + out[outP+1] = byte(bitStream >> 8) + outP += int((bitCount + 7) / 8) + + if charnum > s.symbolLen { + return nil, errors.New("internal error: charnum > s.symbolLen") + } + return out[:outP], nil +} + +// Approximate symbol cost, as fractional value, using fixed-point format (accuracyLog fractional bits) +// note 1 : assume symbolValue is valid (<= maxSymbolValue) +// note 2 : if freq[symbolValue]==0, @return a fake cost of tableLog+1 bits * +func (s *fseEncoder) bitCost(symbolValue uint8, accuracyLog uint32) uint32 { + minNbBits := s.ct.symbolTT[symbolValue].deltaNbBits >> 16 + threshold := (minNbBits + 1) << 16 + if debugAsserts { + if !(s.actualTableLog < 16) { + panic("!s.actualTableLog < 16") + } + // ensure enough room for renormalization double shift + if !(uint8(accuracyLog) < 31-s.actualTableLog) { + panic("!uint8(accuracyLog) < 31-s.actualTableLog") + } + } + tableSize := uint32(1) << s.actualTableLog + deltaFromThreshold := threshold - (s.ct.symbolTT[symbolValue].deltaNbBits + tableSize) + // linear interpolation (very approximate) + normalizedDeltaFromThreshold := (deltaFromThreshold << accuracyLog) >> s.actualTableLog + bitMultiplier := uint32(1) << accuracyLog + if debugAsserts { + if s.ct.symbolTT[symbolValue].deltaNbBits+tableSize > threshold { + panic("s.ct.symbolTT[symbolValue].deltaNbBits+tableSize > threshold") + } + if normalizedDeltaFromThreshold > bitMultiplier { + panic("normalizedDeltaFromThreshold > bitMultiplier") + } + } + return (minNbBits+1)*bitMultiplier - normalizedDeltaFromThreshold +} + +// Returns the cost in bits of encoding the distribution in count using ctable. +// Histogram should only be up to the last non-zero symbol. +// Returns an -1 if ctable cannot represent all the symbols in count. +func (s *fseEncoder) approxSize(hist []uint32) uint32 { + if int(s.symbolLen) < len(hist) { + // More symbols than we have. + return math.MaxUint32 + } + if s.useRLE { + // We will never reuse RLE encoders. + return math.MaxUint32 + } + const kAccuracyLog = 8 + badCost := (uint32(s.actualTableLog) + 1) << kAccuracyLog + var cost uint32 + for i, v := range hist { + if v == 0 { + continue + } + if s.norm[i] == 0 { + return math.MaxUint32 + } + bitCost := s.bitCost(uint8(i), kAccuracyLog) + if bitCost > badCost { + return math.MaxUint32 + } + cost += v * bitCost + } + return cost >> kAccuracyLog +} + +// maxHeaderSize returns the maximum header size in bits. +// This is not exact size, but we want a penalty for new tables anyway. +func (s *fseEncoder) maxHeaderSize() uint32 { + if s.preDefined { + return 0 + } + if s.useRLE { + return 8 + } + return (((uint32(s.symbolLen) * uint32(s.actualTableLog)) >> 3) + 3) * 8 +} + +// cState contains the compression state of a stream. +type cState struct { + bw *bitWriter + stateTable []uint16 + state uint16 +} + +// init will initialize the compression state to the first symbol of the stream. +func (c *cState) init(bw *bitWriter, ct *cTable, first symbolTransform) { + c.bw = bw + c.stateTable = ct.stateTable + if len(c.stateTable) == 1 { + // RLE + c.stateTable[0] = uint16(0) + c.state = 0 + return + } + nbBitsOut := (first.deltaNbBits + (1 << 15)) >> 16 + im := int32((nbBitsOut << 16) - first.deltaNbBits) + lu := (im >> nbBitsOut) + int32(first.deltaFindState) + c.state = c.stateTable[lu] +} + +// encode the output symbol provided and write it to the bitstream. +func (c *cState) encode(symbolTT symbolTransform) { + nbBitsOut := (uint32(c.state) + symbolTT.deltaNbBits) >> 16 + dstState := int32(c.state>>(nbBitsOut&15)) + int32(symbolTT.deltaFindState) + c.bw.addBits16NC(c.state, uint8(nbBitsOut)) + c.state = c.stateTable[dstState] +} + +// flush will write the tablelog to the output and flush the remaining full bytes. +func (c *cState) flush(tableLog uint8) { + c.bw.flush32() + c.bw.addBits16NC(c.state, tableLog) +} diff --git a/vendor/github.com/klauspost/compress/zstd/fse_predefined.go b/vendor/github.com/klauspost/compress/zstd/fse_predefined.go new file mode 100644 index 0000000..474cb77 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/fse_predefined.go @@ -0,0 +1,158 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "fmt" + "math" + "sync" +) + +var ( + // fsePredef are the predefined fse tables as defined here: + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#default-distributions + // These values are already transformed. + fsePredef [3]fseDecoder + + // fsePredefEnc are the predefined encoder based on fse tables as defined here: + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#default-distributions + // These values are already transformed. + fsePredefEnc [3]fseEncoder + + // symbolTableX contain the transformations needed for each type as defined in + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#the-codes-for-literals-lengths-match-lengths-and-offsets + symbolTableX [3][]baseOffset + + // maxTableSymbol is the biggest supported symbol for each table type + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#the-codes-for-literals-lengths-match-lengths-and-offsets + maxTableSymbol = [3]uint8{tableLiteralLengths: maxLiteralLengthSymbol, tableOffsets: maxOffsetLengthSymbol, tableMatchLengths: maxMatchLengthSymbol} + + // bitTables is the bits table for each table. + bitTables = [3][]byte{tableLiteralLengths: llBitsTable[:], tableOffsets: nil, tableMatchLengths: mlBitsTable[:]} +) + +type tableIndex uint8 + +const ( + // indexes for fsePredef and symbolTableX + tableLiteralLengths tableIndex = 0 + tableOffsets tableIndex = 1 + tableMatchLengths tableIndex = 2 + + maxLiteralLengthSymbol = 35 + maxOffsetLengthSymbol = 30 + maxMatchLengthSymbol = 52 +) + +// baseOffset is used for calculating transformations. +type baseOffset struct { + baseLine uint32 + addBits uint8 +} + +// fillBase will precalculate base offsets with the given bit distributions. +func fillBase(dst []baseOffset, base uint32, bits ...uint8) { + if len(bits) != len(dst) { + panic(fmt.Sprintf("len(dst) (%d) != len(bits) (%d)", len(dst), len(bits))) + } + for i, bit := range bits { + if base > math.MaxInt32 { + panic("invalid decoding table, base overflows int32") + } + + dst[i] = baseOffset{ + baseLine: base, + addBits: bit, + } + base += 1 << bit + } +} + +var predef sync.Once + +func initPredefined() { + predef.Do(func() { + // Literals length codes + tmp := make([]baseOffset, 36) + for i := range tmp[:16] { + tmp[i] = baseOffset{ + baseLine: uint32(i), + addBits: 0, + } + } + fillBase(tmp[16:], 16, 1, 1, 1, 1, 2, 2, 3, 3, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) + symbolTableX[tableLiteralLengths] = tmp + + // Match length codes + tmp = make([]baseOffset, 53) + for i := range tmp[:32] { + tmp[i] = baseOffset{ + // The transformation adds the 3 length. + baseLine: uint32(i) + 3, + addBits: 0, + } + } + fillBase(tmp[32:], 35, 1, 1, 1, 1, 2, 2, 3, 3, 4, 4, 5, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) + symbolTableX[tableMatchLengths] = tmp + + // Offset codes + tmp = make([]baseOffset, maxOffsetBits+1) + tmp[1] = baseOffset{ + baseLine: 1, + addBits: 1, + } + fillBase(tmp[2:], 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30) + symbolTableX[tableOffsets] = tmp + + // Fill predefined tables and transform them. + // https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#default-distributions + for i := range fsePredef[:] { + f := &fsePredef[i] + switch tableIndex(i) { + case tableLiteralLengths: + // https://github.com/facebook/zstd/blob/ededcfca57366461021c922720878c81a5854a0a/lib/decompress/zstd_decompress_block.c#L243 + f.actualTableLog = 6 + copy(f.norm[:], []int16{4, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 2, 1, 1, 1, 1, 1, + -1, -1, -1, -1}) + f.symbolLen = 36 + case tableOffsets: + // https://github.com/facebook/zstd/blob/ededcfca57366461021c922720878c81a5854a0a/lib/decompress/zstd_decompress_block.c#L281 + f.actualTableLog = 5 + copy(f.norm[:], []int16{ + 1, 1, 1, 1, 1, 1, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1}) + f.symbolLen = 29 + case tableMatchLengths: + //https://github.com/facebook/zstd/blob/ededcfca57366461021c922720878c81a5854a0a/lib/decompress/zstd_decompress_block.c#L304 + f.actualTableLog = 6 + copy(f.norm[:], []int16{ + 1, 4, 3, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, + -1, -1, -1, -1, -1}) + f.symbolLen = 53 + } + if err := f.buildDtable(); err != nil { + panic(fmt.Errorf("building table %v: %v", tableIndex(i), err)) + } + if err := f.transform(symbolTableX[i]); err != nil { + panic(fmt.Errorf("building table %v: %v", tableIndex(i), err)) + } + f.preDefined = true + + // Create encoder as well + enc := &fsePredefEnc[i] + copy(enc.norm[:], f.norm[:]) + enc.symbolLen = f.symbolLen + enc.actualTableLog = f.actualTableLog + if err := enc.buildCTable(); err != nil { + panic(fmt.Errorf("building encoding table %v: %v", tableIndex(i), err)) + } + enc.setBits(bitTables[i]) + enc.preDefined = true + } + }) +} diff --git a/vendor/github.com/klauspost/compress/zstd/fuzz.go b/vendor/github.com/klauspost/compress/zstd/fuzz.go new file mode 100644 index 0000000..7f2210e --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/fuzz.go @@ -0,0 +1,11 @@ +//go:build ignorecrc +// +build ignorecrc + +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +// ignoreCRC can be used for fuzz testing to ignore CRC values... +const ignoreCRC = true diff --git a/vendor/github.com/klauspost/compress/zstd/fuzz_none.go b/vendor/github.com/klauspost/compress/zstd/fuzz_none.go new file mode 100644 index 0000000..6811c68 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/fuzz_none.go @@ -0,0 +1,11 @@ +//go:build !ignorecrc +// +build !ignorecrc + +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +// ignoreCRC can be used for fuzz testing to ignore CRC values... +const ignoreCRC = false diff --git a/vendor/github.com/klauspost/compress/zstd/hash.go b/vendor/github.com/klauspost/compress/zstd/hash.go new file mode 100644 index 0000000..cf33f29 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/hash.go @@ -0,0 +1,41 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +const ( + prime3bytes = 506832829 + prime4bytes = 2654435761 + prime5bytes = 889523592379 + prime6bytes = 227718039650203 + prime7bytes = 58295818150454627 + prime8bytes = 0xcf1bbcdcb7a56463 +) + +// hashLen returns a hash of the lowest mls bytes of with length output bits. +// mls must be >=3 and <=8. Any other value will return hash for 4 bytes. +// length should always be < 32. +// Preferably length and mls should be a constant for inlining. +func hashLen(u uint64, length, mls uint8) uint32 { + switch mls { + case 3: + return (uint32(u<<8) * prime3bytes) >> (32 - length) + case 5: + return uint32(((u << (64 - 40)) * prime5bytes) >> (64 - length)) + case 6: + return uint32(((u << (64 - 48)) * prime6bytes) >> (64 - length)) + case 7: + return uint32(((u << (64 - 56)) * prime7bytes) >> (64 - length)) + case 8: + return uint32((u * prime8bytes) >> (64 - length)) + default: + return (uint32(u) * prime4bytes) >> (32 - length) + } +} + +// hash3 returns the hash of the lower 3 bytes of u to fit in a hash table with h bits. +// Preferably h should be a constant and should always be <32. +func hash3(u uint32, h uint8) uint32 { + return ((u << (32 - 24)) * prime3bytes) >> ((32 - h) & 31) +} diff --git a/vendor/github.com/klauspost/compress/zstd/history.go b/vendor/github.com/klauspost/compress/zstd/history.go new file mode 100644 index 0000000..28b4015 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/history.go @@ -0,0 +1,119 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "github.com/klauspost/compress/huff0" +) + +// history contains the information transferred between blocks. +type history struct { + // Literal decompression + huffTree *huff0.Scratch + + // Sequence decompression + decoders sequenceDecs + recentOffsets [3]int + + // History buffer... + b []byte + + // ignoreBuffer is meant to ignore a number of bytes + // when checking for matches in history + ignoreBuffer int + + windowSize int + allocFrameBuffer int // needed? + error bool + dict *dict +} + +// reset will reset the history to initial state of a frame. +// The history must already have been initialized to the desired size. +func (h *history) reset() { + h.b = h.b[:0] + h.ignoreBuffer = 0 + h.error = false + h.recentOffsets = [3]int{1, 4, 8} + if f := h.decoders.litLengths.fse; f != nil && !f.preDefined { + fseDecoderPool.Put(f) + } + if f := h.decoders.offsets.fse; f != nil && !f.preDefined { + fseDecoderPool.Put(f) + } + if f := h.decoders.matchLengths.fse; f != nil && !f.preDefined { + fseDecoderPool.Put(f) + } + h.decoders = sequenceDecs{br: h.decoders.br} + if h.huffTree != nil { + if h.dict == nil || h.dict.litEnc != h.huffTree { + huffDecoderPool.Put(h.huffTree) + } + } + h.huffTree = nil + h.dict = nil + //printf("history created: %+v (l: %d, c: %d)", *h, len(h.b), cap(h.b)) +} + +func (h *history) setDict(dict *dict) { + if dict == nil { + return + } + h.dict = dict + h.decoders.litLengths = dict.llDec + h.decoders.offsets = dict.ofDec + h.decoders.matchLengths = dict.mlDec + h.decoders.dict = dict.content + h.recentOffsets = dict.offsets + h.huffTree = dict.litEnc +} + +// append bytes to history. +// This function will make sure there is space for it, +// if the buffer has been allocated with enough extra space. +func (h *history) append(b []byte) { + if len(b) >= h.windowSize { + // Discard all history by simply overwriting + h.b = h.b[:h.windowSize] + copy(h.b, b[len(b)-h.windowSize:]) + return + } + + // If there is space, append it. + if len(b) < cap(h.b)-len(h.b) { + h.b = append(h.b, b...) + return + } + + // Move data down so we only have window size left. + // We know we have less than window size in b at this point. + discard := len(b) + len(h.b) - h.windowSize + copy(h.b, h.b[discard:]) + h.b = h.b[:h.windowSize] + copy(h.b[h.windowSize-len(b):], b) +} + +// ensureBlock will ensure there is space for at least one block... +func (h *history) ensureBlock() { + if cap(h.b) < h.allocFrameBuffer { + h.b = make([]byte, 0, h.allocFrameBuffer) + return + } + + avail := cap(h.b) - len(h.b) + if avail >= h.windowSize || avail > maxCompressedBlockSize { + return + } + // Move data down so we only have window size left. + // We know we have less than window size in b at this point. + discard := len(h.b) - h.windowSize + copy(h.b, h.b[discard:]) + h.b = h.b[:h.windowSize] +} + +// append bytes to history without ever discarding anything. +func (h *history) appendKeep(b []byte) { + h.b = append(h.b, b...) +} diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/LICENSE.txt b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/LICENSE.txt new file mode 100644 index 0000000..24b5306 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/LICENSE.txt @@ -0,0 +1,22 @@ +Copyright (c) 2016 Caleb Spare + +MIT License + +Permission is hereby granted, free of charge, to any person obtaining +a copy of this software and associated documentation files (the +"Software"), to deal in the Software without restriction, including +without limitation the rights to use, copy, modify, merge, publish, +distribute, sublicense, and/or sell copies of the Software, and to +permit persons to whom the Software is furnished to do so, subject to +the following conditions: + +The above copyright notice and this permission notice shall be +included in all copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, +EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF +MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND +NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE +LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION +OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/README.md b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/README.md new file mode 100644 index 0000000..69aa3bb --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/README.md @@ -0,0 +1,58 @@ +# xxhash + +VENDORED: Go to [github.com/cespare/xxhash](https://github.com/cespare/xxhash) for original package. + + +[![GoDoc](https://godoc.org/github.com/cespare/xxhash?status.svg)](https://godoc.org/github.com/cespare/xxhash) +[![Build Status](https://travis-ci.org/cespare/xxhash.svg?branch=master)](https://travis-ci.org/cespare/xxhash) + +xxhash is a Go implementation of the 64-bit +[xxHash](http://cyan4973.github.io/xxHash/) algorithm, XXH64. This is a +high-quality hashing algorithm that is much faster than anything in the Go +standard library. + +This package provides a straightforward API: + +``` +func Sum64(b []byte) uint64 +func Sum64String(s string) uint64 +type Digest struct{ ... } + func New() *Digest +``` + +The `Digest` type implements hash.Hash64. Its key methods are: + +``` +func (*Digest) Write([]byte) (int, error) +func (*Digest) WriteString(string) (int, error) +func (*Digest) Sum64() uint64 +``` + +This implementation provides a fast pure-Go implementation and an even faster +assembly implementation for amd64. + +## Benchmarks + +Here are some quick benchmarks comparing the pure-Go and assembly +implementations of Sum64. + +| input size | purego | asm | +| --- | --- | --- | +| 5 B | 979.66 MB/s | 1291.17 MB/s | +| 100 B | 7475.26 MB/s | 7973.40 MB/s | +| 4 KB | 17573.46 MB/s | 17602.65 MB/s | +| 10 MB | 17131.46 MB/s | 17142.16 MB/s | + +These numbers were generated on Ubuntu 18.04 with an Intel i7-8700K CPU using +the following commands under Go 1.11.2: + +``` +$ go test -tags purego -benchtime 10s -bench '/xxhash,direct,bytes' +$ go test -benchtime 10s -bench '/xxhash,direct,bytes' +``` + +## Projects using this package + +- [InfluxDB](https://github.com/influxdata/influxdb) +- [Prometheus](https://github.com/prometheus/prometheus) +- [FreeCache](https://github.com/coocood/freecache) diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash.go b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash.go new file mode 100644 index 0000000..2c112a0 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash.go @@ -0,0 +1,237 @@ +// Package xxhash implements the 64-bit variant of xxHash (XXH64) as described +// at http://cyan4973.github.io/xxHash/. +// THIS IS VENDORED: Go to github.com/cespare/xxhash for original package. + +package xxhash + +import ( + "encoding/binary" + "errors" + "math/bits" +) + +const ( + prime1 uint64 = 11400714785074694791 + prime2 uint64 = 14029467366897019727 + prime3 uint64 = 1609587929392839161 + prime4 uint64 = 9650029242287828579 + prime5 uint64 = 2870177450012600261 +) + +// NOTE(caleb): I'm using both consts and vars of the primes. Using consts where +// possible in the Go code is worth a small (but measurable) performance boost +// by avoiding some MOVQs. Vars are needed for the asm and also are useful for +// convenience in the Go code in a few places where we need to intentionally +// avoid constant arithmetic (e.g., v1 := prime1 + prime2 fails because the +// result overflows a uint64). +var ( + prime1v = prime1 + prime2v = prime2 + prime3v = prime3 + prime4v = prime4 + prime5v = prime5 +) + +// Digest implements hash.Hash64. +type Digest struct { + v1 uint64 + v2 uint64 + v3 uint64 + v4 uint64 + total uint64 + mem [32]byte + n int // how much of mem is used +} + +// New creates a new Digest that computes the 64-bit xxHash algorithm. +func New() *Digest { + var d Digest + d.Reset() + return &d +} + +// Reset clears the Digest's state so that it can be reused. +func (d *Digest) Reset() { + d.v1 = prime1v + prime2 + d.v2 = prime2 + d.v3 = 0 + d.v4 = -prime1v + d.total = 0 + d.n = 0 +} + +// Size always returns 8 bytes. +func (d *Digest) Size() int { return 8 } + +// BlockSize always returns 32 bytes. +func (d *Digest) BlockSize() int { return 32 } + +// Write adds more data to d. It always returns len(b), nil. +func (d *Digest) Write(b []byte) (n int, err error) { + n = len(b) + d.total += uint64(n) + + if d.n+n < 32 { + // This new data doesn't even fill the current block. + copy(d.mem[d.n:], b) + d.n += n + return + } + + if d.n > 0 { + // Finish off the partial block. + copy(d.mem[d.n:], b) + d.v1 = round(d.v1, u64(d.mem[0:8])) + d.v2 = round(d.v2, u64(d.mem[8:16])) + d.v3 = round(d.v3, u64(d.mem[16:24])) + d.v4 = round(d.v4, u64(d.mem[24:32])) + b = b[32-d.n:] + d.n = 0 + } + + if len(b) >= 32 { + // One or more full blocks left. + nw := writeBlocks(d, b) + b = b[nw:] + } + + // Store any remaining partial block. + copy(d.mem[:], b) + d.n = len(b) + + return +} + +// Sum appends the current hash to b and returns the resulting slice. +func (d *Digest) Sum(b []byte) []byte { + s := d.Sum64() + return append( + b, + byte(s>>56), + byte(s>>48), + byte(s>>40), + byte(s>>32), + byte(s>>24), + byte(s>>16), + byte(s>>8), + byte(s), + ) +} + +// Sum64 returns the current hash. +func (d *Digest) Sum64() uint64 { + var h uint64 + + if d.total >= 32 { + v1, v2, v3, v4 := d.v1, d.v2, d.v3, d.v4 + h = rol1(v1) + rol7(v2) + rol12(v3) + rol18(v4) + h = mergeRound(h, v1) + h = mergeRound(h, v2) + h = mergeRound(h, v3) + h = mergeRound(h, v4) + } else { + h = d.v3 + prime5 + } + + h += d.total + + i, end := 0, d.n + for ; i+8 <= end; i += 8 { + k1 := round(0, u64(d.mem[i:i+8])) + h ^= k1 + h = rol27(h)*prime1 + prime4 + } + if i+4 <= end { + h ^= uint64(u32(d.mem[i:i+4])) * prime1 + h = rol23(h)*prime2 + prime3 + i += 4 + } + for i < end { + h ^= uint64(d.mem[i]) * prime5 + h = rol11(h) * prime1 + i++ + } + + h ^= h >> 33 + h *= prime2 + h ^= h >> 29 + h *= prime3 + h ^= h >> 32 + + return h +} + +const ( + magic = "xxh\x06" + marshaledSize = len(magic) + 8*5 + 32 +) + +// MarshalBinary implements the encoding.BinaryMarshaler interface. +func (d *Digest) MarshalBinary() ([]byte, error) { + b := make([]byte, 0, marshaledSize) + b = append(b, magic...) + b = appendUint64(b, d.v1) + b = appendUint64(b, d.v2) + b = appendUint64(b, d.v3) + b = appendUint64(b, d.v4) + b = appendUint64(b, d.total) + b = append(b, d.mem[:d.n]...) + b = b[:len(b)+len(d.mem)-d.n] + return b, nil +} + +// UnmarshalBinary implements the encoding.BinaryUnmarshaler interface. +func (d *Digest) UnmarshalBinary(b []byte) error { + if len(b) < len(magic) || string(b[:len(magic)]) != magic { + return errors.New("xxhash: invalid hash state identifier") + } + if len(b) != marshaledSize { + return errors.New("xxhash: invalid hash state size") + } + b = b[len(magic):] + b, d.v1 = consumeUint64(b) + b, d.v2 = consumeUint64(b) + b, d.v3 = consumeUint64(b) + b, d.v4 = consumeUint64(b) + b, d.total = consumeUint64(b) + copy(d.mem[:], b) + d.n = int(d.total % uint64(len(d.mem))) + return nil +} + +func appendUint64(b []byte, x uint64) []byte { + var a [8]byte + binary.LittleEndian.PutUint64(a[:], x) + return append(b, a[:]...) +} + +func consumeUint64(b []byte) ([]byte, uint64) { + x := u64(b) + return b[8:], x +} + +func u64(b []byte) uint64 { return binary.LittleEndian.Uint64(b) } +func u32(b []byte) uint32 { return binary.LittleEndian.Uint32(b) } + +func round(acc, input uint64) uint64 { + acc += input * prime2 + acc = rol31(acc) + acc *= prime1 + return acc +} + +func mergeRound(acc, val uint64) uint64 { + val = round(0, val) + acc ^= val + acc = acc*prime1 + prime4 + return acc +} + +func rol1(x uint64) uint64 { return bits.RotateLeft64(x, 1) } +func rol7(x uint64) uint64 { return bits.RotateLeft64(x, 7) } +func rol11(x uint64) uint64 { return bits.RotateLeft64(x, 11) } +func rol12(x uint64) uint64 { return bits.RotateLeft64(x, 12) } +func rol18(x uint64) uint64 { return bits.RotateLeft64(x, 18) } +func rol23(x uint64) uint64 { return bits.RotateLeft64(x, 23) } +func rol27(x uint64) uint64 { return bits.RotateLeft64(x, 27) } +func rol31(x uint64) uint64 { return bits.RotateLeft64(x, 31) } diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_amd64.s b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_amd64.s new file mode 100644 index 0000000..cea1785 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_amd64.s @@ -0,0 +1,216 @@ +// +build !appengine +// +build gc +// +build !purego +// +build !noasm + +#include "textflag.h" + +// Register allocation: +// AX h +// SI pointer to advance through b +// DX n +// BX loop end +// R8 v1, k1 +// R9 v2 +// R10 v3 +// R11 v4 +// R12 tmp +// R13 prime1v +// R14 prime2v +// DI prime4v + +// round reads from and advances the buffer pointer in SI. +// It assumes that R13 has prime1v and R14 has prime2v. +#define round(r) \ + MOVQ (SI), R12 \ + ADDQ $8, SI \ + IMULQ R14, R12 \ + ADDQ R12, r \ + ROLQ $31, r \ + IMULQ R13, r + +// mergeRound applies a merge round on the two registers acc and val. +// It assumes that R13 has prime1v, R14 has prime2v, and DI has prime4v. +#define mergeRound(acc, val) \ + IMULQ R14, val \ + ROLQ $31, val \ + IMULQ R13, val \ + XORQ val, acc \ + IMULQ R13, acc \ + ADDQ DI, acc + +// func Sum64(b []byte) uint64 +TEXT ·Sum64(SB), NOSPLIT, $0-32 + // Load fixed primes. + MOVQ ·prime1v(SB), R13 + MOVQ ·prime2v(SB), R14 + MOVQ ·prime4v(SB), DI + + // Load slice. + MOVQ b_base+0(FP), SI + MOVQ b_len+8(FP), DX + LEAQ (SI)(DX*1), BX + + // The first loop limit will be len(b)-32. + SUBQ $32, BX + + // Check whether we have at least one block. + CMPQ DX, $32 + JLT noBlocks + + // Set up initial state (v1, v2, v3, v4). + MOVQ R13, R8 + ADDQ R14, R8 + MOVQ R14, R9 + XORQ R10, R10 + XORQ R11, R11 + SUBQ R13, R11 + + // Loop until SI > BX. +blockLoop: + round(R8) + round(R9) + round(R10) + round(R11) + + CMPQ SI, BX + JLE blockLoop + + MOVQ R8, AX + ROLQ $1, AX + MOVQ R9, R12 + ROLQ $7, R12 + ADDQ R12, AX + MOVQ R10, R12 + ROLQ $12, R12 + ADDQ R12, AX + MOVQ R11, R12 + ROLQ $18, R12 + ADDQ R12, AX + + mergeRound(AX, R8) + mergeRound(AX, R9) + mergeRound(AX, R10) + mergeRound(AX, R11) + + JMP afterBlocks + +noBlocks: + MOVQ ·prime5v(SB), AX + +afterBlocks: + ADDQ DX, AX + + // Right now BX has len(b)-32, and we want to loop until SI > len(b)-8. + ADDQ $24, BX + + CMPQ SI, BX + JG fourByte + +wordLoop: + // Calculate k1. + MOVQ (SI), R8 + ADDQ $8, SI + IMULQ R14, R8 + ROLQ $31, R8 + IMULQ R13, R8 + + XORQ R8, AX + ROLQ $27, AX + IMULQ R13, AX + ADDQ DI, AX + + CMPQ SI, BX + JLE wordLoop + +fourByte: + ADDQ $4, BX + CMPQ SI, BX + JG singles + + MOVL (SI), R8 + ADDQ $4, SI + IMULQ R13, R8 + XORQ R8, AX + + ROLQ $23, AX + IMULQ R14, AX + ADDQ ·prime3v(SB), AX + +singles: + ADDQ $4, BX + CMPQ SI, BX + JGE finalize + +singlesLoop: + MOVBQZX (SI), R12 + ADDQ $1, SI + IMULQ ·prime5v(SB), R12 + XORQ R12, AX + + ROLQ $11, AX + IMULQ R13, AX + + CMPQ SI, BX + JL singlesLoop + +finalize: + MOVQ AX, R12 + SHRQ $33, R12 + XORQ R12, AX + IMULQ R14, AX + MOVQ AX, R12 + SHRQ $29, R12 + XORQ R12, AX + IMULQ ·prime3v(SB), AX + MOVQ AX, R12 + SHRQ $32, R12 + XORQ R12, AX + + MOVQ AX, ret+24(FP) + RET + +// writeBlocks uses the same registers as above except that it uses AX to store +// the d pointer. + +// func writeBlocks(d *Digest, b []byte) int +TEXT ·writeBlocks(SB), NOSPLIT, $0-40 + // Load fixed primes needed for round. + MOVQ ·prime1v(SB), R13 + MOVQ ·prime2v(SB), R14 + + // Load slice. + MOVQ b_base+8(FP), SI + MOVQ b_len+16(FP), DX + LEAQ (SI)(DX*1), BX + SUBQ $32, BX + + // Load vN from d. + MOVQ d+0(FP), AX + MOVQ 0(AX), R8 // v1 + MOVQ 8(AX), R9 // v2 + MOVQ 16(AX), R10 // v3 + MOVQ 24(AX), R11 // v4 + + // We don't need to check the loop condition here; this function is + // always called with at least one block of data to process. +blockLoop: + round(R8) + round(R9) + round(R10) + round(R11) + + CMPQ SI, BX + JLE blockLoop + + // Copy vN back to d. + MOVQ R8, 0(AX) + MOVQ R9, 8(AX) + MOVQ R10, 16(AX) + MOVQ R11, 24(AX) + + // The number of bytes written is SI minus the old base pointer. + SUBQ b_base+8(FP), SI + MOVQ SI, ret+32(FP) + + RET diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_arm64.s b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_arm64.s new file mode 100644 index 0000000..4d64a17 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_arm64.s @@ -0,0 +1,186 @@ +// +build gc,!purego,!noasm + +#include "textflag.h" + +// Register allocation. +#define digest R1 +#define h R2 // Return value. +#define p R3 // Input pointer. +#define len R4 +#define nblocks R5 // len / 32. +#define prime1 R7 +#define prime2 R8 +#define prime3 R9 +#define prime4 R10 +#define prime5 R11 +#define v1 R12 +#define v2 R13 +#define v3 R14 +#define v4 R15 +#define x1 R20 +#define x2 R21 +#define x3 R22 +#define x4 R23 + +#define round(acc, x) \ + MADD prime2, acc, x, acc \ + ROR $64-31, acc \ + MUL prime1, acc \ + +// x = round(0, x). +#define round0(x) \ + MUL prime2, x \ + ROR $64-31, x \ + MUL prime1, x \ + +#define mergeRound(x) \ + round0(x) \ + EOR x, h \ + MADD h, prime4, prime1, h \ + +// Update v[1-4] with 32-byte blocks. Assumes len >= 32. +#define blocksLoop() \ + LSR $5, len, nblocks \ + PCALIGN $16 \ + loop: \ + LDP.P 32(p), (x1, x2) \ + round(v1, x1) \ + LDP -16(p), (x3, x4) \ + round(v2, x2) \ + SUB $1, nblocks \ + round(v3, x3) \ + round(v4, x4) \ + CBNZ nblocks, loop \ + +// The primes are repeated here to ensure that they're stored +// in a contiguous array, so we can load them with LDP. +DATA primes<> +0(SB)/8, $11400714785074694791 +DATA primes<> +8(SB)/8, $14029467366897019727 +DATA primes<>+16(SB)/8, $1609587929392839161 +DATA primes<>+24(SB)/8, $9650029242287828579 +DATA primes<>+32(SB)/8, $2870177450012600261 +GLOBL primes<>(SB), NOPTR+RODATA, $40 + +// func Sum64(b []byte) uint64 +TEXT ·Sum64(SB), NOFRAME+NOSPLIT, $0-32 + LDP b_base+0(FP), (p, len) + + LDP primes<> +0(SB), (prime1, prime2) + LDP primes<>+16(SB), (prime3, prime4) + MOVD primes<>+32(SB), prime5 + + CMP $32, len + CSEL LO, prime5, ZR, h // if len < 32 { h = prime5 } else { h = 0 } + BLO afterLoop + + ADD prime1, prime2, v1 + MOVD prime2, v2 + MOVD $0, v3 + NEG prime1, v4 + + blocksLoop() + + ROR $64-1, v1, x1 + ROR $64-7, v2, x2 + ADD x1, x2 + ROR $64-12, v3, x3 + ROR $64-18, v4, x4 + ADD x3, x4 + ADD x2, x4, h + + mergeRound(v1) + mergeRound(v2) + mergeRound(v3) + mergeRound(v4) + +afterLoop: + ADD len, h + + TBZ $4, len, try8 + LDP.P 16(p), (x1, x2) + + round0(x1) + ROR $64-27, h + EOR x1 @> 64-27, h, h + MADD h, prime4, prime1, h + + round0(x2) + ROR $64-27, h + EOR x2 @> 64-27, h + MADD h, prime4, prime1, h + +try8: + TBZ $3, len, try4 + MOVD.P 8(p), x1 + + round0(x1) + ROR $64-27, h + EOR x1 @> 64-27, h + MADD h, prime4, prime1, h + +try4: + TBZ $2, len, try2 + MOVWU.P 4(p), x2 + + MUL prime1, x2 + ROR $64-23, h + EOR x2 @> 64-23, h + MADD h, prime3, prime2, h + +try2: + TBZ $1, len, try1 + MOVHU.P 2(p), x3 + AND $255, x3, x1 + LSR $8, x3, x2 + + MUL prime5, x1 + ROR $64-11, h + EOR x1 @> 64-11, h + MUL prime1, h + + MUL prime5, x2 + ROR $64-11, h + EOR x2 @> 64-11, h + MUL prime1, h + +try1: + TBZ $0, len, end + MOVBU (p), x4 + + MUL prime5, x4 + ROR $64-11, h + EOR x4 @> 64-11, h + MUL prime1, h + +end: + EOR h >> 33, h + MUL prime2, h + EOR h >> 29, h + MUL prime3, h + EOR h >> 32, h + + MOVD h, ret+24(FP) + RET + +// func writeBlocks(d *Digest, b []byte) int +// +// Assumes len(b) >= 32. +TEXT ·writeBlocks(SB), NOFRAME+NOSPLIT, $0-40 + LDP primes<>(SB), (prime1, prime2) + + // Load state. Assume v[1-4] are stored contiguously. + MOVD d+0(FP), digest + LDP 0(digest), (v1, v2) + LDP 16(digest), (v3, v4) + + LDP b_base+8(FP), (p, len) + + blocksLoop() + + // Store updated state. + STP (v1, v2), 0(digest) + STP (v3, v4), 16(digest) + + BIC $31, len + MOVD len, ret+32(FP) + RET diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_asm.go b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_asm.go new file mode 100644 index 0000000..1a1fac9 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_asm.go @@ -0,0 +1,16 @@ +//go:build (amd64 || arm64) && !appengine && gc && !purego && !noasm +// +build amd64 arm64 +// +build !appengine +// +build gc +// +build !purego +// +build !noasm + +package xxhash + +// Sum64 computes the 64-bit xxHash digest of b. +// +//go:noescape +func Sum64(b []byte) uint64 + +//go:noescape +func writeBlocks(d *Digest, b []byte) int diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_other.go b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_other.go new file mode 100644 index 0000000..209cb4a --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_other.go @@ -0,0 +1,77 @@ +//go:build (!amd64 && !arm64) || appengine || !gc || purego || noasm +// +build !amd64,!arm64 appengine !gc purego noasm + +package xxhash + +// Sum64 computes the 64-bit xxHash digest of b. +func Sum64(b []byte) uint64 { + // A simpler version would be + // d := New() + // d.Write(b) + // return d.Sum64() + // but this is faster, particularly for small inputs. + + n := len(b) + var h uint64 + + if n >= 32 { + v1 := prime1v + prime2 + v2 := prime2 + v3 := uint64(0) + v4 := -prime1v + for len(b) >= 32 { + v1 = round(v1, u64(b[0:8:len(b)])) + v2 = round(v2, u64(b[8:16:len(b)])) + v3 = round(v3, u64(b[16:24:len(b)])) + v4 = round(v4, u64(b[24:32:len(b)])) + b = b[32:len(b):len(b)] + } + h = rol1(v1) + rol7(v2) + rol12(v3) + rol18(v4) + h = mergeRound(h, v1) + h = mergeRound(h, v2) + h = mergeRound(h, v3) + h = mergeRound(h, v4) + } else { + h = prime5 + } + + h += uint64(n) + + i, end := 0, len(b) + for ; i+8 <= end; i += 8 { + k1 := round(0, u64(b[i:i+8:len(b)])) + h ^= k1 + h = rol27(h)*prime1 + prime4 + } + if i+4 <= end { + h ^= uint64(u32(b[i:i+4:len(b)])) * prime1 + h = rol23(h)*prime2 + prime3 + i += 4 + } + for ; i < end; i++ { + h ^= uint64(b[i]) * prime5 + h = rol11(h) * prime1 + } + + h ^= h >> 33 + h *= prime2 + h ^= h >> 29 + h *= prime3 + h ^= h >> 32 + + return h +} + +func writeBlocks(d *Digest, b []byte) int { + v1, v2, v3, v4 := d.v1, d.v2, d.v3, d.v4 + n := len(b) + for len(b) >= 32 { + v1 = round(v1, u64(b[0:8:len(b)])) + v2 = round(v2, u64(b[8:16:len(b)])) + v3 = round(v3, u64(b[16:24:len(b)])) + v4 = round(v4, u64(b[24:32:len(b)])) + b = b[32:len(b):len(b)] + } + d.v1, d.v2, d.v3, d.v4 = v1, v2, v3, v4 + return n - len(b) +} diff --git a/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_safe.go b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_safe.go new file mode 100644 index 0000000..6f3b0cb --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_safe.go @@ -0,0 +1,11 @@ +package xxhash + +// Sum64String computes the 64-bit xxHash digest of s. +func Sum64String(s string) uint64 { + return Sum64([]byte(s)) +} + +// WriteString adds more data to d. It always returns len(s), nil. +func (d *Digest) WriteString(s string) (n int, err error) { + return d.Write([]byte(s)) +} diff --git a/vendor/github.com/klauspost/compress/zstd/seqdec.go b/vendor/github.com/klauspost/compress/zstd/seqdec.go new file mode 100644 index 0000000..819f146 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/seqdec.go @@ -0,0 +1,715 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "errors" + "fmt" + "io" +) + +type seq struct { + litLen uint32 + matchLen uint32 + offset uint32 + + // Codes are stored here for the encoder + // so they only have to be looked up once. + llCode, mlCode, ofCode uint8 +} + +type seqVals struct { + ll, ml, mo int +} + +func (s seq) String() string { + if s.offset <= 3 { + if s.offset == 0 { + return fmt.Sprint("litLen:", s.litLen, ", matchLen:", s.matchLen+zstdMinMatch, ", offset: INVALID (0)") + } + return fmt.Sprint("litLen:", s.litLen, ", matchLen:", s.matchLen+zstdMinMatch, ", offset:", s.offset, " (repeat)") + } + return fmt.Sprint("litLen:", s.litLen, ", matchLen:", s.matchLen+zstdMinMatch, ", offset:", s.offset-3, " (new)") +} + +type seqCompMode uint8 + +const ( + compModePredefined seqCompMode = iota + compModeRLE + compModeFSE + compModeRepeat +) + +type sequenceDec struct { + // decoder keeps track of the current state and updates it from the bitstream. + fse *fseDecoder + state fseState + repeat bool +} + +// init the state of the decoder with input from stream. +func (s *sequenceDec) init(br *bitReader) error { + if s.fse == nil { + return errors.New("sequence decoder not defined") + } + s.state.init(br, s.fse.actualTableLog, s.fse.dt[:1< 4+((maxOffsetBits+16+16)>>3) { + // inlined function: + // ll, mo, ml = s.nextFast(br, llState, mlState, ofState) + + // Final will not read from stream. + var llB, mlB, moB uint8 + ll, llB = llState.final() + ml, mlB = mlState.final() + mo, moB = ofState.final() + + // extra bits are stored in reverse order. + br.fillFast() + mo += br.getBits(moB) + if s.maxBits > 32 { + br.fillFast() + } + ml += br.getBits(mlB) + ll += br.getBits(llB) + + if moB > 1 { + s.prevOffset[2] = s.prevOffset[1] + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = mo + } else { + // mo = s.adjustOffset(mo, ll, moB) + // Inlined for rather big speedup + if ll == 0 { + // There is an exception though, when current sequence's literals_length = 0. + // In this case, repeated offsets are shifted by one, so an offset_value of 1 means Repeated_Offset2, + // an offset_value of 2 means Repeated_Offset3, and an offset_value of 3 means Repeated_Offset1 - 1_byte. + mo++ + } + + if mo == 0 { + mo = s.prevOffset[0] + } else { + var temp int + if mo == 3 { + temp = s.prevOffset[0] - 1 + } else { + temp = s.prevOffset[mo] + } + + if temp == 0 { + // 0 is not valid; input is corrupted; force offset to 1 + println("WARNING: temp was 0") + temp = 1 + } + + if mo != 1 { + s.prevOffset[2] = s.prevOffset[1] + } + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = temp + mo = temp + } + } + br.fillFast() + } else { + if br.overread() { + if debugDecoder { + printf("reading sequence %d, exceeded available data\n", i) + } + return io.ErrUnexpectedEOF + } + ll, mo, ml = s.next(br, llState, mlState, ofState) + br.fill() + } + + if debugSequences { + println("Seq", i, "Litlen:", ll, "mo:", mo, "(abs) ml:", ml) + } + // Evaluate. + // We might be doing this async, so do it early. + if mo == 0 && ml > 0 { + return fmt.Errorf("zero matchoff and matchlen (%d) > 0", ml) + } + if ml > maxMatchLen { + return fmt.Errorf("match len (%d) bigger than max allowed length", ml) + } + s.seqSize += ll + ml + if s.seqSize > maxBlockSize { + return fmt.Errorf("output (%d) bigger than max block size (%d)", s.seqSize, maxBlockSize) + } + litRemain -= ll + if litRemain < 0 { + return fmt.Errorf("unexpected literal count, want %d bytes, but only %d is available", ll, litRemain+ll) + } + seqs[i] = seqVals{ + ll: ll, + ml: ml, + mo: mo, + } + if i == len(seqs)-1 { + // This is the last sequence, so we shouldn't update state. + break + } + + // Manually inlined, ~ 5-20% faster + // Update all 3 states at once. Approx 20% faster. + nBits := llState.nbBits() + mlState.nbBits() + ofState.nbBits() + if nBits == 0 { + llState = llTable[llState.newState()&maxTableMask] + mlState = mlTable[mlState.newState()&maxTableMask] + ofState = ofTable[ofState.newState()&maxTableMask] + } else { + bits := br.get32BitsFast(nBits) + lowBits := uint16(bits >> ((ofState.nbBits() + mlState.nbBits()) & 31)) + llState = llTable[(llState.newState()+lowBits)&maxTableMask] + + lowBits = uint16(bits >> (ofState.nbBits() & 31)) + lowBits &= bitMask[mlState.nbBits()&15] + mlState = mlTable[(mlState.newState()+lowBits)&maxTableMask] + + lowBits = uint16(bits) & bitMask[ofState.nbBits()&15] + ofState = ofTable[(ofState.newState()+lowBits)&maxTableMask] + } + } + s.seqSize += litRemain + if s.seqSize > maxBlockSize { + return fmt.Errorf("output (%d) bigger than max block size (%d)", s.seqSize, maxBlockSize) + } + err := br.close() + if err != nil { + printf("Closing sequences: %v, %+v\n", err, *br) + } + return err +} + +// execute will execute the decoded sequence with the provided history. +// The sequence must be evaluated before being sent. +func (s *sequenceDecs) execute(seqs []seqVals, hist []byte) error { + // Ensure we have enough output size... + if len(s.out)+s.seqSize > cap(s.out) { + addBytes := s.seqSize + len(s.out) + s.out = append(s.out, make([]byte, addBytes)...) + s.out = s.out[:len(s.out)-addBytes] + } + + if debugDecoder { + printf("Execute %d seqs with hist %d, dict %d, literals: %d into %d bytes\n", len(seqs), len(hist), len(s.dict), len(s.literals), s.seqSize) + } + + var t = len(s.out) + out := s.out[:t+s.seqSize] + + for _, seq := range seqs { + // Add literals + copy(out[t:], s.literals[:seq.ll]) + t += seq.ll + s.literals = s.literals[seq.ll:] + + // Copy from dictionary... + if seq.mo > t+len(hist) || seq.mo > s.windowSize { + if len(s.dict) == 0 { + return fmt.Errorf("match offset (%d) bigger than current history (%d)", seq.mo, t+len(hist)) + } + + // we may be in dictionary. + dictO := len(s.dict) - (seq.mo - (t + len(hist))) + if dictO < 0 || dictO >= len(s.dict) { + return fmt.Errorf("match offset (%d) bigger than current history+dict (%d)", seq.mo, t+len(hist)+len(s.dict)) + } + end := dictO + seq.ml + if end > len(s.dict) { + n := len(s.dict) - dictO + copy(out[t:], s.dict[dictO:]) + t += n + seq.ml -= n + } else { + copy(out[t:], s.dict[dictO:end]) + t += end - dictO + continue + } + } + + // Copy from history. + if v := seq.mo - t; v > 0 { + // v is the start position in history from end. + start := len(hist) - v + if seq.ml > v { + // Some goes into current block. + // Copy remainder of history + copy(out[t:], hist[start:]) + t += v + seq.ml -= v + } else { + copy(out[t:], hist[start:start+seq.ml]) + t += seq.ml + continue + } + } + // We must be in current buffer now + if seq.ml > 0 { + start := t - seq.mo + if seq.ml <= t-start { + // No overlap + copy(out[t:], out[start:start+seq.ml]) + t += seq.ml + continue + } else { + // Overlapping copy + // Extend destination slice and copy one byte at the time. + src := out[start : start+seq.ml] + dst := out[t:] + dst = dst[:len(src)] + t += len(src) + // Destination is the space we just added. + for i := range src { + dst[i] = src[i] + } + } + } + } + // Add final literals + copy(out[t:], s.literals) + if debugDecoder { + t += len(s.literals) + if t != len(out) { + panic(fmt.Errorf("length mismatch, want %d, got %d, ss: %d", len(out), t, s.seqSize)) + } + } + s.out = out + + return nil +} + +// decode sequences from the stream with the provided history. +func (s *sequenceDecs) decodeSync(history *history) error { + br := s.br + seqs := s.nSeqs + startSize := len(s.out) + // Grab full sizes tables, to avoid bounds checks. + llTable, mlTable, ofTable := s.litLengths.fse.dt[:maxTablesize], s.matchLengths.fse.dt[:maxTablesize], s.offsets.fse.dt[:maxTablesize] + llState, mlState, ofState := s.litLengths.state.state, s.matchLengths.state.state, s.offsets.state.state + hist := history.b[history.ignoreBuffer:] + out := s.out + maxBlockSize := maxCompressedBlockSize + if s.windowSize < maxBlockSize { + maxBlockSize = s.windowSize + } + + for i := seqs - 1; i >= 0; i-- { + if br.overread() { + printf("reading sequence %d, exceeded available data\n", seqs-i) + return io.ErrUnexpectedEOF + } + var ll, mo, ml int + if br.off > 4+((maxOffsetBits+16+16)>>3) { + // inlined function: + // ll, mo, ml = s.nextFast(br, llState, mlState, ofState) + + // Final will not read from stream. + var llB, mlB, moB uint8 + ll, llB = llState.final() + ml, mlB = mlState.final() + mo, moB = ofState.final() + + // extra bits are stored in reverse order. + br.fillFast() + mo += br.getBits(moB) + if s.maxBits > 32 { + br.fillFast() + } + ml += br.getBits(mlB) + ll += br.getBits(llB) + + if moB > 1 { + s.prevOffset[2] = s.prevOffset[1] + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = mo + } else { + // mo = s.adjustOffset(mo, ll, moB) + // Inlined for rather big speedup + if ll == 0 { + // There is an exception though, when current sequence's literals_length = 0. + // In this case, repeated offsets are shifted by one, so an offset_value of 1 means Repeated_Offset2, + // an offset_value of 2 means Repeated_Offset3, and an offset_value of 3 means Repeated_Offset1 - 1_byte. + mo++ + } + + if mo == 0 { + mo = s.prevOffset[0] + } else { + var temp int + if mo == 3 { + temp = s.prevOffset[0] - 1 + } else { + temp = s.prevOffset[mo] + } + + if temp == 0 { + // 0 is not valid; input is corrupted; force offset to 1 + println("WARNING: temp was 0") + temp = 1 + } + + if mo != 1 { + s.prevOffset[2] = s.prevOffset[1] + } + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = temp + mo = temp + } + } + br.fillFast() + } else { + ll, mo, ml = s.next(br, llState, mlState, ofState) + br.fill() + } + + if debugSequences { + println("Seq", seqs-i-1, "Litlen:", ll, "mo:", mo, "(abs) ml:", ml) + } + + if ll > len(s.literals) { + return fmt.Errorf("unexpected literal count, want %d bytes, but only %d is available", ll, len(s.literals)) + } + size := ll + ml + len(out) + if size-startSize > maxBlockSize { + return fmt.Errorf("output (%d) bigger than max block size (%d)", size, maxBlockSize) + } + if size > cap(out) { + // Not enough size, which can happen under high volume block streaming conditions + // but could be if destination slice is too small for sync operations. + // over-allocating here can create a large amount of GC pressure so we try to keep + // it as contained as possible + used := len(out) - startSize + addBytes := 256 + ll + ml + used>>2 + // Clamp to max block size. + if used+addBytes > maxBlockSize { + addBytes = maxBlockSize - used + } + out = append(out, make([]byte, addBytes)...) + out = out[:len(out)-addBytes] + } + if ml > maxMatchLen { + return fmt.Errorf("match len (%d) bigger than max allowed length", ml) + } + + // Add literals + out = append(out, s.literals[:ll]...) + s.literals = s.literals[ll:] + + if mo == 0 && ml > 0 { + return fmt.Errorf("zero matchoff and matchlen (%d) > 0", ml) + } + + if mo > len(out)+len(hist) || mo > s.windowSize { + if len(s.dict) == 0 { + return fmt.Errorf("match offset (%d) bigger than current history (%d)", mo, len(out)+len(hist)) + } + + // we may be in dictionary. + dictO := len(s.dict) - (mo - (len(out) + len(hist))) + if dictO < 0 || dictO >= len(s.dict) { + return fmt.Errorf("match offset (%d) bigger than current history (%d)", mo, len(out)+len(hist)) + } + end := dictO + ml + if end > len(s.dict) { + out = append(out, s.dict[dictO:]...) + ml -= len(s.dict) - dictO + } else { + out = append(out, s.dict[dictO:end]...) + mo = 0 + ml = 0 + } + } + + // Copy from history. + // TODO: Blocks without history could be made to ignore this completely. + if v := mo - len(out); v > 0 { + // v is the start position in history from end. + start := len(hist) - v + if ml > v { + // Some goes into current block. + // Copy remainder of history + out = append(out, hist[start:]...) + ml -= v + } else { + out = append(out, hist[start:start+ml]...) + ml = 0 + } + } + // We must be in current buffer now + if ml > 0 { + start := len(out) - mo + if ml <= len(out)-start { + // No overlap + out = append(out, out[start:start+ml]...) + } else { + // Overlapping copy + // Extend destination slice and copy one byte at the time. + out = out[:len(out)+ml] + src := out[start : start+ml] + // Destination is the space we just added. + dst := out[len(out)-ml:] + dst = dst[:len(src)] + for i := range src { + dst[i] = src[i] + } + } + } + if i == 0 { + // This is the last sequence, so we shouldn't update state. + break + } + + // Manually inlined, ~ 5-20% faster + // Update all 3 states at once. Approx 20% faster. + nBits := llState.nbBits() + mlState.nbBits() + ofState.nbBits() + if nBits == 0 { + llState = llTable[llState.newState()&maxTableMask] + mlState = mlTable[mlState.newState()&maxTableMask] + ofState = ofTable[ofState.newState()&maxTableMask] + } else { + bits := br.get32BitsFast(nBits) + lowBits := uint16(bits >> ((ofState.nbBits() + mlState.nbBits()) & 31)) + llState = llTable[(llState.newState()+lowBits)&maxTableMask] + + lowBits = uint16(bits >> (ofState.nbBits() & 31)) + lowBits &= bitMask[mlState.nbBits()&15] + mlState = mlTable[(mlState.newState()+lowBits)&maxTableMask] + + lowBits = uint16(bits) & bitMask[ofState.nbBits()&15] + ofState = ofTable[(ofState.newState()+lowBits)&maxTableMask] + } + } + + // Check if space for literals + if len(s.literals)+len(s.out)-startSize > maxBlockSize { + return fmt.Errorf("output (%d) bigger than max block size (%d)", len(s.out), maxBlockSize) + } + + // Add final literals + s.out = append(out, s.literals...) + return br.close() +} + +// update states, at least 27 bits must be available. +func (s *sequenceDecs) update(br *bitReader) { + // Max 8 bits + s.litLengths.state.next(br) + // Max 9 bits + s.matchLengths.state.next(br) + // Max 8 bits + s.offsets.state.next(br) +} + +var bitMask [16]uint16 + +func init() { + for i := range bitMask[:] { + bitMask[i] = uint16((1 << uint(i)) - 1) + } +} + +// update states, at least 27 bits must be available. +func (s *sequenceDecs) updateAlt(br *bitReader) { + // Update all 3 states at once. Approx 20% faster. + a, b, c := s.litLengths.state.state, s.matchLengths.state.state, s.offsets.state.state + + nBits := a.nbBits() + b.nbBits() + c.nbBits() + if nBits == 0 { + s.litLengths.state.state = s.litLengths.state.dt[a.newState()] + s.matchLengths.state.state = s.matchLengths.state.dt[b.newState()] + s.offsets.state.state = s.offsets.state.dt[c.newState()] + return + } + bits := br.get32BitsFast(nBits) + lowBits := uint16(bits >> ((c.nbBits() + b.nbBits()) & 31)) + s.litLengths.state.state = s.litLengths.state.dt[a.newState()+lowBits] + + lowBits = uint16(bits >> (c.nbBits() & 31)) + lowBits &= bitMask[b.nbBits()&15] + s.matchLengths.state.state = s.matchLengths.state.dt[b.newState()+lowBits] + + lowBits = uint16(bits) & bitMask[c.nbBits()&15] + s.offsets.state.state = s.offsets.state.dt[c.newState()+lowBits] +} + +// nextFast will return new states when there are at least 4 unused bytes left on the stream when done. +func (s *sequenceDecs) nextFast(br *bitReader, llState, mlState, ofState decSymbol) (ll, mo, ml int) { + // Final will not read from stream. + ll, llB := llState.final() + ml, mlB := mlState.final() + mo, moB := ofState.final() + + // extra bits are stored in reverse order. + br.fillFast() + mo += br.getBits(moB) + if s.maxBits > 32 { + br.fillFast() + } + ml += br.getBits(mlB) + ll += br.getBits(llB) + + if moB > 1 { + s.prevOffset[2] = s.prevOffset[1] + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = mo + return + } + // mo = s.adjustOffset(mo, ll, moB) + // Inlined for rather big speedup + if ll == 0 { + // There is an exception though, when current sequence's literals_length = 0. + // In this case, repeated offsets are shifted by one, so an offset_value of 1 means Repeated_Offset2, + // an offset_value of 2 means Repeated_Offset3, and an offset_value of 3 means Repeated_Offset1 - 1_byte. + mo++ + } + + if mo == 0 { + mo = s.prevOffset[0] + return + } + var temp int + if mo == 3 { + temp = s.prevOffset[0] - 1 + } else { + temp = s.prevOffset[mo] + } + + if temp == 0 { + // 0 is not valid; input is corrupted; force offset to 1 + println("temp was 0") + temp = 1 + } + + if mo != 1 { + s.prevOffset[2] = s.prevOffset[1] + } + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = temp + mo = temp + return +} + +func (s *sequenceDecs) next(br *bitReader, llState, mlState, ofState decSymbol) (ll, mo, ml int) { + // Final will not read from stream. + ll, llB := llState.final() + ml, mlB := mlState.final() + mo, moB := ofState.final() + + // extra bits are stored in reverse order. + br.fill() + if s.maxBits <= 32 { + mo += br.getBits(moB) + ml += br.getBits(mlB) + ll += br.getBits(llB) + } else { + mo += br.getBits(moB) + br.fill() + // matchlength+literal length, max 32 bits + ml += br.getBits(mlB) + ll += br.getBits(llB) + + } + mo = s.adjustOffset(mo, ll, moB) + return +} + +func (s *sequenceDecs) adjustOffset(offset, litLen int, offsetB uint8) int { + if offsetB > 1 { + s.prevOffset[2] = s.prevOffset[1] + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = offset + return offset + } + + if litLen == 0 { + // There is an exception though, when current sequence's literals_length = 0. + // In this case, repeated offsets are shifted by one, so an offset_value of 1 means Repeated_Offset2, + // an offset_value of 2 means Repeated_Offset3, and an offset_value of 3 means Repeated_Offset1 - 1_byte. + offset++ + } + + if offset == 0 { + return s.prevOffset[0] + } + var temp int + if offset == 3 { + temp = s.prevOffset[0] - 1 + } else { + temp = s.prevOffset[offset] + } + + if temp == 0 { + // 0 is not valid; input is corrupted; force offset to 1 + println("temp was 0") + temp = 1 + } + + if offset != 1 { + s.prevOffset[2] = s.prevOffset[1] + } + s.prevOffset[1] = s.prevOffset[0] + s.prevOffset[0] = temp + return temp +} diff --git a/vendor/github.com/klauspost/compress/zstd/seqenc.go b/vendor/github.com/klauspost/compress/zstd/seqenc.go new file mode 100644 index 0000000..8014174 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/seqenc.go @@ -0,0 +1,114 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import "math/bits" + +type seqCoders struct { + llEnc, ofEnc, mlEnc *fseEncoder + llPrev, ofPrev, mlPrev *fseEncoder +} + +// swap coders with another (block). +func (s *seqCoders) swap(other *seqCoders) { + *s, *other = *other, *s +} + +// setPrev will update the previous encoders to the actually used ones +// and make sure a fresh one is in the main slot. +func (s *seqCoders) setPrev(ll, ml, of *fseEncoder) { + compareSwap := func(used *fseEncoder, current, prev **fseEncoder) { + // We used the new one, more current to history and reuse the previous history + if *current == used { + *prev, *current = *current, *prev + c := *current + p := *prev + c.reUsed = false + p.reUsed = true + return + } + if used == *prev { + return + } + // Ensure we cannot reuse by accident + prevEnc := *prev + prevEnc.symbolLen = 0 + } + compareSwap(ll, &s.llEnc, &s.llPrev) + compareSwap(ml, &s.mlEnc, &s.mlPrev) + compareSwap(of, &s.ofEnc, &s.ofPrev) +} + +func highBit(val uint32) (n uint32) { + return uint32(bits.Len32(val) - 1) +} + +var llCodeTable = [64]byte{0, 1, 2, 3, 4, 5, 6, 7, + 8, 9, 10, 11, 12, 13, 14, 15, + 16, 16, 17, 17, 18, 18, 19, 19, + 20, 20, 20, 20, 21, 21, 21, 21, + 22, 22, 22, 22, 22, 22, 22, 22, + 23, 23, 23, 23, 23, 23, 23, 23, + 24, 24, 24, 24, 24, 24, 24, 24, + 24, 24, 24, 24, 24, 24, 24, 24} + +// Up to 6 bits +const maxLLCode = 35 + +// llBitsTable translates from ll code to number of bits. +var llBitsTable = [maxLLCode + 1]byte{ + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 1, 1, 1, 1, 2, 2, 3, 3, + 4, 6, 7, 8, 9, 10, 11, 12, + 13, 14, 15, 16} + +// llCode returns the code that represents the literal length requested. +func llCode(litLength uint32) uint8 { + const llDeltaCode = 19 + if litLength <= 63 { + // Compiler insists on bounds check (Go 1.12) + return llCodeTable[litLength&63] + } + return uint8(highBit(litLength)) + llDeltaCode +} + +var mlCodeTable = [128]byte{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, + 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, + 32, 32, 33, 33, 34, 34, 35, 35, 36, 36, 36, 36, 37, 37, 37, 37, + 38, 38, 38, 38, 38, 38, 38, 38, 39, 39, 39, 39, 39, 39, 39, 39, + 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, + 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, + 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, + 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42} + +// Up to 6 bits +const maxMLCode = 52 + +// mlBitsTable translates from ml code to number of bits. +var mlBitsTable = [maxMLCode + 1]byte{ + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 1, 1, 1, 1, 2, 2, 3, 3, + 4, 4, 5, 7, 8, 9, 10, 11, + 12, 13, 14, 15, 16} + +// note : mlBase = matchLength - MINMATCH; +// because it's the format it's stored in seqStore->sequences +func mlCode(mlBase uint32) uint8 { + const mlDeltaCode = 36 + if mlBase <= 127 { + // Compiler insists on bounds check (Go 1.12) + return mlCodeTable[mlBase&127] + } + return uint8(highBit(mlBase)) + mlDeltaCode +} + +func ofCode(offset uint32) uint8 { + // A valid offset will always be > 0. + return uint8(bits.Len32(offset) - 1) +} diff --git a/vendor/github.com/klauspost/compress/zstd/snappy.go b/vendor/github.com/klauspost/compress/zstd/snappy.go new file mode 100644 index 0000000..9e1baad --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/snappy.go @@ -0,0 +1,435 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. +// Based on work by Yann Collet, released under BSD License. + +package zstd + +import ( + "encoding/binary" + "errors" + "hash/crc32" + "io" + + "github.com/klauspost/compress/huff0" + snappy "github.com/klauspost/compress/internal/snapref" +) + +const ( + snappyTagLiteral = 0x00 + snappyTagCopy1 = 0x01 + snappyTagCopy2 = 0x02 + snappyTagCopy4 = 0x03 +) + +const ( + snappyChecksumSize = 4 + snappyMagicBody = "sNaPpY" + + // snappyMaxBlockSize is the maximum size of the input to encodeBlock. It is not + // part of the wire format per se, but some parts of the encoder assume + // that an offset fits into a uint16. + // + // Also, for the framing format (Writer type instead of Encode function), + // https://github.com/google/snappy/blob/master/framing_format.txt says + // that "the uncompressed data in a chunk must be no longer than 65536 + // bytes". + snappyMaxBlockSize = 65536 + + // snappyMaxEncodedLenOfMaxBlockSize equals MaxEncodedLen(snappyMaxBlockSize), but is + // hard coded to be a const instead of a variable, so that obufLen can also + // be a const. Their equivalence is confirmed by + // TestMaxEncodedLenOfMaxBlockSize. + snappyMaxEncodedLenOfMaxBlockSize = 76490 +) + +const ( + chunkTypeCompressedData = 0x00 + chunkTypeUncompressedData = 0x01 + chunkTypePadding = 0xfe + chunkTypeStreamIdentifier = 0xff +) + +var ( + // ErrSnappyCorrupt reports that the input is invalid. + ErrSnappyCorrupt = errors.New("snappy: corrupt input") + // ErrSnappyTooLarge reports that the uncompressed length is too large. + ErrSnappyTooLarge = errors.New("snappy: decoded block is too large") + // ErrSnappyUnsupported reports that the input isn't supported. + ErrSnappyUnsupported = errors.New("snappy: unsupported input") + + errUnsupportedLiteralLength = errors.New("snappy: unsupported literal length") +) + +// SnappyConverter can read SnappyConverter-compressed streams and convert them to zstd. +// Conversion is done by converting the stream directly from Snappy without intermediate +// full decoding. +// Therefore the compression ratio is much less than what can be done by a full decompression +// and compression, and a faulty Snappy stream may lead to a faulty Zstandard stream without +// any errors being generated. +// No CRC value is being generated and not all CRC values of the Snappy stream are checked. +// However, it provides really fast recompression of Snappy streams. +// The converter can be reused to avoid allocations, even after errors. +type SnappyConverter struct { + r io.Reader + err error + buf []byte + block *blockEnc +} + +// Convert the Snappy stream supplied in 'in' and write the zStandard stream to 'w'. +// If any error is detected on the Snappy stream it is returned. +// The number of bytes written is returned. +func (r *SnappyConverter) Convert(in io.Reader, w io.Writer) (int64, error) { + initPredefined() + r.err = nil + r.r = in + if r.block == nil { + r.block = &blockEnc{} + r.block.init() + } + r.block.initNewEncode() + if len(r.buf) != snappyMaxEncodedLenOfMaxBlockSize+snappyChecksumSize { + r.buf = make([]byte, snappyMaxEncodedLenOfMaxBlockSize+snappyChecksumSize) + } + r.block.litEnc.Reuse = huff0.ReusePolicyNone + var written int64 + var readHeader bool + { + var header []byte + var n int + header, r.err = frameHeader{WindowSize: snappyMaxBlockSize}.appendTo(r.buf[:0]) + + n, r.err = w.Write(header) + if r.err != nil { + return written, r.err + } + written += int64(n) + } + + for { + if !r.readFull(r.buf[:4], true) { + // Add empty last block + r.block.reset(nil) + r.block.last = true + err := r.block.encodeLits(r.block.literals, false) + if err != nil { + return written, err + } + n, err := w.Write(r.block.output) + if err != nil { + return written, err + } + written += int64(n) + + return written, r.err + } + chunkType := r.buf[0] + if !readHeader { + if chunkType != chunkTypeStreamIdentifier { + println("chunkType != chunkTypeStreamIdentifier", chunkType) + r.err = ErrSnappyCorrupt + return written, r.err + } + readHeader = true + } + chunkLen := int(r.buf[1]) | int(r.buf[2])<<8 | int(r.buf[3])<<16 + if chunkLen > len(r.buf) { + println("chunkLen > len(r.buf)", chunkType) + r.err = ErrSnappyUnsupported + return written, r.err + } + + // The chunk types are specified at + // https://github.com/google/snappy/blob/master/framing_format.txt + switch chunkType { + case chunkTypeCompressedData: + // Section 4.2. Compressed data (chunk type 0x00). + if chunkLen < snappyChecksumSize { + println("chunkLen < snappyChecksumSize", chunkLen, snappyChecksumSize) + r.err = ErrSnappyCorrupt + return written, r.err + } + buf := r.buf[:chunkLen] + if !r.readFull(buf, false) { + return written, r.err + } + //checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24 + buf = buf[snappyChecksumSize:] + + n, hdr, err := snappyDecodedLen(buf) + if err != nil { + r.err = err + return written, r.err + } + buf = buf[hdr:] + if n > snappyMaxBlockSize { + println("n > snappyMaxBlockSize", n, snappyMaxBlockSize) + r.err = ErrSnappyCorrupt + return written, r.err + } + r.block.reset(nil) + r.block.pushOffsets() + if err := decodeSnappy(r.block, buf); err != nil { + r.err = err + return written, r.err + } + if r.block.size+r.block.extraLits != n { + printf("invalid size, want %d, got %d\n", n, r.block.size+r.block.extraLits) + r.err = ErrSnappyCorrupt + return written, r.err + } + err = r.block.encode(nil, false, false) + switch err { + case errIncompressible: + r.block.popOffsets() + r.block.reset(nil) + r.block.literals, err = snappy.Decode(r.block.literals[:n], r.buf[snappyChecksumSize:chunkLen]) + if err != nil { + return written, err + } + err = r.block.encodeLits(r.block.literals, false) + if err != nil { + return written, err + } + case nil: + default: + return written, err + } + + n, r.err = w.Write(r.block.output) + if r.err != nil { + return written, err + } + written += int64(n) + continue + case chunkTypeUncompressedData: + if debugEncoder { + println("Uncompressed, chunklen", chunkLen) + } + // Section 4.3. Uncompressed data (chunk type 0x01). + if chunkLen < snappyChecksumSize { + println("chunkLen < snappyChecksumSize", chunkLen, snappyChecksumSize) + r.err = ErrSnappyCorrupt + return written, r.err + } + r.block.reset(nil) + buf := r.buf[:snappyChecksumSize] + if !r.readFull(buf, false) { + return written, r.err + } + checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24 + // Read directly into r.decoded instead of via r.buf. + n := chunkLen - snappyChecksumSize + if n > snappyMaxBlockSize { + println("n > snappyMaxBlockSize", n, snappyMaxBlockSize) + r.err = ErrSnappyCorrupt + return written, r.err + } + r.block.literals = r.block.literals[:n] + if !r.readFull(r.block.literals, false) { + return written, r.err + } + if snappyCRC(r.block.literals) != checksum { + println("literals crc mismatch") + r.err = ErrSnappyCorrupt + return written, r.err + } + err := r.block.encodeLits(r.block.literals, false) + if err != nil { + return written, err + } + n, r.err = w.Write(r.block.output) + if r.err != nil { + return written, err + } + written += int64(n) + continue + + case chunkTypeStreamIdentifier: + if debugEncoder { + println("stream id", chunkLen, len(snappyMagicBody)) + } + // Section 4.1. Stream identifier (chunk type 0xff). + if chunkLen != len(snappyMagicBody) { + println("chunkLen != len(snappyMagicBody)", chunkLen, len(snappyMagicBody)) + r.err = ErrSnappyCorrupt + return written, r.err + } + if !r.readFull(r.buf[:len(snappyMagicBody)], false) { + return written, r.err + } + for i := 0; i < len(snappyMagicBody); i++ { + if r.buf[i] != snappyMagicBody[i] { + println("r.buf[i] != snappyMagicBody[i]", r.buf[i], snappyMagicBody[i], i) + r.err = ErrSnappyCorrupt + return written, r.err + } + } + continue + } + + if chunkType <= 0x7f { + // Section 4.5. Reserved unskippable chunks (chunk types 0x02-0x7f). + println("chunkType <= 0x7f") + r.err = ErrSnappyUnsupported + return written, r.err + } + // Section 4.4 Padding (chunk type 0xfe). + // Section 4.6. Reserved skippable chunks (chunk types 0x80-0xfd). + if !r.readFull(r.buf[:chunkLen], false) { + return written, r.err + } + } +} + +// decodeSnappy writes the decoding of src to dst. It assumes that the varint-encoded +// length of the decompressed bytes has already been read. +func decodeSnappy(blk *blockEnc, src []byte) error { + //decodeRef(make([]byte, snappyMaxBlockSize), src) + var s, length int + lits := blk.extraLits + var offset uint32 + for s < len(src) { + switch src[s] & 0x03 { + case snappyTagLiteral: + x := uint32(src[s] >> 2) + switch { + case x < 60: + s++ + case x == 60: + s += 2 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, src) + return ErrSnappyCorrupt + } + x = uint32(src[s-1]) + case x == 61: + s += 3 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, src) + return ErrSnappyCorrupt + } + x = uint32(src[s-2]) | uint32(src[s-1])<<8 + case x == 62: + s += 4 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, src) + return ErrSnappyCorrupt + } + x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16 + case x == 63: + s += 5 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, src) + return ErrSnappyCorrupt + } + x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24 + } + if x > snappyMaxBlockSize { + println("x > snappyMaxBlockSize", x, snappyMaxBlockSize) + return ErrSnappyCorrupt + } + length = int(x) + 1 + if length <= 0 { + println("length <= 0 ", length) + + return errUnsupportedLiteralLength + } + //if length > snappyMaxBlockSize-d || uint32(length) > len(src)-s { + // return ErrSnappyCorrupt + //} + + blk.literals = append(blk.literals, src[s:s+length]...) + //println(length, "litLen") + lits += length + s += length + continue + + case snappyTagCopy1: + s += 2 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, len(src)) + return ErrSnappyCorrupt + } + length = 4 + int(src[s-2])>>2&0x7 + offset = uint32(src[s-2])&0xe0<<3 | uint32(src[s-1]) + + case snappyTagCopy2: + s += 3 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, len(src)) + return ErrSnappyCorrupt + } + length = 1 + int(src[s-3])>>2 + offset = uint32(src[s-2]) | uint32(src[s-1])<<8 + + case snappyTagCopy4: + s += 5 + if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line. + println("uint(s) > uint(len(src)", s, len(src)) + return ErrSnappyCorrupt + } + length = 1 + int(src[s-5])>>2 + offset = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24 + } + + if offset <= 0 || blk.size+lits < int(offset) /*|| length > len(blk)-d */ { + println("offset <= 0 || blk.size+lits < int(offset)", offset, blk.size+lits, int(offset), blk.size, lits) + + return ErrSnappyCorrupt + } + + // Check if offset is one of the recent offsets. + // Adjusts the output offset accordingly. + // Gives a tiny bit of compression, typically around 1%. + if false { + offset = blk.matchOffset(offset, uint32(lits)) + } else { + offset += 3 + } + + blk.sequences = append(blk.sequences, seq{ + litLen: uint32(lits), + offset: offset, + matchLen: uint32(length) - zstdMinMatch, + }) + blk.size += length + lits + lits = 0 + } + blk.extraLits = lits + return nil +} + +func (r *SnappyConverter) readFull(p []byte, allowEOF bool) (ok bool) { + if _, r.err = io.ReadFull(r.r, p); r.err != nil { + if r.err == io.ErrUnexpectedEOF || (r.err == io.EOF && !allowEOF) { + r.err = ErrSnappyCorrupt + } + return false + } + return true +} + +var crcTable = crc32.MakeTable(crc32.Castagnoli) + +// crc implements the checksum specified in section 3 of +// https://github.com/google/snappy/blob/master/framing_format.txt +func snappyCRC(b []byte) uint32 { + c := crc32.Update(0, crcTable, b) + return c>>15 | c<<17 + 0xa282ead8 +} + +// snappyDecodedLen returns the length of the decoded block and the number of bytes +// that the length header occupied. +func snappyDecodedLen(src []byte) (blockLen, headerLen int, err error) { + v, n := binary.Uvarint(src) + if n <= 0 || v > 0xffffffff { + return 0, 0, ErrSnappyCorrupt + } + + const wordSize = 32 << (^uint(0) >> 32 & 1) + if wordSize == 32 && v > 0x7fffffff { + return 0, 0, ErrSnappyTooLarge + } + return int(v), n, nil +} diff --git a/vendor/github.com/klauspost/compress/zstd/zip.go b/vendor/github.com/klauspost/compress/zstd/zip.go new file mode 100644 index 0000000..ffffcbc --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/zip.go @@ -0,0 +1,120 @@ +// Copyright 2019+ Klaus Post. All rights reserved. +// License information can be found in the LICENSE file. + +package zstd + +import ( + "errors" + "io" + "sync" +) + +// ZipMethodWinZip is the method for Zstandard compressed data inside Zip files for WinZip. +// See https://www.winzip.com/win/en/comp_info.html +const ZipMethodWinZip = 93 + +// ZipMethodPKWare is the original method number used by PKWARE to indicate Zstandard compression. +// Deprecated: This has been deprecated by PKWARE, use ZipMethodWinZip instead for compression. +// See https://pkware.cachefly.net/webdocs/APPNOTE/APPNOTE-6.3.9.TXT +const ZipMethodPKWare = 20 + +var zipReaderPool sync.Pool + +// newZipReader creates a pooled zip decompressor. +func newZipReader(r io.Reader) io.ReadCloser { + dec, ok := zipReaderPool.Get().(*Decoder) + if ok { + dec.Reset(r) + } else { + d, err := NewReader(r, WithDecoderConcurrency(1), WithDecoderLowmem(true)) + if err != nil { + panic(err) + } + dec = d + } + return &pooledZipReader{dec: dec} +} + +type pooledZipReader struct { + mu sync.Mutex // guards Close and Read + dec *Decoder +} + +func (r *pooledZipReader) Read(p []byte) (n int, err error) { + r.mu.Lock() + defer r.mu.Unlock() + if r.dec == nil { + return 0, errors.New("read after close or EOF") + } + dec, err := r.dec.Read(p) + if err == io.EOF { + err = r.dec.Reset(nil) + zipReaderPool.Put(r.dec) + r.dec = nil + } + return dec, err +} + +func (r *pooledZipReader) Close() error { + r.mu.Lock() + defer r.mu.Unlock() + var err error + if r.dec != nil { + err = r.dec.Reset(nil) + zipReaderPool.Put(r.dec) + r.dec = nil + } + return err +} + +type pooledZipWriter struct { + mu sync.Mutex // guards Close and Read + enc *Encoder + pool *sync.Pool +} + +func (w *pooledZipWriter) Write(p []byte) (n int, err error) { + w.mu.Lock() + defer w.mu.Unlock() + if w.enc == nil { + return 0, errors.New("Write after Close") + } + return w.enc.Write(p) +} + +func (w *pooledZipWriter) Close() error { + w.mu.Lock() + defer w.mu.Unlock() + var err error + if w.enc != nil { + err = w.enc.Close() + w.pool.Put(w.enc) + w.enc = nil + } + return err +} + +// ZipCompressor returns a compressor that can be registered with zip libraries. +// The provided encoder options will be used on all encodes. +func ZipCompressor(opts ...EOption) func(w io.Writer) (io.WriteCloser, error) { + var pool sync.Pool + return func(w io.Writer) (io.WriteCloser, error) { + enc, ok := pool.Get().(*Encoder) + if ok { + enc.Reset(w) + } else { + var err error + enc, err = NewWriter(w, opts...) + if err != nil { + return nil, err + } + } + return &pooledZipWriter{enc: enc, pool: &pool}, nil + } +} + +// ZipDecompressor returns a decompressor that can be registered with zip libraries. +// See ZipCompressor for example. +func ZipDecompressor() func(r io.Reader) io.ReadCloser { + return newZipReader +} diff --git a/vendor/github.com/klauspost/compress/zstd/zstd.go b/vendor/github.com/klauspost/compress/zstd/zstd.go new file mode 100644 index 0000000..c1c90b4 --- /dev/null +++ b/vendor/github.com/klauspost/compress/zstd/zstd.go @@ -0,0 +1,163 @@ +// Package zstd provides decompression of zstandard files. +// +// For advanced usage and examples, go to the README: https://github.com/klauspost/compress/tree/master/zstd#zstd +package zstd + +import ( + "bytes" + "encoding/binary" + "errors" + "log" + "math" + "math/bits" +) + +// enable debug printing +const debug = false + +// enable encoding debug printing +const debugEncoder = debug + +// enable decoding debug printing +const debugDecoder = debug + +// Enable extra assertions. +const debugAsserts = debug || false + +// print sequence details +const debugSequences = false + +// print detailed matching information +const debugMatches = false + +// force encoder to use predefined tables. +const forcePreDef = false + +// zstdMinMatch is the minimum zstd match length. +const zstdMinMatch = 3 + +// Reset the buffer offset when reaching this. +const bufferReset = math.MaxInt32 - MaxWindowSize + +// fcsUnknown is used for unknown frame content size. +const fcsUnknown = math.MaxUint64 + +var ( + // ErrReservedBlockType is returned when a reserved block type is found. + // Typically this indicates wrong or corrupted input. + ErrReservedBlockType = errors.New("invalid input: reserved block type encountered") + + // ErrCompressedSizeTooBig is returned when a block is bigger than allowed. + // Typically this indicates wrong or corrupted input. + ErrCompressedSizeTooBig = errors.New("invalid input: compressed size too big") + + // ErrBlockTooSmall is returned when a block is too small to be decoded. + // Typically returned on invalid input. + ErrBlockTooSmall = errors.New("block too small") + + // ErrUnexpectedBlockSize is returned when a block has unexpected size. + // Typically returned on invalid input. + ErrUnexpectedBlockSize = errors.New("unexpected block size") + + // ErrMagicMismatch is returned when a "magic" number isn't what is expected. + // Typically this indicates wrong or corrupted input. + ErrMagicMismatch = errors.New("invalid input: magic number mismatch") + + // ErrWindowSizeExceeded is returned when a reference exceeds the valid window size. + // Typically this indicates wrong or corrupted input. + ErrWindowSizeExceeded = errors.New("window size exceeded") + + // ErrWindowSizeTooSmall is returned when no window size is specified. + // Typically this indicates wrong or corrupted input. + ErrWindowSizeTooSmall = errors.New("invalid input: window size was too small") + + // ErrDecoderSizeExceeded is returned if decompressed size exceeds the configured limit. + ErrDecoderSizeExceeded = errors.New("decompressed size exceeds configured limit") + + // ErrUnknownDictionary is returned if the dictionary ID is unknown. + // For the time being dictionaries are not supported. + ErrUnknownDictionary = errors.New("unknown dictionary") + + // ErrFrameSizeExceeded is returned if the stated frame size is exceeded. + // This is only returned if SingleSegment is specified on the frame. + ErrFrameSizeExceeded = errors.New("frame size exceeded") + + // ErrFrameSizeMismatch is returned if the stated frame size does not match the expected size. + // This is only returned if SingleSegment is specified on the frame. + ErrFrameSizeMismatch = errors.New("frame size does not match size on stream") + + // ErrCRCMismatch is returned if CRC mismatches. + ErrCRCMismatch = errors.New("CRC check failed") + + // ErrDecoderClosed will be returned if the Decoder was used after + // Close has been called. + ErrDecoderClosed = errors.New("decoder used after Close") + + // ErrDecoderNilInput is returned when a nil Reader was provided + // and an operation other than Reset/DecodeAll/Close was attempted. + ErrDecoderNilInput = errors.New("nil input provided as reader") +) + +func println(a ...interface{}) { + if debug || debugDecoder || debugEncoder { + log.Println(a...) + } +} + +func printf(format string, a ...interface{}) { + if debug || debugDecoder || debugEncoder { + log.Printf(format, a...) + } +} + +// matchLenFast does matching, but will not match the last up to 7 bytes. +func matchLenFast(a, b []byte) int { + endI := len(a) & (math.MaxInt32 - 7) + for i := 0; i < endI; i += 8 { + if diff := load64(a, i) ^ load64(b, i); diff != 0 { + return i + bits.TrailingZeros64(diff)>>3 + } + } + return endI +} + +// matchLen returns the maximum length. +// a must be the shortest of the two. +// The function also returns whether all bytes matched. +func matchLen(a, b []byte) int { + b = b[:len(a)] + for i := 0; i < len(a)-7; i += 8 { + if diff := load64(a, i) ^ load64(b, i); diff != 0 { + return i + (bits.TrailingZeros64(diff) >> 3) + } + } + + checked := (len(a) >> 3) << 3 + a = a[checked:] + b = b[checked:] + for i := range a { + if a[i] != b[i] { + return i + checked + } + } + return len(a) + checked +} + +func load3232(b []byte, i int32) uint32 { + return binary.LittleEndian.Uint32(b[i:]) +} + +func load6432(b []byte, i int32) uint64 { + return binary.LittleEndian.Uint64(b[i:]) +} + +func load64(b []byte, i int) uint64 { + return binary.LittleEndian.Uint64(b[i:]) +} + +type byter interface { + Bytes() []byte + Len() int +} + +var _ byter = &bytes.Buffer{} diff --git a/vendor/github.com/pkg/errors/.gitignore b/vendor/github.com/pkg/errors/.gitignore new file mode 100644 index 0000000..daf913b --- /dev/null +++ b/vendor/github.com/pkg/errors/.gitignore @@ -0,0 +1,24 @@ +# Compiled Object files, Static and Dynamic libs (Shared Objects) +*.o +*.a +*.so + +# Folders +_obj +_test + +# Architecture specific extensions/prefixes +*.[568vq] +[568vq].out + +*.cgo1.go +*.cgo2.c +_cgo_defun.c +_cgo_gotypes.go +_cgo_export.* + +_testmain.go + +*.exe +*.test +*.prof diff --git a/vendor/github.com/pkg/errors/.travis.yml b/vendor/github.com/pkg/errors/.travis.yml new file mode 100644 index 0000000..9159de0 --- /dev/null +++ b/vendor/github.com/pkg/errors/.travis.yml @@ -0,0 +1,10 @@ +language: go +go_import_path: github.com/pkg/errors +go: + - 1.11.x + - 1.12.x + - 1.13.x + - tip + +script: + - make check diff --git a/vendor/github.com/pkg/errors/LICENSE b/vendor/github.com/pkg/errors/LICENSE new file mode 100644 index 0000000..835ba3e --- /dev/null +++ b/vendor/github.com/pkg/errors/LICENSE @@ -0,0 +1,23 @@ +Copyright (c) 2015, Dave Cheney +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + +* Redistributions of source code must retain the above copyright notice, this + list of conditions and the following disclaimer. + +* Redistributions in binary form must reproduce the above copyright notice, + this list of conditions and the following disclaimer in the documentation + and/or other materials provided with the distribution. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE +DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE +FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR +SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER +CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, +OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/github.com/pkg/errors/Makefile b/vendor/github.com/pkg/errors/Makefile new file mode 100644 index 0000000..ce9d7cd --- /dev/null +++ b/vendor/github.com/pkg/errors/Makefile @@ -0,0 +1,44 @@ +PKGS := github.com/pkg/errors +SRCDIRS := $(shell go list -f '{{.Dir}}' $(PKGS)) +GO := go + +check: test vet gofmt misspell unconvert staticcheck ineffassign unparam + +test: + $(GO) test $(PKGS) + +vet: | test + $(GO) vet $(PKGS) + +staticcheck: + $(GO) get honnef.co/go/tools/cmd/staticcheck + staticcheck -checks all $(PKGS) + +misspell: + $(GO) get github.com/client9/misspell/cmd/misspell + misspell \ + -locale GB \ + -error \ + *.md *.go + +unconvert: + $(GO) get github.com/mdempsky/unconvert + unconvert -v $(PKGS) + +ineffassign: + $(GO) get github.com/gordonklaus/ineffassign + find $(SRCDIRS) -name '*.go' | xargs ineffassign + +pedantic: check errcheck + +unparam: + $(GO) get mvdan.cc/unparam + unparam ./... + +errcheck: + $(GO) get github.com/kisielk/errcheck + errcheck $(PKGS) + +gofmt: + @echo Checking code is gofmted + @test -z "$(shell gofmt -s -l -d -e $(SRCDIRS) | tee /dev/stderr)" diff --git a/vendor/github.com/pkg/errors/README.md b/vendor/github.com/pkg/errors/README.md new file mode 100644 index 0000000..54dfdcb --- /dev/null +++ b/vendor/github.com/pkg/errors/README.md @@ -0,0 +1,59 @@ +# errors [![Travis-CI](https://travis-ci.org/pkg/errors.svg)](https://travis-ci.org/pkg/errors) [![AppVeyor](https://ci.appveyor.com/api/projects/status/b98mptawhudj53ep/branch/master?svg=true)](https://ci.appveyor.com/project/davecheney/errors/branch/master) [![GoDoc](https://godoc.org/github.com/pkg/errors?status.svg)](http://godoc.org/github.com/pkg/errors) [![Report card](https://goreportcard.com/badge/github.com/pkg/errors)](https://goreportcard.com/report/github.com/pkg/errors) [![Sourcegraph](https://sourcegraph.com/github.com/pkg/errors/-/badge.svg)](https://sourcegraph.com/github.com/pkg/errors?badge) + +Package errors provides simple error handling primitives. + +`go get github.com/pkg/errors` + +The traditional error handling idiom in Go is roughly akin to +```go +if err != nil { + return err +} +``` +which applied recursively up the call stack results in error reports without context or debugging information. The errors package allows programmers to add context to the failure path in their code in a way that does not destroy the original value of the error. + +## Adding context to an error + +The errors.Wrap function returns a new error that adds context to the original error. For example +```go +_, err := ioutil.ReadAll(r) +if err != nil { + return errors.Wrap(err, "read failed") +} +``` +## Retrieving the cause of an error + +Using `errors.Wrap` constructs a stack of errors, adding context to the preceding error. Depending on the nature of the error it may be necessary to reverse the operation of errors.Wrap to retrieve the original error for inspection. Any error value which implements this interface can be inspected by `errors.Cause`. +```go +type causer interface { + Cause() error +} +``` +`errors.Cause` will recursively retrieve the topmost error which does not implement `causer`, which is assumed to be the original cause. For example: +```go +switch err := errors.Cause(err).(type) { +case *MyError: + // handle specifically +default: + // unknown error +} +``` + +[Read the package documentation for more information](https://godoc.org/github.com/pkg/errors). + +## Roadmap + +With the upcoming [Go2 error proposals](https://go.googlesource.com/proposal/+/master/design/go2draft.md) this package is moving into maintenance mode. The roadmap for a 1.0 release is as follows: + +- 0.9. Remove pre Go 1.9 and Go 1.10 support, address outstanding pull requests (if possible) +- 1.0. Final release. + +## Contributing + +Because of the Go2 errors changes, this package is not accepting proposals for new functionality. With that said, we welcome pull requests, bug fixes and issue reports. + +Before sending a PR, please discuss your change by raising an issue. + +## License + +BSD-2-Clause diff --git a/vendor/github.com/pkg/errors/appveyor.yml b/vendor/github.com/pkg/errors/appveyor.yml new file mode 100644 index 0000000..a932ead --- /dev/null +++ b/vendor/github.com/pkg/errors/appveyor.yml @@ -0,0 +1,32 @@ +version: build-{build}.{branch} + +clone_folder: C:\gopath\src\github.com\pkg\errors +shallow_clone: true # for startup speed + +environment: + GOPATH: C:\gopath + +platform: + - x64 + +# http://www.appveyor.com/docs/installed-software +install: + # some helpful output for debugging builds + - go version + - go env + # pre-installed MinGW at C:\MinGW is 32bit only + # but MSYS2 at C:\msys64 has mingw64 + - set PATH=C:\msys64\mingw64\bin;%PATH% + - gcc --version + - g++ --version + +build_script: + - go install -v ./... + +test_script: + - set PATH=C:\gopath\bin;%PATH% + - go test -v ./... + +#artifacts: +# - path: '%GOPATH%\bin\*.exe' +deploy: off diff --git a/vendor/github.com/pkg/errors/errors.go b/vendor/github.com/pkg/errors/errors.go new file mode 100644 index 0000000..161aea2 --- /dev/null +++ b/vendor/github.com/pkg/errors/errors.go @@ -0,0 +1,288 @@ +// Package errors provides simple error handling primitives. +// +// The traditional error handling idiom in Go is roughly akin to +// +// if err != nil { +// return err +// } +// +// which when applied recursively up the call stack results in error reports +// without context or debugging information. The errors package allows +// programmers to add context to the failure path in their code in a way +// that does not destroy the original value of the error. +// +// Adding context to an error +// +// The errors.Wrap function returns a new error that adds context to the +// original error by recording a stack trace at the point Wrap is called, +// together with the supplied message. For example +// +// _, err := ioutil.ReadAll(r) +// if err != nil { +// return errors.Wrap(err, "read failed") +// } +// +// If additional control is required, the errors.WithStack and +// errors.WithMessage functions destructure errors.Wrap into its component +// operations: annotating an error with a stack trace and with a message, +// respectively. +// +// Retrieving the cause of an error +// +// Using errors.Wrap constructs a stack of errors, adding context to the +// preceding error. Depending on the nature of the error it may be necessary +// to reverse the operation of errors.Wrap to retrieve the original error +// for inspection. Any error value which implements this interface +// +// type causer interface { +// Cause() error +// } +// +// can be inspected by errors.Cause. errors.Cause will recursively retrieve +// the topmost error that does not implement causer, which is assumed to be +// the original cause. For example: +// +// switch err := errors.Cause(err).(type) { +// case *MyError: +// // handle specifically +// default: +// // unknown error +// } +// +// Although the causer interface is not exported by this package, it is +// considered a part of its stable public interface. +// +// Formatted printing of errors +// +// All error values returned from this package implement fmt.Formatter and can +// be formatted by the fmt package. The following verbs are supported: +// +// %s print the error. If the error has a Cause it will be +// printed recursively. +// %v see %s +// %+v extended format. Each Frame of the error's StackTrace will +// be printed in detail. +// +// Retrieving the stack trace of an error or wrapper +// +// New, Errorf, Wrap, and Wrapf record a stack trace at the point they are +// invoked. This information can be retrieved with the following interface: +// +// type stackTracer interface { +// StackTrace() errors.StackTrace +// } +// +// The returned errors.StackTrace type is defined as +// +// type StackTrace []Frame +// +// The Frame type represents a call site in the stack trace. Frame supports +// the fmt.Formatter interface that can be used for printing information about +// the stack trace of this error. For example: +// +// if err, ok := err.(stackTracer); ok { +// for _, f := range err.StackTrace() { +// fmt.Printf("%+s:%d\n", f, f) +// } +// } +// +// Although the stackTracer interface is not exported by this package, it is +// considered a part of its stable public interface. +// +// See the documentation for Frame.Format for more details. +package errors + +import ( + "fmt" + "io" +) + +// New returns an error with the supplied message. +// New also records the stack trace at the point it was called. +func New(message string) error { + return &fundamental{ + msg: message, + stack: callers(), + } +} + +// Errorf formats according to a format specifier and returns the string +// as a value that satisfies error. +// Errorf also records the stack trace at the point it was called. +func Errorf(format string, args ...interface{}) error { + return &fundamental{ + msg: fmt.Sprintf(format, args...), + stack: callers(), + } +} + +// fundamental is an error that has a message and a stack, but no caller. +type fundamental struct { + msg string + *stack +} + +func (f *fundamental) Error() string { return f.msg } + +func (f *fundamental) Format(s fmt.State, verb rune) { + switch verb { + case 'v': + if s.Flag('+') { + io.WriteString(s, f.msg) + f.stack.Format(s, verb) + return + } + fallthrough + case 's': + io.WriteString(s, f.msg) + case 'q': + fmt.Fprintf(s, "%q", f.msg) + } +} + +// WithStack annotates err with a stack trace at the point WithStack was called. +// If err is nil, WithStack returns nil. +func WithStack(err error) error { + if err == nil { + return nil + } + return &withStack{ + err, + callers(), + } +} + +type withStack struct { + error + *stack +} + +func (w *withStack) Cause() error { return w.error } + +// Unwrap provides compatibility for Go 1.13 error chains. +func (w *withStack) Unwrap() error { return w.error } + +func (w *withStack) Format(s fmt.State, verb rune) { + switch verb { + case 'v': + if s.Flag('+') { + fmt.Fprintf(s, "%+v", w.Cause()) + w.stack.Format(s, verb) + return + } + fallthrough + case 's': + io.WriteString(s, w.Error()) + case 'q': + fmt.Fprintf(s, "%q", w.Error()) + } +} + +// Wrap returns an error annotating err with a stack trace +// at the point Wrap is called, and the supplied message. +// If err is nil, Wrap returns nil. +func Wrap(err error, message string) error { + if err == nil { + return nil + } + err = &withMessage{ + cause: err, + msg: message, + } + return &withStack{ + err, + callers(), + } +} + +// Wrapf returns an error annotating err with a stack trace +// at the point Wrapf is called, and the format specifier. +// If err is nil, Wrapf returns nil. +func Wrapf(err error, format string, args ...interface{}) error { + if err == nil { + return nil + } + err = &withMessage{ + cause: err, + msg: fmt.Sprintf(format, args...), + } + return &withStack{ + err, + callers(), + } +} + +// WithMessage annotates err with a new message. +// If err is nil, WithMessage returns nil. +func WithMessage(err error, message string) error { + if err == nil { + return nil + } + return &withMessage{ + cause: err, + msg: message, + } +} + +// WithMessagef annotates err with the format specifier. +// If err is nil, WithMessagef returns nil. +func WithMessagef(err error, format string, args ...interface{}) error { + if err == nil { + return nil + } + return &withMessage{ + cause: err, + msg: fmt.Sprintf(format, args...), + } +} + +type withMessage struct { + cause error + msg string +} + +func (w *withMessage) Error() string { return w.msg + ": " + w.cause.Error() } +func (w *withMessage) Cause() error { return w.cause } + +// Unwrap provides compatibility for Go 1.13 error chains. +func (w *withMessage) Unwrap() error { return w.cause } + +func (w *withMessage) Format(s fmt.State, verb rune) { + switch verb { + case 'v': + if s.Flag('+') { + fmt.Fprintf(s, "%+v\n", w.Cause()) + io.WriteString(s, w.msg) + return + } + fallthrough + case 's', 'q': + io.WriteString(s, w.Error()) + } +} + +// Cause returns the underlying cause of the error, if possible. +// An error value has a cause if it implements the following +// interface: +// +// type causer interface { +// Cause() error +// } +// +// If the error does not implement Cause, the original error will +// be returned. If the error is nil, nil will be returned without further +// investigation. +func Cause(err error) error { + type causer interface { + Cause() error + } + + for err != nil { + cause, ok := err.(causer) + if !ok { + break + } + err = cause.Cause() + } + return err +} diff --git a/vendor/github.com/pkg/errors/go113.go b/vendor/github.com/pkg/errors/go113.go new file mode 100644 index 0000000..be0d10d --- /dev/null +++ b/vendor/github.com/pkg/errors/go113.go @@ -0,0 +1,38 @@ +// +build go1.13 + +package errors + +import ( + stderrors "errors" +) + +// Is reports whether any error in err's chain matches target. +// +// The chain consists of err itself followed by the sequence of errors obtained by +// repeatedly calling Unwrap. +// +// An error is considered to match a target if it is equal to that target or if +// it implements a method Is(error) bool such that Is(target) returns true. +func Is(err, target error) bool { return stderrors.Is(err, target) } + +// As finds the first error in err's chain that matches target, and if so, sets +// target to that error value and returns true. +// +// The chain consists of err itself followed by the sequence of errors obtained by +// repeatedly calling Unwrap. +// +// An error matches target if the error's concrete value is assignable to the value +// pointed to by target, or if the error has a method As(interface{}) bool such that +// As(target) returns true. In the latter case, the As method is responsible for +// setting target. +// +// As will panic if target is not a non-nil pointer to either a type that implements +// error, or to any interface type. As returns false if err is nil. +func As(err error, target interface{}) bool { return stderrors.As(err, target) } + +// Unwrap returns the result of calling the Unwrap method on err, if err's +// type contains an Unwrap method returning error. +// Otherwise, Unwrap returns nil. +func Unwrap(err error) error { + return stderrors.Unwrap(err) +} diff --git a/vendor/github.com/pkg/errors/stack.go b/vendor/github.com/pkg/errors/stack.go new file mode 100644 index 0000000..779a834 --- /dev/null +++ b/vendor/github.com/pkg/errors/stack.go @@ -0,0 +1,177 @@ +package errors + +import ( + "fmt" + "io" + "path" + "runtime" + "strconv" + "strings" +) + +// Frame represents a program counter inside a stack frame. +// For historical reasons if Frame is interpreted as a uintptr +// its value represents the program counter + 1. +type Frame uintptr + +// pc returns the program counter for this frame; +// multiple frames may have the same PC value. +func (f Frame) pc() uintptr { return uintptr(f) - 1 } + +// file returns the full path to the file that contains the +// function for this Frame's pc. +func (f Frame) file() string { + fn := runtime.FuncForPC(f.pc()) + if fn == nil { + return "unknown" + } + file, _ := fn.FileLine(f.pc()) + return file +} + +// line returns the line number of source code of the +// function for this Frame's pc. +func (f Frame) line() int { + fn := runtime.FuncForPC(f.pc()) + if fn == nil { + return 0 + } + _, line := fn.FileLine(f.pc()) + return line +} + +// name returns the name of this function, if known. +func (f Frame) name() string { + fn := runtime.FuncForPC(f.pc()) + if fn == nil { + return "unknown" + } + return fn.Name() +} + +// Format formats the frame according to the fmt.Formatter interface. +// +// %s source file +// %d source line +// %n function name +// %v equivalent to %s:%d +// +// Format accepts flags that alter the printing of some verbs, as follows: +// +// %+s function name and path of source file relative to the compile time +// GOPATH separated by \n\t (\n\t) +// %+v equivalent to %+s:%d +func (f Frame) Format(s fmt.State, verb rune) { + switch verb { + case 's': + switch { + case s.Flag('+'): + io.WriteString(s, f.name()) + io.WriteString(s, "\n\t") + io.WriteString(s, f.file()) + default: + io.WriteString(s, path.Base(f.file())) + } + case 'd': + io.WriteString(s, strconv.Itoa(f.line())) + case 'n': + io.WriteString(s, funcname(f.name())) + case 'v': + f.Format(s, 's') + io.WriteString(s, ":") + f.Format(s, 'd') + } +} + +// MarshalText formats a stacktrace Frame as a text string. The output is the +// same as that of fmt.Sprintf("%+v", f), but without newlines or tabs. +func (f Frame) MarshalText() ([]byte, error) { + name := f.name() + if name == "unknown" { + return []byte(name), nil + } + return []byte(fmt.Sprintf("%s %s:%d", name, f.file(), f.line())), nil +} + +// StackTrace is stack of Frames from innermost (newest) to outermost (oldest). +type StackTrace []Frame + +// Format formats the stack of Frames according to the fmt.Formatter interface. +// +// %s lists source files for each Frame in the stack +// %v lists the source file and line number for each Frame in the stack +// +// Format accepts flags that alter the printing of some verbs, as follows: +// +// %+v Prints filename, function, and line number for each Frame in the stack. +func (st StackTrace) Format(s fmt.State, verb rune) { + switch verb { + case 'v': + switch { + case s.Flag('+'): + for _, f := range st { + io.WriteString(s, "\n") + f.Format(s, verb) + } + case s.Flag('#'): + fmt.Fprintf(s, "%#v", []Frame(st)) + default: + st.formatSlice(s, verb) + } + case 's': + st.formatSlice(s, verb) + } +} + +// formatSlice will format this StackTrace into the given buffer as a slice of +// Frame, only valid when called with '%s' or '%v'. +func (st StackTrace) formatSlice(s fmt.State, verb rune) { + io.WriteString(s, "[") + for i, f := range st { + if i > 0 { + io.WriteString(s, " ") + } + f.Format(s, verb) + } + io.WriteString(s, "]") +} + +// stack represents a stack of program counters. +type stack []uintptr + +func (s *stack) Format(st fmt.State, verb rune) { + switch verb { + case 'v': + switch { + case st.Flag('+'): + for _, pc := range *s { + f := Frame(pc) + fmt.Fprintf(st, "\n%+v", f) + } + } + } +} + +func (s *stack) StackTrace() StackTrace { + f := make([]Frame, len(*s)) + for i := 0; i < len(f); i++ { + f[i] = Frame((*s)[i]) + } + return f +} + +func callers() *stack { + const depth = 32 + var pcs [depth]uintptr + n := runtime.Callers(3, pcs[:]) + var st stack = pcs[0:n] + return &st +} + +// funcname removes the path prefix component of a function's name reported by func.Name(). +func funcname(name string) string { + i := strings.LastIndex(name, "/") + name = name[i+1:] + i = strings.Index(name, ".") + return name[i+1:] +} diff --git a/vendor/github.com/rs/zerolog/.gitignore b/vendor/github.com/rs/zerolog/.gitignore new file mode 100644 index 0000000..8ebe58b --- /dev/null +++ b/vendor/github.com/rs/zerolog/.gitignore @@ -0,0 +1,25 @@ +# Compiled Object files, Static and Dynamic libs (Shared Objects) +*.o +*.a +*.so + +# Folders +_obj +_test +tmp + +# Architecture specific extensions/prefixes +*.[568vq] +[568vq].out + +*.cgo1.go +*.cgo2.c +_cgo_defun.c +_cgo_gotypes.go +_cgo_export.* + +_testmain.go + +*.exe +*.test +*.prof diff --git a/vendor/github.com/rs/zerolog/CNAME b/vendor/github.com/rs/zerolog/CNAME new file mode 100644 index 0000000..9ce57a6 --- /dev/null +++ b/vendor/github.com/rs/zerolog/CNAME @@ -0,0 +1 @@ +zerolog.io \ No newline at end of file diff --git a/vendor/github.com/rs/zerolog/LICENSE b/vendor/github.com/rs/zerolog/LICENSE new file mode 100644 index 0000000..677e07f --- /dev/null +++ b/vendor/github.com/rs/zerolog/LICENSE @@ -0,0 +1,21 @@ +MIT License + +Copyright (c) 2017 Olivier Poitrey + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. diff --git a/vendor/github.com/rs/zerolog/README.md b/vendor/github.com/rs/zerolog/README.md new file mode 100644 index 0000000..47b1297 --- /dev/null +++ b/vendor/github.com/rs/zerolog/README.md @@ -0,0 +1,695 @@ +# Zero Allocation JSON Logger + +[![godoc](http://img.shields.io/badge/godoc-reference-blue.svg?style=flat)](https://godoc.org/github.com/rs/zerolog) [![license](http://img.shields.io/badge/license-MIT-red.svg?style=flat)](https://raw.githubusercontent.com/rs/zerolog/master/LICENSE) [![Build Status](https://travis-ci.org/rs/zerolog.svg?branch=master)](https://travis-ci.org/rs/zerolog) [![Coverage](http://gocover.io/_badge/github.com/rs/zerolog)](http://gocover.io/github.com/rs/zerolog) + +The zerolog package provides a fast and simple logger dedicated to JSON output. + +Zerolog's API is designed to provide both a great developer experience and stunning [performance](#benchmarks). Its unique chaining API allows zerolog to write JSON (or CBOR) log events by avoiding allocations and reflection. + +Uber's [zap](https://godoc.org/go.uber.org/zap) library pioneered this approach. Zerolog is taking this concept to the next level with a simpler to use API and even better performance. + +To keep the code base and the API simple, zerolog focuses on efficient structured logging only. Pretty logging on the console is made possible using the provided (but inefficient) [`zerolog.ConsoleWriter`](#pretty-logging). + +![Pretty Logging Image](pretty.png) + +## Who uses zerolog + +Find out [who uses zerolog](https://github.com/rs/zerolog/wiki/Who-uses-zerolog) and add your company / project to the list. + +## Features + +* [Blazing fast](#benchmarks) +* [Low to zero allocation](#benchmarks) +* [Leveled logging](#leveled-logging) +* [Sampling](#log-sampling) +* [Hooks](#hooks) +* [Contextual fields](#contextual-logging) +* `context.Context` integration +* [Integration with `net/http`](#integration-with-nethttp) +* [JSON and CBOR encoding formats](#binary-encoding) +* [Pretty logging for development](#pretty-logging) +* [Error Logging (with optional Stacktrace)](#error-logging) + +## Installation + +```bash +go get -u github.com/rs/zerolog/log +``` + +## Getting Started + +### Simple Logging Example + +For simple logging, import the global logger package **github.com/rs/zerolog/log** + +```go +package main + +import ( + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + // UNIX Time is faster and smaller than most timestamps + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + + log.Print("hello world") +} + +// Output: {"time":1516134303,"level":"debug","message":"hello world"} +``` +> Note: By default log writes to `os.Stderr` +> Note: The default log level for `log.Print` is *debug* + +### Contextual Logging + +**zerolog** allows data to be added to log messages in the form of key:value pairs. The data added to the message adds "context" about the log event that can be critical for debugging as well as myriad other purposes. An example of this is below: + +```go +package main + +import ( + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + + log.Debug(). + Str("Scale", "833 cents"). + Float64("Interval", 833.09). + Msg("Fibonacci is everywhere") + + log.Debug(). + Str("Name", "Tom"). + Send() +} + +// Output: {"level":"debug","Scale":"833 cents","Interval":833.09,"time":1562212768,"message":"Fibonacci is everywhere"} +// Output: {"level":"debug","Name":"Tom","time":1562212768} +``` + +> You'll note in the above example that when adding contextual fields, the fields are strongly typed. You can find the full list of supported fields [here](#standard-types) + +### Leveled Logging + +#### Simple Leveled Logging Example + +```go +package main + +import ( + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + + log.Info().Msg("hello world") +} + +// Output: {"time":1516134303,"level":"info","message":"hello world"} +``` + +> It is very important to note that when using the **zerolog** chaining API, as shown above (`log.Info().Msg("hello world"`), the chain must have either the `Msg` or `Msgf` method call. If you forget to add either of these, the log will not occur and there is no compile time error to alert you of this. + +**zerolog** allows for logging at the following levels (from highest to lowest): + +* panic (`zerolog.PanicLevel`, 5) +* fatal (`zerolog.FatalLevel`, 4) +* error (`zerolog.ErrorLevel`, 3) +* warn (`zerolog.WarnLevel`, 2) +* info (`zerolog.InfoLevel`, 1) +* debug (`zerolog.DebugLevel`, 0) +* trace (`zerolog.TraceLevel`, -1) + +You can set the Global logging level to any of these options using the `SetGlobalLevel` function in the zerolog package, passing in one of the given constants above, e.g. `zerolog.InfoLevel` would be the "info" level. Whichever level is chosen, all logs with a level greater than or equal to that level will be written. To turn off logging entirely, pass the `zerolog.Disabled` constant. + +#### Setting Global Log Level + +This example uses command-line flags to demonstrate various outputs depending on the chosen log level. + +```go +package main + +import ( + "flag" + + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + debug := flag.Bool("debug", false, "sets log level to debug") + + flag.Parse() + + // Default level for this example is info, unless debug flag is present + zerolog.SetGlobalLevel(zerolog.InfoLevel) + if *debug { + zerolog.SetGlobalLevel(zerolog.DebugLevel) + } + + log.Debug().Msg("This message appears only when log level set to Debug") + log.Info().Msg("This message appears when log level set to Debug or Info") + + if e := log.Debug(); e.Enabled() { + // Compute log output only if enabled. + value := "bar" + e.Str("foo", value).Msg("some debug message") + } +} +``` + +Info Output (no flag) + +```bash +$ ./logLevelExample +{"time":1516387492,"level":"info","message":"This message appears when log level set to Debug or Info"} +``` + +Debug Output (debug flag set) + +```bash +$ ./logLevelExample -debug +{"time":1516387573,"level":"debug","message":"This message appears only when log level set to Debug"} +{"time":1516387573,"level":"info","message":"This message appears when log level set to Debug or Info"} +{"time":1516387573,"level":"debug","foo":"bar","message":"some debug message"} +``` + +#### Logging without Level or Message + +You may choose to log without a specific level by using the `Log` method. You may also write without a message by setting an empty string in the `msg string` parameter of the `Msg` method. Both are demonstrated in the example below. + +```go +package main + +import ( + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + + log.Log(). + Str("foo", "bar"). + Msg("") +} + +// Output: {"time":1494567715,"foo":"bar"} +``` + +### Error Logging + +You can log errors using the `Err` method + +```go +package main + +import ( + "errors" + + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + + err := errors.New("seems we have an error here") + log.Error().Err(err).Msg("") +} + +// Output: {"level":"error","error":"seems we have an error here","time":1609085256} +``` + +> The default field name for errors is `error`, you can change this by setting `zerolog.ErrorFieldName` to meet your needs. + +#### Error Logging with Stacktrace + +Using `github.com/pkg/errors`, you can add a formatted stacktrace to your errors. + +```go +package main + +import ( + "github.com/pkg/errors" + "github.com/rs/zerolog/pkgerrors" + + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + zerolog.ErrorStackMarshaler = pkgerrors.MarshalStack + + err := outer() + log.Error().Stack().Err(err).Msg("") +} + +func inner() error { + return errors.New("seems we have an error here") +} + +func middle() error { + err := inner() + if err != nil { + return err + } + return nil +} + +func outer() error { + err := middle() + if err != nil { + return err + } + return nil +} + +// Output: {"level":"error","stack":[{"func":"inner","line":"20","source":"errors.go"},{"func":"middle","line":"24","source":"errors.go"},{"func":"outer","line":"32","source":"errors.go"},{"func":"main","line":"15","source":"errors.go"},{"func":"main","line":"204","source":"proc.go"},{"func":"goexit","line":"1374","source":"asm_amd64.s"}],"error":"seems we have an error here","time":1609086683} +``` + +> zerolog.ErrorStackMarshaler must be set in order for the stack to output anything. + +#### Logging Fatal Messages + +```go +package main + +import ( + "errors" + + "github.com/rs/zerolog" + "github.com/rs/zerolog/log" +) + +func main() { + err := errors.New("A repo man spends his life getting into tense situations") + service := "myservice" + + zerolog.TimeFieldFormat = zerolog.TimeFormatUnix + + log.Fatal(). + Err(err). + Str("service", service). + Msgf("Cannot start %s", service) +} + +// Output: {"time":1516133263,"level":"fatal","error":"A repo man spends his life getting into tense situations","service":"myservice","message":"Cannot start myservice"} +// exit status 1 +``` + +> NOTE: Using `Msgf` generates one allocation even when the logger is disabled. + + +### Create logger instance to manage different outputs + +```go +logger := zerolog.New(os.Stderr).With().Timestamp().Logger() + +logger.Info().Str("foo", "bar").Msg("hello world") + +// Output: {"level":"info","time":1494567715,"message":"hello world","foo":"bar"} +``` + +### Sub-loggers let you chain loggers with additional context + +```go +sublogger := log.With(). + Str("component", "foo"). + Logger() +sublogger.Info().Msg("hello world") + +// Output: {"level":"info","time":1494567715,"message":"hello world","component":"foo"} +``` + +### Pretty logging + +To log a human-friendly, colorized output, use `zerolog.ConsoleWriter`: + +```go +log.Logger = log.Output(zerolog.ConsoleWriter{Out: os.Stderr}) + +log.Info().Str("foo", "bar").Msg("Hello world") + +// Output: 3:04PM INF Hello World foo=bar +``` + +To customize the configuration and formatting: + +```go +output := zerolog.ConsoleWriter{Out: os.Stdout, TimeFormat: time.RFC3339} +output.FormatLevel = func(i interface{}) string { + return strings.ToUpper(fmt.Sprintf("| %-6s|", i)) +} +output.FormatMessage = func(i interface{}) string { + return fmt.Sprintf("***%s****", i) +} +output.FormatFieldName = func(i interface{}) string { + return fmt.Sprintf("%s:", i) +} +output.FormatFieldValue = func(i interface{}) string { + return strings.ToUpper(fmt.Sprintf("%s", i)) +} + +log := zerolog.New(output).With().Timestamp().Logger() + +log.Info().Str("foo", "bar").Msg("Hello World") + +// Output: 2006-01-02T15:04:05Z07:00 | INFO | ***Hello World**** foo:BAR +``` + +### Sub dictionary + +```go +log.Info(). + Str("foo", "bar"). + Dict("dict", zerolog.Dict(). + Str("bar", "baz"). + Int("n", 1), + ).Msg("hello world") + +// Output: {"level":"info","time":1494567715,"foo":"bar","dict":{"bar":"baz","n":1},"message":"hello world"} +``` + +### Customize automatic field names + +```go +zerolog.TimestampFieldName = "t" +zerolog.LevelFieldName = "l" +zerolog.MessageFieldName = "m" + +log.Info().Msg("hello world") + +// Output: {"l":"info","t":1494567715,"m":"hello world"} +``` + +### Add contextual fields to the global logger + +```go +log.Logger = log.With().Str("foo", "bar").Logger() +``` + +### Add file and line number to log + +```go +log.Logger = log.With().Caller().Logger() +log.Info().Msg("hello world") + +// Output: {"level": "info", "message": "hello world", "caller": "/go/src/your_project/some_file:21"} +``` + + +### Thread-safe, lock-free, non-blocking writer + +If your writer might be slow or not thread-safe and you need your log producers to never get slowed down by a slow writer, you can use a `diode.Writer` as follow: + +```go +wr := diode.NewWriter(os.Stdout, 1000, 10*time.Millisecond, func(missed int) { + fmt.Printf("Logger Dropped %d messages", missed) + }) +log := zerolog.New(wr) +log.Print("test") +``` + +You will need to install `code.cloudfoundry.org/go-diodes` to use this feature. + +### Log Sampling + +```go +sampled := log.Sample(&zerolog.BasicSampler{N: 10}) +sampled.Info().Msg("will be logged every 10 messages") + +// Output: {"time":1494567715,"level":"info","message":"will be logged every 10 messages"} +``` + +More advanced sampling: + +```go +// Will let 5 debug messages per period of 1 second. +// Over 5 debug message, 1 every 100 debug messages are logged. +// Other levels are not sampled. +sampled := log.Sample(zerolog.LevelSampler{ + DebugSampler: &zerolog.BurstSampler{ + Burst: 5, + Period: 1*time.Second, + NextSampler: &zerolog.BasicSampler{N: 100}, + }, +}) +sampled.Debug().Msg("hello world") + +// Output: {"time":1494567715,"level":"debug","message":"hello world"} +``` + +### Hooks + +```go +type SeverityHook struct{} + +func (h SeverityHook) Run(e *zerolog.Event, level zerolog.Level, msg string) { + if level != zerolog.NoLevel { + e.Str("severity", level.String()) + } +} + +hooked := log.Hook(SeverityHook{}) +hooked.Warn().Msg("") + +// Output: {"level":"warn","severity":"warn"} +``` + +### Pass a sub-logger by context + +```go +ctx := log.With().Str("component", "module").Logger().WithContext(ctx) + +log.Ctx(ctx).Info().Msg("hello world") + +// Output: {"component":"module","level":"info","message":"hello world"} +``` + +### Set as standard logger output + +```go +log := zerolog.New(os.Stdout).With(). + Str("foo", "bar"). + Logger() + +stdlog.SetFlags(0) +stdlog.SetOutput(log) + +stdlog.Print("hello world") + +// Output: {"foo":"bar","message":"hello world"} +``` + +### Integration with `net/http` + +The `github.com/rs/zerolog/hlog` package provides some helpers to integrate zerolog with `http.Handler`. + +In this example we use [alice](https://github.com/justinas/alice) to install logger for better readability. + +```go +log := zerolog.New(os.Stdout).With(). + Timestamp(). + Str("role", "my-service"). + Str("host", host). + Logger() + +c := alice.New() + +// Install the logger handler with default output on the console +c = c.Append(hlog.NewHandler(log)) + +// Install some provided extra handler to set some request's context fields. +// Thanks to that handler, all our logs will come with some prepopulated fields. +c = c.Append(hlog.AccessHandler(func(r *http.Request, status, size int, duration time.Duration) { + hlog.FromRequest(r).Info(). + Str("method", r.Method). + Stringer("url", r.URL). + Int("status", status). + Int("size", size). + Dur("duration", duration). + Msg("") +})) +c = c.Append(hlog.RemoteAddrHandler("ip")) +c = c.Append(hlog.UserAgentHandler("user_agent")) +c = c.Append(hlog.RefererHandler("referer")) +c = c.Append(hlog.RequestIDHandler("req_id", "Request-Id")) + +// Here is your final handler +h := c.Then(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { + // Get the logger from the request's context. You can safely assume it + // will be always there: if the handler is removed, hlog.FromRequest + // will return a no-op logger. + hlog.FromRequest(r).Info(). + Str("user", "current user"). + Str("status", "ok"). + Msg("Something happened") + + // Output: {"level":"info","time":"2001-02-03T04:05:06Z","role":"my-service","host":"local-hostname","req_id":"b4g0l5t6tfid6dtrapu0","user":"current user","status":"ok","message":"Something happened"} +})) +http.Handle("/", h) + +if err := http.ListenAndServe(":8080", nil); err != nil { + log.Fatal().Err(err).Msg("Startup failed") +} +``` + +## Multiple Log Output +`zerolog.MultiLevelWriter` may be used to send the log message to multiple outputs. +In this example, we send the log message to both `os.Stdout` and the in-built ConsoleWriter. +```go +func main() { + consoleWriter := zerolog.ConsoleWriter{Out: os.Stdout} + + multi := zerolog.MultiLevelWriter(consoleWriter, os.Stdout) + + logger := zerolog.New(multi).With().Timestamp().Logger() + + logger.Info().Msg("Hello World!") +} + +// Output (Line 1: Console; Line 2: Stdout) +// 12:36PM INF Hello World! +// {"level":"info","time":"2019-11-07T12:36:38+03:00","message":"Hello World!"} +``` + +## Global Settings + +Some settings can be changed and will by applied to all loggers: + +* `log.Logger`: You can set this value to customize the global logger (the one used by package level methods). +* `zerolog.SetGlobalLevel`: Can raise the minimum level of all loggers. Call this with `zerolog.Disabled` to disable logging altogether (quiet mode). +* `zerolog.DisableSampling`: If argument is `true`, all sampled loggers will stop sampling and issue 100% of their log events. +* `zerolog.TimestampFieldName`: Can be set to customize `Timestamp` field name. +* `zerolog.LevelFieldName`: Can be set to customize level field name. +* `zerolog.MessageFieldName`: Can be set to customize message field name. +* `zerolog.ErrorFieldName`: Can be set to customize `Err` field name. +* `zerolog.TimeFieldFormat`: Can be set to customize `Time` field value formatting. If set with `zerolog.TimeFormatUnix`, `zerolog.TimeFormatUnixMs` or `zerolog.TimeFormatUnixMicro`, times are formated as UNIX timestamp. +* `zerolog.DurationFieldUnit`: Can be set to customize the unit for time.Duration type fields added by `Dur` (default: `time.Millisecond`). +* `zerolog.DurationFieldInteger`: If set to `true`, `Dur` fields are formatted as integers instead of floats (default: `false`). +* `zerolog.ErrorHandler`: Called whenever zerolog fails to write an event on its output. If not set, an error is printed on the stderr. This handler must be thread safe and non-blocking. + +## Field Types + +### Standard Types + +* `Str` +* `Bool` +* `Int`, `Int8`, `Int16`, `Int32`, `Int64` +* `Uint`, `Uint8`, `Uint16`, `Uint32`, `Uint64` +* `Float32`, `Float64` + +### Advanced Fields + +* `Err`: Takes an `error` and renders it as a string using the `zerolog.ErrorFieldName` field name. +* `Func`: Run a `func` only if the level is enabled. +* `Timestamp`: Inserts a timestamp field with `zerolog.TimestampFieldName` field name, formatted using `zerolog.TimeFieldFormat`. +* `Time`: Adds a field with time formatted with `zerolog.TimeFieldFormat`. +* `Dur`: Adds a field with `time.Duration`. +* `Dict`: Adds a sub-key/value as a field of the event. +* `RawJSON`: Adds a field with an already encoded JSON (`[]byte`) +* `Hex`: Adds a field with value formatted as a hexadecimal string (`[]byte`) +* `Interface`: Uses reflection to marshal the type. + +Most fields are also available in the slice format (`Strs` for `[]string`, `Errs` for `[]error` etc.) + +## Binary Encoding + +In addition to the default JSON encoding, `zerolog` can produce binary logs using [CBOR](http://cbor.io) encoding. The choice of encoding can be decided at compile time using the build tag `binary_log` as follows: + +```bash +go build -tags binary_log . +``` + +To Decode binary encoded log files you can use any CBOR decoder. One has been tested to work +with zerolog library is [CSD](https://github.com/toravir/csd/). + +## Related Projects + +* [grpc-zerolog](https://github.com/cheapRoc/grpc-zerolog): Implementation of `grpclog.LoggerV2` interface using `zerolog` +* [overlog](https://github.com/Trendyol/overlog): Implementation of `Mapped Diagnostic Context` interface using `zerolog` +* [zerologr](https://github.com/go-logr/zerologr): Implementation of `logr.LogSink` interface using `zerolog` + +## Benchmarks + +See [logbench](http://hackemist.com/logbench/) for more comprehensive and up-to-date benchmarks. + +All operations are allocation free (those numbers *include* JSON encoding): + +```text +BenchmarkLogEmpty-8 100000000 19.1 ns/op 0 B/op 0 allocs/op +BenchmarkDisabled-8 500000000 4.07 ns/op 0 B/op 0 allocs/op +BenchmarkInfo-8 30000000 42.5 ns/op 0 B/op 0 allocs/op +BenchmarkContextFields-8 30000000 44.9 ns/op 0 B/op 0 allocs/op +BenchmarkLogFields-8 10000000 184 ns/op 0 B/op 0 allocs/op +``` + +There are a few Go logging benchmarks and comparisons that include zerolog. + +* [imkira/go-loggers-bench](https://github.com/imkira/go-loggers-bench) +* [uber-common/zap](https://github.com/uber-go/zap#performance) + +Using Uber's zap comparison benchmark: + +Log a message and 10 fields: + +| Library | Time | Bytes Allocated | Objects Allocated | +| :--- | :---: | :---: | :---: | +| zerolog | 767 ns/op | 552 B/op | 6 allocs/op | +| :zap: zap | 848 ns/op | 704 B/op | 2 allocs/op | +| :zap: zap (sugared) | 1363 ns/op | 1610 B/op | 20 allocs/op | +| go-kit | 3614 ns/op | 2895 B/op | 66 allocs/op | +| lion | 5392 ns/op | 5807 B/op | 63 allocs/op | +| logrus | 5661 ns/op | 6092 B/op | 78 allocs/op | +| apex/log | 15332 ns/op | 3832 B/op | 65 allocs/op | +| log15 | 20657 ns/op | 5632 B/op | 93 allocs/op | + +Log a message with a logger that already has 10 fields of context: + +| Library | Time | Bytes Allocated | Objects Allocated | +| :--- | :---: | :---: | :---: | +| zerolog | 52 ns/op | 0 B/op | 0 allocs/op | +| :zap: zap | 283 ns/op | 0 B/op | 0 allocs/op | +| :zap: zap (sugared) | 337 ns/op | 80 B/op | 2 allocs/op | +| lion | 2702 ns/op | 4074 B/op | 38 allocs/op | +| go-kit | 3378 ns/op | 3046 B/op | 52 allocs/op | +| logrus | 4309 ns/op | 4564 B/op | 63 allocs/op | +| apex/log | 13456 ns/op | 2898 B/op | 51 allocs/op | +| log15 | 14179 ns/op | 2642 B/op | 44 allocs/op | + +Log a static string, without any context or `printf`-style templating: + +| Library | Time | Bytes Allocated | Objects Allocated | +| :--- | :---: | :---: | :---: | +| zerolog | 50 ns/op | 0 B/op | 0 allocs/op | +| :zap: zap | 236 ns/op | 0 B/op | 0 allocs/op | +| standard library | 453 ns/op | 80 B/op | 2 allocs/op | +| :zap: zap (sugared) | 337 ns/op | 80 B/op | 2 allocs/op | +| go-kit | 508 ns/op | 656 B/op | 13 allocs/op | +| lion | 771 ns/op | 1224 B/op | 10 allocs/op | +| logrus | 1244 ns/op | 1505 B/op | 27 allocs/op | +| apex/log | 2751 ns/op | 584 B/op | 11 allocs/op | +| log15 | 5181 ns/op | 1592 B/op | 26 allocs/op | + +## Caveats + +Note that zerolog does no de-duplication of fields. Using the same key multiple times creates multiple keys in final JSON: + +```go +logger := zerolog.New(os.Stderr).With().Timestamp().Logger() +logger.Info(). + Timestamp(). + Msg("dup") +// Output: {"level":"info","time":1494567715,"time":1494567715,"message":"dup"} +``` + +In this case, many consumers will take the last value, but this is not guaranteed; check yours if in doubt. diff --git a/vendor/github.com/rs/zerolog/_config.yml b/vendor/github.com/rs/zerolog/_config.yml new file mode 100644 index 0000000..a1e896d --- /dev/null +++ b/vendor/github.com/rs/zerolog/_config.yml @@ -0,0 +1 @@ +remote_theme: rs/gh-readme diff --git a/vendor/github.com/rs/zerolog/array.go b/vendor/github.com/rs/zerolog/array.go new file mode 100644 index 0000000..666dded --- /dev/null +++ b/vendor/github.com/rs/zerolog/array.go @@ -0,0 +1,240 @@ +package zerolog + +import ( + "net" + "sync" + "time" +) + +var arrayPool = &sync.Pool{ + New: func() interface{} { + return &Array{ + buf: make([]byte, 0, 500), + } + }, +} + +// Array is used to prepopulate an array of items +// which can be re-used to add to log messages. +type Array struct { + buf []byte +} + +func putArray(a *Array) { + // Proper usage of a sync.Pool requires each entry to have approximately + // the same memory cost. To obtain this property when the stored type + // contains a variably-sized buffer, we add a hard limit on the maximum buffer + // to place back in the pool. + // + // See https://golang.org/issue/23199 + const maxSize = 1 << 16 // 64KiB + if cap(a.buf) > maxSize { + return + } + arrayPool.Put(a) +} + +// Arr creates an array to be added to an Event or Context. +func Arr() *Array { + a := arrayPool.Get().(*Array) + a.buf = a.buf[:0] + return a +} + +// MarshalZerologArray method here is no-op - since data is +// already in the needed format. +func (*Array) MarshalZerologArray(*Array) { +} + +func (a *Array) write(dst []byte) []byte { + dst = enc.AppendArrayStart(dst) + if len(a.buf) > 0 { + dst = append(append(dst, a.buf...)) + } + dst = enc.AppendArrayEnd(dst) + putArray(a) + return dst +} + +// Object marshals an object that implement the LogObjectMarshaler +// interface and append append it to the array. +func (a *Array) Object(obj LogObjectMarshaler) *Array { + e := Dict() + obj.MarshalZerologObject(e) + e.buf = enc.AppendEndMarker(e.buf) + a.buf = append(enc.AppendArrayDelim(a.buf), e.buf...) + putEvent(e) + return a +} + +// Str append append the val as a string to the array. +func (a *Array) Str(val string) *Array { + a.buf = enc.AppendString(enc.AppendArrayDelim(a.buf), val) + return a +} + +// Bytes append append the val as a string to the array. +func (a *Array) Bytes(val []byte) *Array { + a.buf = enc.AppendBytes(enc.AppendArrayDelim(a.buf), val) + return a +} + +// Hex append append the val as a hex string to the array. +func (a *Array) Hex(val []byte) *Array { + a.buf = enc.AppendHex(enc.AppendArrayDelim(a.buf), val) + return a +} + +// RawJSON adds already encoded JSON to the array. +func (a *Array) RawJSON(val []byte) *Array { + a.buf = appendJSON(enc.AppendArrayDelim(a.buf), val) + return a +} + +// Err serializes and appends the err to the array. +func (a *Array) Err(err error) *Array { + switch m := ErrorMarshalFunc(err).(type) { + case LogObjectMarshaler: + e := newEvent(nil, 0) + e.buf = e.buf[:0] + e.appendObject(m) + a.buf = append(enc.AppendArrayDelim(a.buf), e.buf...) + putEvent(e) + case error: + if m == nil || isNilValue(m) { + a.buf = enc.AppendNil(enc.AppendArrayDelim(a.buf)) + } else { + a.buf = enc.AppendString(enc.AppendArrayDelim(a.buf), m.Error()) + } + case string: + a.buf = enc.AppendString(enc.AppendArrayDelim(a.buf), m) + default: + a.buf = enc.AppendInterface(enc.AppendArrayDelim(a.buf), m) + } + + return a +} + +// Bool append append the val as a bool to the array. +func (a *Array) Bool(b bool) *Array { + a.buf = enc.AppendBool(enc.AppendArrayDelim(a.buf), b) + return a +} + +// Int append append i as a int to the array. +func (a *Array) Int(i int) *Array { + a.buf = enc.AppendInt(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Int8 append append i as a int8 to the array. +func (a *Array) Int8(i int8) *Array { + a.buf = enc.AppendInt8(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Int16 append append i as a int16 to the array. +func (a *Array) Int16(i int16) *Array { + a.buf = enc.AppendInt16(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Int32 append append i as a int32 to the array. +func (a *Array) Int32(i int32) *Array { + a.buf = enc.AppendInt32(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Int64 append append i as a int64 to the array. +func (a *Array) Int64(i int64) *Array { + a.buf = enc.AppendInt64(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Uint append append i as a uint to the array. +func (a *Array) Uint(i uint) *Array { + a.buf = enc.AppendUint(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Uint8 append append i as a uint8 to the array. +func (a *Array) Uint8(i uint8) *Array { + a.buf = enc.AppendUint8(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Uint16 append append i as a uint16 to the array. +func (a *Array) Uint16(i uint16) *Array { + a.buf = enc.AppendUint16(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Uint32 append append i as a uint32 to the array. +func (a *Array) Uint32(i uint32) *Array { + a.buf = enc.AppendUint32(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Uint64 append append i as a uint64 to the array. +func (a *Array) Uint64(i uint64) *Array { + a.buf = enc.AppendUint64(enc.AppendArrayDelim(a.buf), i) + return a +} + +// Float32 append append f as a float32 to the array. +func (a *Array) Float32(f float32) *Array { + a.buf = enc.AppendFloat32(enc.AppendArrayDelim(a.buf), f) + return a +} + +// Float64 append append f as a float64 to the array. +func (a *Array) Float64(f float64) *Array { + a.buf = enc.AppendFloat64(enc.AppendArrayDelim(a.buf), f) + return a +} + +// Time append append t formated as string using zerolog.TimeFieldFormat. +func (a *Array) Time(t time.Time) *Array { + a.buf = enc.AppendTime(enc.AppendArrayDelim(a.buf), t, TimeFieldFormat) + return a +} + +// Dur append append d to the array. +func (a *Array) Dur(d time.Duration) *Array { + a.buf = enc.AppendDuration(enc.AppendArrayDelim(a.buf), d, DurationFieldUnit, DurationFieldInteger) + return a +} + +// Interface append append i marshaled using reflection. +func (a *Array) Interface(i interface{}) *Array { + if obj, ok := i.(LogObjectMarshaler); ok { + return a.Object(obj) + } + a.buf = enc.AppendInterface(enc.AppendArrayDelim(a.buf), i) + return a +} + +// IPAddr adds IPv4 or IPv6 address to the array +func (a *Array) IPAddr(ip net.IP) *Array { + a.buf = enc.AppendIPAddr(enc.AppendArrayDelim(a.buf), ip) + return a +} + +// IPPrefix adds IPv4 or IPv6 Prefix (IP + mask) to the array +func (a *Array) IPPrefix(pfx net.IPNet) *Array { + a.buf = enc.AppendIPPrefix(enc.AppendArrayDelim(a.buf), pfx) + return a +} + +// MACAddr adds a MAC (Ethernet) address to the array +func (a *Array) MACAddr(ha net.HardwareAddr) *Array { + a.buf = enc.AppendMACAddr(enc.AppendArrayDelim(a.buf), ha) + return a +} + +// Dict adds the dict Event to the array +func (a *Array) Dict(dict *Event) *Array { + dict.buf = enc.AppendEndMarker(dict.buf) + a.buf = append(enc.AppendArrayDelim(a.buf), dict.buf...) + return a +} diff --git a/vendor/github.com/rs/zerolog/console.go b/vendor/github.com/rs/zerolog/console.go new file mode 100644 index 0000000..a8e12b2 --- /dev/null +++ b/vendor/github.com/rs/zerolog/console.go @@ -0,0 +1,409 @@ +package zerolog + +import ( + "bytes" + "encoding/json" + "fmt" + "io" + "os" + "path/filepath" + "sort" + "strconv" + "strings" + "sync" + "time" +) + +const ( + colorBlack = iota + 30 + colorRed + colorGreen + colorYellow + colorBlue + colorMagenta + colorCyan + colorWhite + + colorBold = 1 + colorDarkGray = 90 +) + +var ( + consoleBufPool = sync.Pool{ + New: func() interface{} { + return bytes.NewBuffer(make([]byte, 0, 100)) + }, + } +) + +const ( + consoleDefaultTimeFormat = time.Kitchen +) + +// Formatter transforms the input into a formatted string. +type Formatter func(interface{}) string + +// ConsoleWriter parses the JSON input and writes it in an +// (optionally) colorized, human-friendly format to Out. +type ConsoleWriter struct { + // Out is the output destination. + Out io.Writer + + // NoColor disables the colorized output. + NoColor bool + + // TimeFormat specifies the format for timestamp in output. + TimeFormat string + + // PartsOrder defines the order of parts in output. + PartsOrder []string + + // PartsExclude defines parts to not display in output. + PartsExclude []string + + FormatTimestamp Formatter + FormatLevel Formatter + FormatCaller Formatter + FormatMessage Formatter + FormatFieldName Formatter + FormatFieldValue Formatter + FormatErrFieldName Formatter + FormatErrFieldValue Formatter +} + +// NewConsoleWriter creates and initializes a new ConsoleWriter. +func NewConsoleWriter(options ...func(w *ConsoleWriter)) ConsoleWriter { + w := ConsoleWriter{ + Out: os.Stdout, + TimeFormat: consoleDefaultTimeFormat, + PartsOrder: consoleDefaultPartsOrder(), + } + + for _, opt := range options { + opt(&w) + } + + return w +} + +// Write transforms the JSON input with formatters and appends to w.Out. +func (w ConsoleWriter) Write(p []byte) (n int, err error) { + if w.PartsOrder == nil { + w.PartsOrder = consoleDefaultPartsOrder() + } + + var buf = consoleBufPool.Get().(*bytes.Buffer) + defer func() { + buf.Reset() + consoleBufPool.Put(buf) + }() + + var evt map[string]interface{} + p = decodeIfBinaryToBytes(p) + d := json.NewDecoder(bytes.NewReader(p)) + d.UseNumber() + err = d.Decode(&evt) + if err != nil { + return n, fmt.Errorf("cannot decode event: %s", err) + } + + for _, p := range w.PartsOrder { + w.writePart(buf, evt, p) + } + + w.writeFields(evt, buf) + + err = buf.WriteByte('\n') + if err != nil { + return n, err + } + _, err = buf.WriteTo(w.Out) + return len(p), err +} + +// writeFields appends formatted key-value pairs to buf. +func (w ConsoleWriter) writeFields(evt map[string]interface{}, buf *bytes.Buffer) { + var fields = make([]string, 0, len(evt)) + for field := range evt { + switch field { + case LevelFieldName, TimestampFieldName, MessageFieldName, CallerFieldName: + continue + } + fields = append(fields, field) + } + sort.Strings(fields) + + if len(fields) > 0 { + buf.WriteByte(' ') + } + + // Move the "error" field to the front + ei := sort.Search(len(fields), func(i int) bool { return fields[i] >= ErrorFieldName }) + if ei < len(fields) && fields[ei] == ErrorFieldName { + fields[ei] = "" + fields = append([]string{ErrorFieldName}, fields...) + var xfields = make([]string, 0, len(fields)) + for _, field := range fields { + if field == "" { // Skip empty fields + continue + } + xfields = append(xfields, field) + } + fields = xfields + } + + for i, field := range fields { + var fn Formatter + var fv Formatter + + if field == ErrorFieldName { + if w.FormatErrFieldName == nil { + fn = consoleDefaultFormatErrFieldName(w.NoColor) + } else { + fn = w.FormatErrFieldName + } + + if w.FormatErrFieldValue == nil { + fv = consoleDefaultFormatErrFieldValue(w.NoColor) + } else { + fv = w.FormatErrFieldValue + } + } else { + if w.FormatFieldName == nil { + fn = consoleDefaultFormatFieldName(w.NoColor) + } else { + fn = w.FormatFieldName + } + + if w.FormatFieldValue == nil { + fv = consoleDefaultFormatFieldValue + } else { + fv = w.FormatFieldValue + } + } + + buf.WriteString(fn(field)) + + switch fValue := evt[field].(type) { + case string: + if needsQuote(fValue) { + buf.WriteString(fv(strconv.Quote(fValue))) + } else { + buf.WriteString(fv(fValue)) + } + case json.Number: + buf.WriteString(fv(fValue)) + default: + b, err := json.Marshal(fValue) + if err != nil { + fmt.Fprintf(buf, colorize("[error: %v]", colorRed, w.NoColor), err) + } else { + fmt.Fprint(buf, fv(b)) + } + } + + if i < len(fields)-1 { // Skip space for last field + buf.WriteByte(' ') + } + } +} + +// writePart appends a formatted part to buf. +func (w ConsoleWriter) writePart(buf *bytes.Buffer, evt map[string]interface{}, p string) { + var f Formatter + + if w.PartsExclude != nil && len(w.PartsExclude) > 0 { + for _, exclude := range w.PartsExclude { + if exclude == p { + return + } + } + } + + switch p { + case LevelFieldName: + if w.FormatLevel == nil { + f = consoleDefaultFormatLevel(w.NoColor) + } else { + f = w.FormatLevel + } + case TimestampFieldName: + if w.FormatTimestamp == nil { + f = consoleDefaultFormatTimestamp(w.TimeFormat, w.NoColor) + } else { + f = w.FormatTimestamp + } + case MessageFieldName: + if w.FormatMessage == nil { + f = consoleDefaultFormatMessage + } else { + f = w.FormatMessage + } + case CallerFieldName: + if w.FormatCaller == nil { + f = consoleDefaultFormatCaller(w.NoColor) + } else { + f = w.FormatCaller + } + default: + if w.FormatFieldValue == nil { + f = consoleDefaultFormatFieldValue + } else { + f = w.FormatFieldValue + } + } + + var s = f(evt[p]) + + if len(s) > 0 { + buf.WriteString(s) + if p != w.PartsOrder[len(w.PartsOrder)-1] { // Skip space for last part + buf.WriteByte(' ') + } + } +} + +// needsQuote returns true when the string s should be quoted in output. +func needsQuote(s string) bool { + for i := range s { + if s[i] < 0x20 || s[i] > 0x7e || s[i] == ' ' || s[i] == '\\' || s[i] == '"' { + return true + } + } + return false +} + +// colorize returns the string s wrapped in ANSI code c, unless disabled is true. +func colorize(s interface{}, c int, disabled bool) string { + if disabled { + return fmt.Sprintf("%s", s) + } + return fmt.Sprintf("\x1b[%dm%v\x1b[0m", c, s) +} + +// ----- DEFAULT FORMATTERS --------------------------------------------------- + +func consoleDefaultPartsOrder() []string { + return []string{ + TimestampFieldName, + LevelFieldName, + CallerFieldName, + MessageFieldName, + } +} + +func consoleDefaultFormatTimestamp(timeFormat string, noColor bool) Formatter { + if timeFormat == "" { + timeFormat = consoleDefaultTimeFormat + } + return func(i interface{}) string { + t := "" + switch tt := i.(type) { + case string: + ts, err := time.Parse(TimeFieldFormat, tt) + if err != nil { + t = tt + } else { + t = ts.Format(timeFormat) + } + case json.Number: + i, err := tt.Int64() + if err != nil { + t = tt.String() + } else { + var sec, nsec int64 = i, 0 + switch TimeFieldFormat { + case TimeFormatUnixMs: + nsec = int64(time.Duration(i) * time.Millisecond) + sec = 0 + case TimeFormatUnixMicro: + nsec = int64(time.Duration(i) * time.Microsecond) + sec = 0 + } + ts := time.Unix(sec, nsec).UTC() + t = ts.Format(timeFormat) + } + } + return colorize(t, colorDarkGray, noColor) + } +} + +func consoleDefaultFormatLevel(noColor bool) Formatter { + return func(i interface{}) string { + var l string + if ll, ok := i.(string); ok { + switch ll { + case LevelTraceValue: + l = colorize("TRC", colorMagenta, noColor) + case LevelDebugValue: + l = colorize("DBG", colorYellow, noColor) + case LevelInfoValue: + l = colorize("INF", colorGreen, noColor) + case LevelWarnValue: + l = colorize("WRN", colorRed, noColor) + case LevelErrorValue: + l = colorize(colorize("ERR", colorRed, noColor), colorBold, noColor) + case LevelFatalValue: + l = colorize(colorize("FTL", colorRed, noColor), colorBold, noColor) + case LevelPanicValue: + l = colorize(colorize("PNC", colorRed, noColor), colorBold, noColor) + default: + l = colorize("???", colorBold, noColor) + } + } else { + if i == nil { + l = colorize("???", colorBold, noColor) + } else { + l = strings.ToUpper(fmt.Sprintf("%s", i))[0:3] + } + } + return l + } +} + +func consoleDefaultFormatCaller(noColor bool) Formatter { + return func(i interface{}) string { + var c string + if cc, ok := i.(string); ok { + c = cc + } + if len(c) > 0 { + if cwd, err := os.Getwd(); err == nil { + if rel, err := filepath.Rel(cwd, c); err == nil { + c = rel + } + } + c = colorize(c, colorBold, noColor) + colorize(" >", colorCyan, noColor) + } + return c + } +} + +func consoleDefaultFormatMessage(i interface{}) string { + if i == nil { + return "" + } + return fmt.Sprintf("%s", i) +} + +func consoleDefaultFormatFieldName(noColor bool) Formatter { + return func(i interface{}) string { + return colorize(fmt.Sprintf("%s=", i), colorCyan, noColor) + } +} + +func consoleDefaultFormatFieldValue(i interface{}) string { + return fmt.Sprintf("%s", i) +} + +func consoleDefaultFormatErrFieldName(noColor bool) Formatter { + return func(i interface{}) string { + return colorize(fmt.Sprintf("%s=", i), colorCyan, noColor) + } +} + +func consoleDefaultFormatErrFieldValue(noColor bool) Formatter { + return func(i interface{}) string { + return colorize(fmt.Sprintf("%s", i), colorRed, noColor) + } +} diff --git a/vendor/github.com/rs/zerolog/context.go b/vendor/github.com/rs/zerolog/context.go new file mode 100644 index 0000000..f398e31 --- /dev/null +++ b/vendor/github.com/rs/zerolog/context.go @@ -0,0 +1,433 @@ +package zerolog + +import ( + "fmt" + "io/ioutil" + "math" + "net" + "time" +) + +// Context configures a new sub-logger with contextual fields. +type Context struct { + l Logger +} + +// Logger returns the logger with the context previously set. +func (c Context) Logger() Logger { + return c.l +} + +// Fields is a helper function to use a map or slice to set fields using type assertion. +// Only map[string]interface{} and []interface{} are accepted. []interface{} must +// alternate string keys and arbitrary values, and extraneous ones are ignored. +func (c Context) Fields(fields interface{}) Context { + c.l.context = appendFields(c.l.context, fields) + return c +} + +// Dict adds the field key with the dict to the logger context. +func (c Context) Dict(key string, dict *Event) Context { + dict.buf = enc.AppendEndMarker(dict.buf) + c.l.context = append(enc.AppendKey(c.l.context, key), dict.buf...) + putEvent(dict) + return c +} + +// Array adds the field key with an array to the event context. +// Use zerolog.Arr() to create the array or pass a type that +// implement the LogArrayMarshaler interface. +func (c Context) Array(key string, arr LogArrayMarshaler) Context { + c.l.context = enc.AppendKey(c.l.context, key) + if arr, ok := arr.(*Array); ok { + c.l.context = arr.write(c.l.context) + return c + } + var a *Array + if aa, ok := arr.(*Array); ok { + a = aa + } else { + a = Arr() + arr.MarshalZerologArray(a) + } + c.l.context = a.write(c.l.context) + return c +} + +// Object marshals an object that implement the LogObjectMarshaler interface. +func (c Context) Object(key string, obj LogObjectMarshaler) Context { + e := newEvent(levelWriterAdapter{ioutil.Discard}, 0) + e.Object(key, obj) + c.l.context = enc.AppendObjectData(c.l.context, e.buf) + putEvent(e) + return c +} + +// EmbedObject marshals and Embeds an object that implement the LogObjectMarshaler interface. +func (c Context) EmbedObject(obj LogObjectMarshaler) Context { + e := newEvent(levelWriterAdapter{ioutil.Discard}, 0) + e.EmbedObject(obj) + c.l.context = enc.AppendObjectData(c.l.context, e.buf) + putEvent(e) + return c +} + +// Str adds the field key with val as a string to the logger context. +func (c Context) Str(key, val string) Context { + c.l.context = enc.AppendString(enc.AppendKey(c.l.context, key), val) + return c +} + +// Strs adds the field key with val as a string to the logger context. +func (c Context) Strs(key string, vals []string) Context { + c.l.context = enc.AppendStrings(enc.AppendKey(c.l.context, key), vals) + return c +} + +// Stringer adds the field key with val.String() (or null if val is nil) to the logger context. +func (c Context) Stringer(key string, val fmt.Stringer) Context { + if val != nil { + c.l.context = enc.AppendString(enc.AppendKey(c.l.context, key), val.String()) + return c + } + + c.l.context = enc.AppendInterface(enc.AppendKey(c.l.context, key), nil) + return c +} + +// Bytes adds the field key with val as a []byte to the logger context. +func (c Context) Bytes(key string, val []byte) Context { + c.l.context = enc.AppendBytes(enc.AppendKey(c.l.context, key), val) + return c +} + +// Hex adds the field key with val as a hex string to the logger context. +func (c Context) Hex(key string, val []byte) Context { + c.l.context = enc.AppendHex(enc.AppendKey(c.l.context, key), val) + return c +} + +// RawJSON adds already encoded JSON to context. +// +// No sanity check is performed on b; it must not contain carriage returns and +// be valid JSON. +func (c Context) RawJSON(key string, b []byte) Context { + c.l.context = appendJSON(enc.AppendKey(c.l.context, key), b) + return c +} + +// AnErr adds the field key with serialized err to the logger context. +func (c Context) AnErr(key string, err error) Context { + switch m := ErrorMarshalFunc(err).(type) { + case nil: + return c + case LogObjectMarshaler: + return c.Object(key, m) + case error: + if m == nil || isNilValue(m) { + return c + } else { + return c.Str(key, m.Error()) + } + case string: + return c.Str(key, m) + default: + return c.Interface(key, m) + } +} + +// Errs adds the field key with errs as an array of serialized errors to the +// logger context. +func (c Context) Errs(key string, errs []error) Context { + arr := Arr() + for _, err := range errs { + switch m := ErrorMarshalFunc(err).(type) { + case LogObjectMarshaler: + arr = arr.Object(m) + case error: + if m == nil || isNilValue(m) { + arr = arr.Interface(nil) + } else { + arr = arr.Str(m.Error()) + } + case string: + arr = arr.Str(m) + default: + arr = arr.Interface(m) + } + } + + return c.Array(key, arr) +} + +// Err adds the field "error" with serialized err to the logger context. +func (c Context) Err(err error) Context { + return c.AnErr(ErrorFieldName, err) +} + +// Bool adds the field key with val as a bool to the logger context. +func (c Context) Bool(key string, b bool) Context { + c.l.context = enc.AppendBool(enc.AppendKey(c.l.context, key), b) + return c +} + +// Bools adds the field key with val as a []bool to the logger context. +func (c Context) Bools(key string, b []bool) Context { + c.l.context = enc.AppendBools(enc.AppendKey(c.l.context, key), b) + return c +} + +// Int adds the field key with i as a int to the logger context. +func (c Context) Int(key string, i int) Context { + c.l.context = enc.AppendInt(enc.AppendKey(c.l.context, key), i) + return c +} + +// Ints adds the field key with i as a []int to the logger context. +func (c Context) Ints(key string, i []int) Context { + c.l.context = enc.AppendInts(enc.AppendKey(c.l.context, key), i) + return c +} + +// Int8 adds the field key with i as a int8 to the logger context. +func (c Context) Int8(key string, i int8) Context { + c.l.context = enc.AppendInt8(enc.AppendKey(c.l.context, key), i) + return c +} + +// Ints8 adds the field key with i as a []int8 to the logger context. +func (c Context) Ints8(key string, i []int8) Context { + c.l.context = enc.AppendInts8(enc.AppendKey(c.l.context, key), i) + return c +} + +// Int16 adds the field key with i as a int16 to the logger context. +func (c Context) Int16(key string, i int16) Context { + c.l.context = enc.AppendInt16(enc.AppendKey(c.l.context, key), i) + return c +} + +// Ints16 adds the field key with i as a []int16 to the logger context. +func (c Context) Ints16(key string, i []int16) Context { + c.l.context = enc.AppendInts16(enc.AppendKey(c.l.context, key), i) + return c +} + +// Int32 adds the field key with i as a int32 to the logger context. +func (c Context) Int32(key string, i int32) Context { + c.l.context = enc.AppendInt32(enc.AppendKey(c.l.context, key), i) + return c +} + +// Ints32 adds the field key with i as a []int32 to the logger context. +func (c Context) Ints32(key string, i []int32) Context { + c.l.context = enc.AppendInts32(enc.AppendKey(c.l.context, key), i) + return c +} + +// Int64 adds the field key with i as a int64 to the logger context. +func (c Context) Int64(key string, i int64) Context { + c.l.context = enc.AppendInt64(enc.AppendKey(c.l.context, key), i) + return c +} + +// Ints64 adds the field key with i as a []int64 to the logger context. +func (c Context) Ints64(key string, i []int64) Context { + c.l.context = enc.AppendInts64(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uint adds the field key with i as a uint to the logger context. +func (c Context) Uint(key string, i uint) Context { + c.l.context = enc.AppendUint(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uints adds the field key with i as a []uint to the logger context. +func (c Context) Uints(key string, i []uint) Context { + c.l.context = enc.AppendUints(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uint8 adds the field key with i as a uint8 to the logger context. +func (c Context) Uint8(key string, i uint8) Context { + c.l.context = enc.AppendUint8(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uints8 adds the field key with i as a []uint8 to the logger context. +func (c Context) Uints8(key string, i []uint8) Context { + c.l.context = enc.AppendUints8(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uint16 adds the field key with i as a uint16 to the logger context. +func (c Context) Uint16(key string, i uint16) Context { + c.l.context = enc.AppendUint16(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uints16 adds the field key with i as a []uint16 to the logger context. +func (c Context) Uints16(key string, i []uint16) Context { + c.l.context = enc.AppendUints16(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uint32 adds the field key with i as a uint32 to the logger context. +func (c Context) Uint32(key string, i uint32) Context { + c.l.context = enc.AppendUint32(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uints32 adds the field key with i as a []uint32 to the logger context. +func (c Context) Uints32(key string, i []uint32) Context { + c.l.context = enc.AppendUints32(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uint64 adds the field key with i as a uint64 to the logger context. +func (c Context) Uint64(key string, i uint64) Context { + c.l.context = enc.AppendUint64(enc.AppendKey(c.l.context, key), i) + return c +} + +// Uints64 adds the field key with i as a []uint64 to the logger context. +func (c Context) Uints64(key string, i []uint64) Context { + c.l.context = enc.AppendUints64(enc.AppendKey(c.l.context, key), i) + return c +} + +// Float32 adds the field key with f as a float32 to the logger context. +func (c Context) Float32(key string, f float32) Context { + c.l.context = enc.AppendFloat32(enc.AppendKey(c.l.context, key), f) + return c +} + +// Floats32 adds the field key with f as a []float32 to the logger context. +func (c Context) Floats32(key string, f []float32) Context { + c.l.context = enc.AppendFloats32(enc.AppendKey(c.l.context, key), f) + return c +} + +// Float64 adds the field key with f as a float64 to the logger context. +func (c Context) Float64(key string, f float64) Context { + c.l.context = enc.AppendFloat64(enc.AppendKey(c.l.context, key), f) + return c +} + +// Floats64 adds the field key with f as a []float64 to the logger context. +func (c Context) Floats64(key string, f []float64) Context { + c.l.context = enc.AppendFloats64(enc.AppendKey(c.l.context, key), f) + return c +} + +type timestampHook struct{} + +func (ts timestampHook) Run(e *Event, level Level, msg string) { + e.Timestamp() +} + +var th = timestampHook{} + +// Timestamp adds the current local time as UNIX timestamp to the logger context with the "time" key. +// To customize the key name, change zerolog.TimestampFieldName. +// +// NOTE: It won't dedupe the "time" key if the *Context has one already. +func (c Context) Timestamp() Context { + c.l = c.l.Hook(th) + return c +} + +// Time adds the field key with t formated as string using zerolog.TimeFieldFormat. +func (c Context) Time(key string, t time.Time) Context { + c.l.context = enc.AppendTime(enc.AppendKey(c.l.context, key), t, TimeFieldFormat) + return c +} + +// Times adds the field key with t formated as string using zerolog.TimeFieldFormat. +func (c Context) Times(key string, t []time.Time) Context { + c.l.context = enc.AppendTimes(enc.AppendKey(c.l.context, key), t, TimeFieldFormat) + return c +} + +// Dur adds the fields key with d divided by unit and stored as a float. +func (c Context) Dur(key string, d time.Duration) Context { + c.l.context = enc.AppendDuration(enc.AppendKey(c.l.context, key), d, DurationFieldUnit, DurationFieldInteger) + return c +} + +// Durs adds the fields key with d divided by unit and stored as a float. +func (c Context) Durs(key string, d []time.Duration) Context { + c.l.context = enc.AppendDurations(enc.AppendKey(c.l.context, key), d, DurationFieldUnit, DurationFieldInteger) + return c +} + +// Interface adds the field key with obj marshaled using reflection. +func (c Context) Interface(key string, i interface{}) Context { + c.l.context = enc.AppendInterface(enc.AppendKey(c.l.context, key), i) + return c +} + +type callerHook struct { + callerSkipFrameCount int +} + +func newCallerHook(skipFrameCount int) callerHook { + return callerHook{callerSkipFrameCount: skipFrameCount} +} + +func (ch callerHook) Run(e *Event, level Level, msg string) { + switch ch.callerSkipFrameCount { + case useGlobalSkipFrameCount: + // Extra frames to skip (added by hook infra). + e.caller(CallerSkipFrameCount + contextCallerSkipFrameCount) + default: + // Extra frames to skip (added by hook infra). + e.caller(ch.callerSkipFrameCount + contextCallerSkipFrameCount) + } +} + +// useGlobalSkipFrameCount acts as a flag to informat callerHook.Run +// to use the global CallerSkipFrameCount. +const useGlobalSkipFrameCount = math.MinInt32 + +// ch is the default caller hook using the global CallerSkipFrameCount. +var ch = newCallerHook(useGlobalSkipFrameCount) + +// Caller adds the file:line of the caller with the zerolog.CallerFieldName key. +func (c Context) Caller() Context { + c.l = c.l.Hook(ch) + return c +} + +// CallerWithSkipFrameCount adds the file:line of the caller with the zerolog.CallerFieldName key. +// The specified skipFrameCount int will override the global CallerSkipFrameCount for this context's respective logger. +// If set to -1 the global CallerSkipFrameCount will be used. +func (c Context) CallerWithSkipFrameCount(skipFrameCount int) Context { + c.l = c.l.Hook(newCallerHook(skipFrameCount)) + return c +} + +// Stack enables stack trace printing for the error passed to Err(). +func (c Context) Stack() Context { + c.l.stack = true + return c +} + +// IPAddr adds IPv4 or IPv6 Address to the context +func (c Context) IPAddr(key string, ip net.IP) Context { + c.l.context = enc.AppendIPAddr(enc.AppendKey(c.l.context, key), ip) + return c +} + +// IPPrefix adds IPv4 or IPv6 Prefix (address and mask) to the context +func (c Context) IPPrefix(key string, pfx net.IPNet) Context { + c.l.context = enc.AppendIPPrefix(enc.AppendKey(c.l.context, key), pfx) + return c +} + +// MACAddr adds MAC address to the context +func (c Context) MACAddr(key string, ha net.HardwareAddr) Context { + c.l.context = enc.AppendMACAddr(enc.AppendKey(c.l.context, key), ha) + return c +} diff --git a/vendor/github.com/rs/zerolog/ctx.go b/vendor/github.com/rs/zerolog/ctx.go new file mode 100644 index 0000000..966d17e --- /dev/null +++ b/vendor/github.com/rs/zerolog/ctx.go @@ -0,0 +1,51 @@ +package zerolog + +import ( + "context" +) + +var disabledLogger *Logger + +func init() { + SetGlobalLevel(TraceLevel) + l := Nop() + disabledLogger = &l +} + +type ctxKey struct{} + +// WithContext returns a copy of ctx with l associated. If an instance of Logger +// is already in the context, the context is not updated. +// +// For instance, to add a field to an existing logger in the context, use this +// notation: +// +// ctx := r.Context() +// l := zerolog.Ctx(ctx) +// l.UpdateContext(func(c Context) Context { +// return c.Str("bar", "baz") +// }) +func (l *Logger) WithContext(ctx context.Context) context.Context { + if lp, ok := ctx.Value(ctxKey{}).(*Logger); ok { + if lp == l { + // Do not store same logger. + return ctx + } + } else if l.level == Disabled { + // Do not store disabled logger. + return ctx + } + return context.WithValue(ctx, ctxKey{}, l) +} + +// Ctx returns the Logger associated with the ctx. If no logger +// is associated, DefaultContextLogger is returned, unless DefaultContextLogger +// is nil, in which case a disabled logger is returned. +func Ctx(ctx context.Context) *Logger { + if l, ok := ctx.Value(ctxKey{}).(*Logger); ok { + return l + } else if l = DefaultContextLogger; l != nil { + return l + } + return disabledLogger +} diff --git a/vendor/github.com/rs/zerolog/encoder.go b/vendor/github.com/rs/zerolog/encoder.go new file mode 100644 index 0000000..09b24e8 --- /dev/null +++ b/vendor/github.com/rs/zerolog/encoder.go @@ -0,0 +1,56 @@ +package zerolog + +import ( + "net" + "time" +) + +type encoder interface { + AppendArrayDelim(dst []byte) []byte + AppendArrayEnd(dst []byte) []byte + AppendArrayStart(dst []byte) []byte + AppendBeginMarker(dst []byte) []byte + AppendBool(dst []byte, val bool) []byte + AppendBools(dst []byte, vals []bool) []byte + AppendBytes(dst, s []byte) []byte + AppendDuration(dst []byte, d time.Duration, unit time.Duration, useInt bool) []byte + AppendDurations(dst []byte, vals []time.Duration, unit time.Duration, useInt bool) []byte + AppendEndMarker(dst []byte) []byte + AppendFloat32(dst []byte, val float32) []byte + AppendFloat64(dst []byte, val float64) []byte + AppendFloats32(dst []byte, vals []float32) []byte + AppendFloats64(dst []byte, vals []float64) []byte + AppendHex(dst, s []byte) []byte + AppendIPAddr(dst []byte, ip net.IP) []byte + AppendIPPrefix(dst []byte, pfx net.IPNet) []byte + AppendInt(dst []byte, val int) []byte + AppendInt16(dst []byte, val int16) []byte + AppendInt32(dst []byte, val int32) []byte + AppendInt64(dst []byte, val int64) []byte + AppendInt8(dst []byte, val int8) []byte + AppendInterface(dst []byte, i interface{}) []byte + AppendInts(dst []byte, vals []int) []byte + AppendInts16(dst []byte, vals []int16) []byte + AppendInts32(dst []byte, vals []int32) []byte + AppendInts64(dst []byte, vals []int64) []byte + AppendInts8(dst []byte, vals []int8) []byte + AppendKey(dst []byte, key string) []byte + AppendLineBreak(dst []byte) []byte + AppendMACAddr(dst []byte, ha net.HardwareAddr) []byte + AppendNil(dst []byte) []byte + AppendObjectData(dst []byte, o []byte) []byte + AppendString(dst []byte, s string) []byte + AppendStrings(dst []byte, vals []string) []byte + AppendTime(dst []byte, t time.Time, format string) []byte + AppendTimes(dst []byte, vals []time.Time, format string) []byte + AppendUint(dst []byte, val uint) []byte + AppendUint16(dst []byte, val uint16) []byte + AppendUint32(dst []byte, val uint32) []byte + AppendUint64(dst []byte, val uint64) []byte + AppendUint8(dst []byte, val uint8) []byte + AppendUints(dst []byte, vals []uint) []byte + AppendUints16(dst []byte, vals []uint16) []byte + AppendUints32(dst []byte, vals []uint32) []byte + AppendUints64(dst []byte, vals []uint64) []byte + AppendUints8(dst []byte, vals []uint8) []byte +} diff --git a/vendor/github.com/rs/zerolog/encoder_cbor.go b/vendor/github.com/rs/zerolog/encoder_cbor.go new file mode 100644 index 0000000..7b0dafe --- /dev/null +++ b/vendor/github.com/rs/zerolog/encoder_cbor.go @@ -0,0 +1,42 @@ +// +build binary_log + +package zerolog + +// This file contains bindings to do binary encoding. + +import ( + "github.com/rs/zerolog/internal/cbor" +) + +var ( + _ encoder = (*cbor.Encoder)(nil) + + enc = cbor.Encoder{} +) + +func init() { + // using closure to reflect the changes at runtime. + cbor.JSONMarshalFunc = func(v interface{}) ([]byte, error) { + return InterfaceMarshalFunc(v) + } +} + +func appendJSON(dst []byte, j []byte) []byte { + return cbor.AppendEmbeddedJSON(dst, j) +} + +// decodeIfBinaryToString - converts a binary formatted log msg to a +// JSON formatted String Log message. +func decodeIfBinaryToString(in []byte) string { + return cbor.DecodeIfBinaryToString(in) +} + +func decodeObjectToStr(in []byte) string { + return cbor.DecodeObjectToStr(in) +} + +// decodeIfBinaryToBytes - converts a binary formatted log msg to a +// JSON formatted Bytes Log message. +func decodeIfBinaryToBytes(in []byte) []byte { + return cbor.DecodeIfBinaryToBytes(in) +} diff --git a/vendor/github.com/rs/zerolog/encoder_json.go b/vendor/github.com/rs/zerolog/encoder_json.go new file mode 100644 index 0000000..0e0450e --- /dev/null +++ b/vendor/github.com/rs/zerolog/encoder_json.go @@ -0,0 +1,39 @@ +// +build !binary_log + +package zerolog + +// encoder_json.go file contains bindings to generate +// JSON encoded byte stream. + +import ( + "github.com/rs/zerolog/internal/json" +) + +var ( + _ encoder = (*json.Encoder)(nil) + + enc = json.Encoder{} +) + +func init() { + // using closure to reflect the changes at runtime. + json.JSONMarshalFunc = func(v interface{}) ([]byte, error) { + return InterfaceMarshalFunc(v) + } +} + +func appendJSON(dst []byte, j []byte) []byte { + return append(dst, j...) +} + +func decodeIfBinaryToString(in []byte) string { + return string(in) +} + +func decodeObjectToStr(in []byte) string { + return string(in) +} + +func decodeIfBinaryToBytes(in []byte) []byte { + return in +} diff --git a/vendor/github.com/rs/zerolog/event.go b/vendor/github.com/rs/zerolog/event.go new file mode 100644 index 0000000..b409955 --- /dev/null +++ b/vendor/github.com/rs/zerolog/event.go @@ -0,0 +1,773 @@ +package zerolog + +import ( + "fmt" + "net" + "os" + "runtime" + "sync" + "time" +) + +var eventPool = &sync.Pool{ + New: func() interface{} { + return &Event{ + buf: make([]byte, 0, 500), + } + }, +} + +// Event represents a log event. It is instanced by one of the level method of +// Logger and finalized by the Msg or Msgf method. +type Event struct { + buf []byte + w LevelWriter + level Level + done func(msg string) + stack bool // enable error stack trace + ch []Hook // hooks from context + skipFrame int // The number of additional frames to skip when printing the caller. +} + +func putEvent(e *Event) { + // Proper usage of a sync.Pool requires each entry to have approximately + // the same memory cost. To obtain this property when the stored type + // contains a variably-sized buffer, we add a hard limit on the maximum buffer + // to place back in the pool. + // + // See https://golang.org/issue/23199 + const maxSize = 1 << 16 // 64KiB + if cap(e.buf) > maxSize { + return + } + eventPool.Put(e) +} + +// LogObjectMarshaler provides a strongly-typed and encoding-agnostic interface +// to be implemented by types used with Event/Context's Object methods. +type LogObjectMarshaler interface { + MarshalZerologObject(e *Event) +} + +// LogArrayMarshaler provides a strongly-typed and encoding-agnostic interface +// to be implemented by types used with Event/Context's Array methods. +type LogArrayMarshaler interface { + MarshalZerologArray(a *Array) +} + +func newEvent(w LevelWriter, level Level) *Event { + e := eventPool.Get().(*Event) + e.buf = e.buf[:0] + e.ch = nil + e.buf = enc.AppendBeginMarker(e.buf) + e.w = w + e.level = level + e.stack = false + e.skipFrame = 0 + return e +} + +func (e *Event) write() (err error) { + if e == nil { + return nil + } + if e.level != Disabled { + e.buf = enc.AppendEndMarker(e.buf) + e.buf = enc.AppendLineBreak(e.buf) + if e.w != nil { + _, err = e.w.WriteLevel(e.level, e.buf) + } + } + putEvent(e) + return +} + +// Enabled return false if the *Event is going to be filtered out by +// log level or sampling. +func (e *Event) Enabled() bool { + return e != nil && e.level != Disabled +} + +// Discard disables the event so Msg(f) won't print it. +func (e *Event) Discard() *Event { + if e == nil { + return e + } + e.level = Disabled + return nil +} + +// Msg sends the *Event with msg added as the message field if not empty. +// +// NOTICE: once this method is called, the *Event should be disposed. +// Calling Msg twice can have unexpected result. +func (e *Event) Msg(msg string) { + if e == nil { + return + } + e.msg(msg) +} + +// Send is equivalent to calling Msg(""). +// +// NOTICE: once this method is called, the *Event should be disposed. +func (e *Event) Send() { + if e == nil { + return + } + e.msg("") +} + +// Msgf sends the event with formatted msg added as the message field if not empty. +// +// NOTICE: once this method is called, the *Event should be disposed. +// Calling Msgf twice can have unexpected result. +func (e *Event) Msgf(format string, v ...interface{}) { + if e == nil { + return + } + e.msg(fmt.Sprintf(format, v...)) +} + +func (e *Event) msg(msg string) { + for _, hook := range e.ch { + hook.Run(e, e.level, msg) + } + if msg != "" { + e.buf = enc.AppendString(enc.AppendKey(e.buf, MessageFieldName), msg) + } + if e.done != nil { + defer e.done(msg) + } + if err := e.write(); err != nil { + if ErrorHandler != nil { + ErrorHandler(err) + } else { + fmt.Fprintf(os.Stderr, "zerolog: could not write event: %v\n", err) + } + } +} + +// Fields is a helper function to use a map or slice to set fields using type assertion. +// Only map[string]interface{} and []interface{} are accepted. []interface{} must +// alternate string keys and arbitrary values, and extraneous ones are ignored. +func (e *Event) Fields(fields interface{}) *Event { + if e == nil { + return e + } + e.buf = appendFields(e.buf, fields) + return e +} + +// Dict adds the field key with a dict to the event context. +// Use zerolog.Dict() to create the dictionary. +func (e *Event) Dict(key string, dict *Event) *Event { + if e == nil { + return e + } + dict.buf = enc.AppendEndMarker(dict.buf) + e.buf = append(enc.AppendKey(e.buf, key), dict.buf...) + putEvent(dict) + return e +} + +// Dict creates an Event to be used with the *Event.Dict method. +// Call usual field methods like Str, Int etc to add fields to this +// event and give it as argument the *Event.Dict method. +func Dict() *Event { + return newEvent(nil, 0) +} + +// Array adds the field key with an array to the event context. +// Use zerolog.Arr() to create the array or pass a type that +// implement the LogArrayMarshaler interface. +func (e *Event) Array(key string, arr LogArrayMarshaler) *Event { + if e == nil { + return e + } + e.buf = enc.AppendKey(e.buf, key) + var a *Array + if aa, ok := arr.(*Array); ok { + a = aa + } else { + a = Arr() + arr.MarshalZerologArray(a) + } + e.buf = a.write(e.buf) + return e +} + +func (e *Event) appendObject(obj LogObjectMarshaler) { + e.buf = enc.AppendBeginMarker(e.buf) + obj.MarshalZerologObject(e) + e.buf = enc.AppendEndMarker(e.buf) +} + +// Object marshals an object that implement the LogObjectMarshaler interface. +func (e *Event) Object(key string, obj LogObjectMarshaler) *Event { + if e == nil { + return e + } + e.buf = enc.AppendKey(e.buf, key) + if obj == nil { + e.buf = enc.AppendNil(e.buf) + + return e + } + + e.appendObject(obj) + return e +} + +// Func allows an anonymous func to run only if the event is enabled. +func (e *Event) Func(f func(e *Event)) *Event { + if e != nil && e.Enabled() { + f(e) + } + return e +} + +// EmbedObject marshals an object that implement the LogObjectMarshaler interface. +func (e *Event) EmbedObject(obj LogObjectMarshaler) *Event { + if e == nil { + return e + } + if obj == nil { + return e + } + obj.MarshalZerologObject(e) + return e +} + +// Str adds the field key with val as a string to the *Event context. +func (e *Event) Str(key, val string) *Event { + if e == nil { + return e + } + e.buf = enc.AppendString(enc.AppendKey(e.buf, key), val) + return e +} + +// Strs adds the field key with vals as a []string to the *Event context. +func (e *Event) Strs(key string, vals []string) *Event { + if e == nil { + return e + } + e.buf = enc.AppendStrings(enc.AppendKey(e.buf, key), vals) + return e +} + +// Stringer adds the field key with val.String() (or null if val is nil) +// to the *Event context. +func (e *Event) Stringer(key string, val fmt.Stringer) *Event { + if e == nil { + return e + } + e.buf = enc.AppendStringer(enc.AppendKey(e.buf, key), val) + return e +} + +// Stringers adds the field key with vals where each individual val +// is used as val.String() (or null if val is empty) to the *Event +// context. +func (e *Event) Stringers(key string, vals []fmt.Stringer) *Event { + if e == nil { + return e + } + e.buf = enc.AppendStringers(enc.AppendKey(e.buf, key), vals) + return e +} + +// Bytes adds the field key with val as a string to the *Event context. +// +// Runes outside of normal ASCII ranges will be hex-encoded in the resulting +// JSON. +func (e *Event) Bytes(key string, val []byte) *Event { + if e == nil { + return e + } + e.buf = enc.AppendBytes(enc.AppendKey(e.buf, key), val) + return e +} + +// Hex adds the field key with val as a hex string to the *Event context. +func (e *Event) Hex(key string, val []byte) *Event { + if e == nil { + return e + } + e.buf = enc.AppendHex(enc.AppendKey(e.buf, key), val) + return e +} + +// RawJSON adds already encoded JSON to the log line under key. +// +// No sanity check is performed on b; it must not contain carriage returns and +// be valid JSON. +func (e *Event) RawJSON(key string, b []byte) *Event { + if e == nil { + return e + } + e.buf = appendJSON(enc.AppendKey(e.buf, key), b) + return e +} + +// AnErr adds the field key with serialized err to the *Event context. +// If err is nil, no field is added. +func (e *Event) AnErr(key string, err error) *Event { + if e == nil { + return e + } + switch m := ErrorMarshalFunc(err).(type) { + case nil: + return e + case LogObjectMarshaler: + return e.Object(key, m) + case error: + if m == nil || isNilValue(m) { + return e + } else { + return e.Str(key, m.Error()) + } + case string: + return e.Str(key, m) + default: + return e.Interface(key, m) + } +} + +// Errs adds the field key with errs as an array of serialized errors to the +// *Event context. +func (e *Event) Errs(key string, errs []error) *Event { + if e == nil { + return e + } + arr := Arr() + for _, err := range errs { + switch m := ErrorMarshalFunc(err).(type) { + case LogObjectMarshaler: + arr = arr.Object(m) + case error: + arr = arr.Err(m) + case string: + arr = arr.Str(m) + default: + arr = arr.Interface(m) + } + } + + return e.Array(key, arr) +} + +// Err adds the field "error" with serialized err to the *Event context. +// If err is nil, no field is added. +// +// To customize the key name, change zerolog.ErrorFieldName. +// +// If Stack() has been called before and zerolog.ErrorStackMarshaler is defined, +// the err is passed to ErrorStackMarshaler and the result is appended to the +// zerolog.ErrorStackFieldName. +func (e *Event) Err(err error) *Event { + if e == nil { + return e + } + if e.stack && ErrorStackMarshaler != nil { + switch m := ErrorStackMarshaler(err).(type) { + case nil: + case LogObjectMarshaler: + e.Object(ErrorStackFieldName, m) + case error: + if m != nil && !isNilValue(m) { + e.Str(ErrorStackFieldName, m.Error()) + } + case string: + e.Str(ErrorStackFieldName, m) + default: + e.Interface(ErrorStackFieldName, m) + } + } + return e.AnErr(ErrorFieldName, err) +} + +// Stack enables stack trace printing for the error passed to Err(). +// +// ErrorStackMarshaler must be set for this method to do something. +func (e *Event) Stack() *Event { + if e != nil { + e.stack = true + } + return e +} + +// Bool adds the field key with val as a bool to the *Event context. +func (e *Event) Bool(key string, b bool) *Event { + if e == nil { + return e + } + e.buf = enc.AppendBool(enc.AppendKey(e.buf, key), b) + return e +} + +// Bools adds the field key with val as a []bool to the *Event context. +func (e *Event) Bools(key string, b []bool) *Event { + if e == nil { + return e + } + e.buf = enc.AppendBools(enc.AppendKey(e.buf, key), b) + return e +} + +// Int adds the field key with i as a int to the *Event context. +func (e *Event) Int(key string, i int) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInt(enc.AppendKey(e.buf, key), i) + return e +} + +// Ints adds the field key with i as a []int to the *Event context. +func (e *Event) Ints(key string, i []int) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInts(enc.AppendKey(e.buf, key), i) + return e +} + +// Int8 adds the field key with i as a int8 to the *Event context. +func (e *Event) Int8(key string, i int8) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInt8(enc.AppendKey(e.buf, key), i) + return e +} + +// Ints8 adds the field key with i as a []int8 to the *Event context. +func (e *Event) Ints8(key string, i []int8) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInts8(enc.AppendKey(e.buf, key), i) + return e +} + +// Int16 adds the field key with i as a int16 to the *Event context. +func (e *Event) Int16(key string, i int16) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInt16(enc.AppendKey(e.buf, key), i) + return e +} + +// Ints16 adds the field key with i as a []int16 to the *Event context. +func (e *Event) Ints16(key string, i []int16) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInts16(enc.AppendKey(e.buf, key), i) + return e +} + +// Int32 adds the field key with i as a int32 to the *Event context. +func (e *Event) Int32(key string, i int32) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInt32(enc.AppendKey(e.buf, key), i) + return e +} + +// Ints32 adds the field key with i as a []int32 to the *Event context. +func (e *Event) Ints32(key string, i []int32) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInts32(enc.AppendKey(e.buf, key), i) + return e +} + +// Int64 adds the field key with i as a int64 to the *Event context. +func (e *Event) Int64(key string, i int64) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInt64(enc.AppendKey(e.buf, key), i) + return e +} + +// Ints64 adds the field key with i as a []int64 to the *Event context. +func (e *Event) Ints64(key string, i []int64) *Event { + if e == nil { + return e + } + e.buf = enc.AppendInts64(enc.AppendKey(e.buf, key), i) + return e +} + +// Uint adds the field key with i as a uint to the *Event context. +func (e *Event) Uint(key string, i uint) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUint(enc.AppendKey(e.buf, key), i) + return e +} + +// Uints adds the field key with i as a []int to the *Event context. +func (e *Event) Uints(key string, i []uint) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUints(enc.AppendKey(e.buf, key), i) + return e +} + +// Uint8 adds the field key with i as a uint8 to the *Event context. +func (e *Event) Uint8(key string, i uint8) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUint8(enc.AppendKey(e.buf, key), i) + return e +} + +// Uints8 adds the field key with i as a []int8 to the *Event context. +func (e *Event) Uints8(key string, i []uint8) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUints8(enc.AppendKey(e.buf, key), i) + return e +} + +// Uint16 adds the field key with i as a uint16 to the *Event context. +func (e *Event) Uint16(key string, i uint16) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUint16(enc.AppendKey(e.buf, key), i) + return e +} + +// Uints16 adds the field key with i as a []int16 to the *Event context. +func (e *Event) Uints16(key string, i []uint16) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUints16(enc.AppendKey(e.buf, key), i) + return e +} + +// Uint32 adds the field key with i as a uint32 to the *Event context. +func (e *Event) Uint32(key string, i uint32) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUint32(enc.AppendKey(e.buf, key), i) + return e +} + +// Uints32 adds the field key with i as a []int32 to the *Event context. +func (e *Event) Uints32(key string, i []uint32) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUints32(enc.AppendKey(e.buf, key), i) + return e +} + +// Uint64 adds the field key with i as a uint64 to the *Event context. +func (e *Event) Uint64(key string, i uint64) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUint64(enc.AppendKey(e.buf, key), i) + return e +} + +// Uints64 adds the field key with i as a []int64 to the *Event context. +func (e *Event) Uints64(key string, i []uint64) *Event { + if e == nil { + return e + } + e.buf = enc.AppendUints64(enc.AppendKey(e.buf, key), i) + return e +} + +// Float32 adds the field key with f as a float32 to the *Event context. +func (e *Event) Float32(key string, f float32) *Event { + if e == nil { + return e + } + e.buf = enc.AppendFloat32(enc.AppendKey(e.buf, key), f) + return e +} + +// Floats32 adds the field key with f as a []float32 to the *Event context. +func (e *Event) Floats32(key string, f []float32) *Event { + if e == nil { + return e + } + e.buf = enc.AppendFloats32(enc.AppendKey(e.buf, key), f) + return e +} + +// Float64 adds the field key with f as a float64 to the *Event context. +func (e *Event) Float64(key string, f float64) *Event { + if e == nil { + return e + } + e.buf = enc.AppendFloat64(enc.AppendKey(e.buf, key), f) + return e +} + +// Floats64 adds the field key with f as a []float64 to the *Event context. +func (e *Event) Floats64(key string, f []float64) *Event { + if e == nil { + return e + } + e.buf = enc.AppendFloats64(enc.AppendKey(e.buf, key), f) + return e +} + +// Timestamp adds the current local time as UNIX timestamp to the *Event context with the "time" key. +// To customize the key name, change zerolog.TimestampFieldName. +// +// NOTE: It won't dedupe the "time" key if the *Event (or *Context) has one +// already. +func (e *Event) Timestamp() *Event { + if e == nil { + return e + } + e.buf = enc.AppendTime(enc.AppendKey(e.buf, TimestampFieldName), TimestampFunc(), TimeFieldFormat) + return e +} + +// Time adds the field key with t formated as string using zerolog.TimeFieldFormat. +func (e *Event) Time(key string, t time.Time) *Event { + if e == nil { + return e + } + e.buf = enc.AppendTime(enc.AppendKey(e.buf, key), t, TimeFieldFormat) + return e +} + +// Times adds the field key with t formated as string using zerolog.TimeFieldFormat. +func (e *Event) Times(key string, t []time.Time) *Event { + if e == nil { + return e + } + e.buf = enc.AppendTimes(enc.AppendKey(e.buf, key), t, TimeFieldFormat) + return e +} + +// Dur adds the field key with duration d stored as zerolog.DurationFieldUnit. +// If zerolog.DurationFieldInteger is true, durations are rendered as integer +// instead of float. +func (e *Event) Dur(key string, d time.Duration) *Event { + if e == nil { + return e + } + e.buf = enc.AppendDuration(enc.AppendKey(e.buf, key), d, DurationFieldUnit, DurationFieldInteger) + return e +} + +// Durs adds the field key with duration d stored as zerolog.DurationFieldUnit. +// If zerolog.DurationFieldInteger is true, durations are rendered as integer +// instead of float. +func (e *Event) Durs(key string, d []time.Duration) *Event { + if e == nil { + return e + } + e.buf = enc.AppendDurations(enc.AppendKey(e.buf, key), d, DurationFieldUnit, DurationFieldInteger) + return e +} + +// TimeDiff adds the field key with positive duration between time t and start. +// If time t is not greater than start, duration will be 0. +// Duration format follows the same principle as Dur(). +func (e *Event) TimeDiff(key string, t time.Time, start time.Time) *Event { + if e == nil { + return e + } + var d time.Duration + if t.After(start) { + d = t.Sub(start) + } + e.buf = enc.AppendDuration(enc.AppendKey(e.buf, key), d, DurationFieldUnit, DurationFieldInteger) + return e +} + +// Interface adds the field key with i marshaled using reflection. +func (e *Event) Interface(key string, i interface{}) *Event { + if e == nil { + return e + } + if obj, ok := i.(LogObjectMarshaler); ok { + return e.Object(key, obj) + } + e.buf = enc.AppendInterface(enc.AppendKey(e.buf, key), i) + return e +} + +// CallerSkipFrame instructs any future Caller calls to skip the specified number of frames. +// This includes those added via hooks from the context. +func (e *Event) CallerSkipFrame(skip int) *Event { + if e == nil { + return e + } + e.skipFrame += skip + return e +} + +// Caller adds the file:line of the caller with the zerolog.CallerFieldName key. +// The argument skip is the number of stack frames to ascend +// Skip If not passed, use the global variable CallerSkipFrameCount +func (e *Event) Caller(skip ...int) *Event { + sk := CallerSkipFrameCount + if len(skip) > 0 { + sk = skip[0] + CallerSkipFrameCount + } + return e.caller(sk) +} + +func (e *Event) caller(skip int) *Event { + if e == nil { + return e + } + _, file, line, ok := runtime.Caller(skip + e.skipFrame) + if !ok { + return e + } + e.buf = enc.AppendString(enc.AppendKey(e.buf, CallerFieldName), CallerMarshalFunc(file, line)) + return e +} + +// IPAddr adds IPv4 or IPv6 Address to the event +func (e *Event) IPAddr(key string, ip net.IP) *Event { + if e == nil { + return e + } + e.buf = enc.AppendIPAddr(enc.AppendKey(e.buf, key), ip) + return e +} + +// IPPrefix adds IPv4 or IPv6 Prefix (address and mask) to the event +func (e *Event) IPPrefix(key string, pfx net.IPNet) *Event { + if e == nil { + return e + } + e.buf = enc.AppendIPPrefix(enc.AppendKey(e.buf, key), pfx) + return e +} + +// MACAddr adds MAC address to the event +func (e *Event) MACAddr(key string, ha net.HardwareAddr) *Event { + if e == nil { + return e + } + e.buf = enc.AppendMACAddr(enc.AppendKey(e.buf, key), ha) + return e +} diff --git a/vendor/github.com/rs/zerolog/fields.go b/vendor/github.com/rs/zerolog/fields.go new file mode 100644 index 0000000..c1eb5ce --- /dev/null +++ b/vendor/github.com/rs/zerolog/fields.go @@ -0,0 +1,277 @@ +package zerolog + +import ( + "encoding/json" + "net" + "sort" + "time" + "unsafe" +) + +func isNilValue(i interface{}) bool { + return (*[2]uintptr)(unsafe.Pointer(&i))[1] == 0 +} + +func appendFields(dst []byte, fields interface{}) []byte { + switch fields := fields.(type) { + case []interface{}: + if n := len(fields); n&0x1 == 1 { // odd number + fields = fields[:n-1] + } + dst = appendFieldList(dst, fields) + case map[string]interface{}: + keys := make([]string, 0, len(fields)) + for key := range fields { + keys = append(keys, key) + } + sort.Strings(keys) + kv := make([]interface{}, 2) + for _, key := range keys { + kv[0], kv[1] = key, fields[key] + dst = appendFieldList(dst, kv) + } + } + return dst +} + +func appendFieldList(dst []byte, kvList []interface{}) []byte { + for i, n := 0, len(kvList); i < n; i += 2 { + key, val := kvList[i], kvList[i+1] + if key, ok := key.(string); ok { + dst = enc.AppendKey(dst, key) + } else { + continue + } + if val, ok := val.(LogObjectMarshaler); ok { + e := newEvent(nil, 0) + e.buf = e.buf[:0] + e.appendObject(val) + dst = append(dst, e.buf...) + putEvent(e) + continue + } + switch val := val.(type) { + case string: + dst = enc.AppendString(dst, val) + case []byte: + dst = enc.AppendBytes(dst, val) + case error: + switch m := ErrorMarshalFunc(val).(type) { + case LogObjectMarshaler: + e := newEvent(nil, 0) + e.buf = e.buf[:0] + e.appendObject(m) + dst = append(dst, e.buf...) + putEvent(e) + case error: + if m == nil || isNilValue(m) { + dst = enc.AppendNil(dst) + } else { + dst = enc.AppendString(dst, m.Error()) + } + case string: + dst = enc.AppendString(dst, m) + default: + dst = enc.AppendInterface(dst, m) + } + case []error: + dst = enc.AppendArrayStart(dst) + for i, err := range val { + switch m := ErrorMarshalFunc(err).(type) { + case LogObjectMarshaler: + e := newEvent(nil, 0) + e.buf = e.buf[:0] + e.appendObject(m) + dst = append(dst, e.buf...) + putEvent(e) + case error: + if m == nil || isNilValue(m) { + dst = enc.AppendNil(dst) + } else { + dst = enc.AppendString(dst, m.Error()) + } + case string: + dst = enc.AppendString(dst, m) + default: + dst = enc.AppendInterface(dst, m) + } + + if i < (len(val) - 1) { + enc.AppendArrayDelim(dst) + } + } + dst = enc.AppendArrayEnd(dst) + case bool: + dst = enc.AppendBool(dst, val) + case int: + dst = enc.AppendInt(dst, val) + case int8: + dst = enc.AppendInt8(dst, val) + case int16: + dst = enc.AppendInt16(dst, val) + case int32: + dst = enc.AppendInt32(dst, val) + case int64: + dst = enc.AppendInt64(dst, val) + case uint: + dst = enc.AppendUint(dst, val) + case uint8: + dst = enc.AppendUint8(dst, val) + case uint16: + dst = enc.AppendUint16(dst, val) + case uint32: + dst = enc.AppendUint32(dst, val) + case uint64: + dst = enc.AppendUint64(dst, val) + case float32: + dst = enc.AppendFloat32(dst, val) + case float64: + dst = enc.AppendFloat64(dst, val) + case time.Time: + dst = enc.AppendTime(dst, val, TimeFieldFormat) + case time.Duration: + dst = enc.AppendDuration(dst, val, DurationFieldUnit, DurationFieldInteger) + case *string: + if val != nil { + dst = enc.AppendString(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *bool: + if val != nil { + dst = enc.AppendBool(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *int: + if val != nil { + dst = enc.AppendInt(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *int8: + if val != nil { + dst = enc.AppendInt8(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *int16: + if val != nil { + dst = enc.AppendInt16(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *int32: + if val != nil { + dst = enc.AppendInt32(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *int64: + if val != nil { + dst = enc.AppendInt64(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *uint: + if val != nil { + dst = enc.AppendUint(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *uint8: + if val != nil { + dst = enc.AppendUint8(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *uint16: + if val != nil { + dst = enc.AppendUint16(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *uint32: + if val != nil { + dst = enc.AppendUint32(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *uint64: + if val != nil { + dst = enc.AppendUint64(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *float32: + if val != nil { + dst = enc.AppendFloat32(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *float64: + if val != nil { + dst = enc.AppendFloat64(dst, *val) + } else { + dst = enc.AppendNil(dst) + } + case *time.Time: + if val != nil { + dst = enc.AppendTime(dst, *val, TimeFieldFormat) + } else { + dst = enc.AppendNil(dst) + } + case *time.Duration: + if val != nil { + dst = enc.AppendDuration(dst, *val, DurationFieldUnit, DurationFieldInteger) + } else { + dst = enc.AppendNil(dst) + } + case []string: + dst = enc.AppendStrings(dst, val) + case []bool: + dst = enc.AppendBools(dst, val) + case []int: + dst = enc.AppendInts(dst, val) + case []int8: + dst = enc.AppendInts8(dst, val) + case []int16: + dst = enc.AppendInts16(dst, val) + case []int32: + dst = enc.AppendInts32(dst, val) + case []int64: + dst = enc.AppendInts64(dst, val) + case []uint: + dst = enc.AppendUints(dst, val) + // case []uint8: + // dst = enc.AppendUints8(dst, val) + case []uint16: + dst = enc.AppendUints16(dst, val) + case []uint32: + dst = enc.AppendUints32(dst, val) + case []uint64: + dst = enc.AppendUints64(dst, val) + case []float32: + dst = enc.AppendFloats32(dst, val) + case []float64: + dst = enc.AppendFloats64(dst, val) + case []time.Time: + dst = enc.AppendTimes(dst, val, TimeFieldFormat) + case []time.Duration: + dst = enc.AppendDurations(dst, val, DurationFieldUnit, DurationFieldInteger) + case nil: + dst = enc.AppendNil(dst) + case net.IP: + dst = enc.AppendIPAddr(dst, val) + case net.IPNet: + dst = enc.AppendIPPrefix(dst, val) + case net.HardwareAddr: + dst = enc.AppendMACAddr(dst, val) + case json.RawMessage: + dst = appendJSON(dst, val) + default: + dst = enc.AppendInterface(dst, val) + } + } + return dst +} diff --git a/vendor/github.com/rs/zerolog/globals.go b/vendor/github.com/rs/zerolog/globals.go new file mode 100644 index 0000000..e561d8f --- /dev/null +++ b/vendor/github.com/rs/zerolog/globals.go @@ -0,0 +1,138 @@ +package zerolog + +import ( + "encoding/json" + "strconv" + "sync/atomic" + "time" +) + +const ( + // TimeFormatUnix defines a time format that makes time fields to be + // serialized as Unix timestamp integers. + TimeFormatUnix = "" + + // TimeFormatUnixMs defines a time format that makes time fields to be + // serialized as Unix timestamp integers in milliseconds. + TimeFormatUnixMs = "UNIXMS" + + // TimeFormatUnixMicro defines a time format that makes time fields to be + // serialized as Unix timestamp integers in microseconds. + TimeFormatUnixMicro = "UNIXMICRO" +) + +var ( + // TimestampFieldName is the field name used for the timestamp field. + TimestampFieldName = "time" + + // LevelFieldName is the field name used for the level field. + LevelFieldName = "level" + + // LevelTraceValue is the value used for the trace level field. + LevelTraceValue = "trace" + // LevelDebugValue is the value used for the debug level field. + LevelDebugValue = "debug" + // LevelInfoValue is the value used for the info level field. + LevelInfoValue = "info" + // LevelWarnValue is the value used for the warn level field. + LevelWarnValue = "warn" + // LevelErrorValue is the value used for the error level field. + LevelErrorValue = "error" + // LevelFatalValue is the value used for the fatal level field. + LevelFatalValue = "fatal" + // LevelPanicValue is the value used for the panic level field. + LevelPanicValue = "panic" + + // LevelFieldMarshalFunc allows customization of global level field marshaling. + LevelFieldMarshalFunc = func(l Level) string { + return l.String() + } + + // MessageFieldName is the field name used for the message field. + MessageFieldName = "message" + + // ErrorFieldName is the field name used for error fields. + ErrorFieldName = "error" + + // CallerFieldName is the field name used for caller field. + CallerFieldName = "caller" + + // CallerSkipFrameCount is the number of stack frames to skip to find the caller. + CallerSkipFrameCount = 2 + + // CallerMarshalFunc allows customization of global caller marshaling + CallerMarshalFunc = func(file string, line int) string { + return file + ":" + strconv.Itoa(line) + } + + // ErrorStackFieldName is the field name used for error stacks. + ErrorStackFieldName = "stack" + + // ErrorStackMarshaler extract the stack from err if any. + ErrorStackMarshaler func(err error) interface{} + + // ErrorMarshalFunc allows customization of global error marshaling + ErrorMarshalFunc = func(err error) interface{} { + return err + } + + // InterfaceMarshalFunc allows customization of interface marshaling. + // Default: "encoding/json.Marshal" + InterfaceMarshalFunc = json.Marshal + + // TimeFieldFormat defines the time format of the Time field type. If set to + // TimeFormatUnix, TimeFormatUnixMs or TimeFormatUnixMicro, the time is formatted as an UNIX + // timestamp as integer. + TimeFieldFormat = time.RFC3339 + + // TimestampFunc defines the function called to generate a timestamp. + TimestampFunc = time.Now + + // DurationFieldUnit defines the unit for time.Duration type fields added + // using the Dur method. + DurationFieldUnit = time.Millisecond + + // DurationFieldInteger renders Dur fields as integer instead of float if + // set to true. + DurationFieldInteger = false + + // ErrorHandler is called whenever zerolog fails to write an event on its + // output. If not set, an error is printed on the stderr. This handler must + // be thread safe and non-blocking. + ErrorHandler func(err error) + + // DefaultContextLogger is returned from Ctx() if there is no logger associated + // with the context. + DefaultContextLogger *Logger +) + +var ( + gLevel = new(int32) + disableSampling = new(int32) +) + +// SetGlobalLevel sets the global override for log level. If this +// values is raised, all Loggers will use at least this value. +// +// To globally disable logs, set GlobalLevel to Disabled. +func SetGlobalLevel(l Level) { + atomic.StoreInt32(gLevel, int32(l)) +} + +// GlobalLevel returns the current global log level +func GlobalLevel() Level { + return Level(atomic.LoadInt32(gLevel)) +} + +// DisableSampling will disable sampling in all Loggers if true. +func DisableSampling(v bool) { + var i int32 + if v { + i = 1 + } + atomic.StoreInt32(disableSampling, i) +} + +func samplingDisabled() bool { + return atomic.LoadInt32(disableSampling) == 1 +} diff --git a/vendor/github.com/rs/zerolog/go112.go b/vendor/github.com/rs/zerolog/go112.go new file mode 100644 index 0000000..e7b5a1b --- /dev/null +++ b/vendor/github.com/rs/zerolog/go112.go @@ -0,0 +1,7 @@ +// +build go1.12 + +package zerolog + +// Since go 1.12, some auto generated init functions are hidden from +// runtime.Caller. +const contextCallerSkipFrameCount = 2 diff --git a/vendor/github.com/rs/zerolog/hook.go b/vendor/github.com/rs/zerolog/hook.go new file mode 100644 index 0000000..ec6effc --- /dev/null +++ b/vendor/github.com/rs/zerolog/hook.go @@ -0,0 +1,64 @@ +package zerolog + +// Hook defines an interface to a log hook. +type Hook interface { + // Run runs the hook with the event. + Run(e *Event, level Level, message string) +} + +// HookFunc is an adaptor to allow the use of an ordinary function +// as a Hook. +type HookFunc func(e *Event, level Level, message string) + +// Run implements the Hook interface. +func (h HookFunc) Run(e *Event, level Level, message string) { + h(e, level, message) +} + +// LevelHook applies a different hook for each level. +type LevelHook struct { + NoLevelHook, TraceHook, DebugHook, InfoHook, WarnHook, ErrorHook, FatalHook, PanicHook Hook +} + +// Run implements the Hook interface. +func (h LevelHook) Run(e *Event, level Level, message string) { + switch level { + case TraceLevel: + if h.TraceHook != nil { + h.TraceHook.Run(e, level, message) + } + case DebugLevel: + if h.DebugHook != nil { + h.DebugHook.Run(e, level, message) + } + case InfoLevel: + if h.InfoHook != nil { + h.InfoHook.Run(e, level, message) + } + case WarnLevel: + if h.WarnHook != nil { + h.WarnHook.Run(e, level, message) + } + case ErrorLevel: + if h.ErrorHook != nil { + h.ErrorHook.Run(e, level, message) + } + case FatalLevel: + if h.FatalHook != nil { + h.FatalHook.Run(e, level, message) + } + case PanicLevel: + if h.PanicHook != nil { + h.PanicHook.Run(e, level, message) + } + case NoLevel: + if h.NoLevelHook != nil { + h.NoLevelHook.Run(e, level, message) + } + } +} + +// NewLevelHook returns a new LevelHook. +func NewLevelHook() LevelHook { + return LevelHook{} +} diff --git a/vendor/github.com/rs/zerolog/internal/cbor/README.md b/vendor/github.com/rs/zerolog/internal/cbor/README.md new file mode 100644 index 0000000..92c2e8c --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/README.md @@ -0,0 +1,56 @@ +## Reference: + CBOR Encoding is described in [RFC7049](https://tools.ietf.org/html/rfc7049) + +## Comparison of JSON vs CBOR + +Two main areas of reduction are: + +1. CPU usage to write a log msg +2. Size (in bytes) of log messages. + + +CPU Usage savings are below: +``` +name JSON time/op CBOR time/op delta +Info-32 15.3ns ± 1% 11.7ns ± 3% -23.78% (p=0.000 n=9+10) +ContextFields-32 16.2ns ± 2% 12.3ns ± 3% -23.97% (p=0.000 n=9+9) +ContextAppend-32 6.70ns ± 0% 6.20ns ± 0% -7.44% (p=0.000 n=9+9) +LogFields-32 66.4ns ± 0% 24.6ns ± 2% -62.89% (p=0.000 n=10+9) +LogArrayObject-32 911ns ±11% 768ns ± 6% -15.64% (p=0.000 n=10+10) +LogFieldType/Floats-32 70.3ns ± 2% 29.5ns ± 1% -57.98% (p=0.000 n=10+10) +LogFieldType/Err-32 14.0ns ± 3% 12.1ns ± 8% -13.20% (p=0.000 n=8+10) +LogFieldType/Dur-32 17.2ns ± 2% 13.1ns ± 1% -24.27% (p=0.000 n=10+9) +LogFieldType/Object-32 54.3ns ±11% 52.3ns ± 7% ~ (p=0.239 n=10+10) +LogFieldType/Ints-32 20.3ns ± 2% 15.1ns ± 2% -25.50% (p=0.000 n=9+10) +LogFieldType/Interfaces-32 642ns ±11% 621ns ± 9% ~ (p=0.118 n=10+10) +LogFieldType/Interface(Objects)-32 635ns ±13% 632ns ± 9% ~ (p=0.592 n=10+10) +LogFieldType/Times-32 294ns ± 0% 27ns ± 1% -90.71% (p=0.000 n=10+9) +LogFieldType/Durs-32 121ns ± 0% 33ns ± 2% -72.44% (p=0.000 n=9+9) +LogFieldType/Interface(Object)-32 56.6ns ± 8% 52.3ns ± 8% -7.54% (p=0.007 n=10+10) +LogFieldType/Errs-32 17.8ns ± 3% 16.1ns ± 2% -9.71% (p=0.000 n=10+9) +LogFieldType/Time-32 40.5ns ± 1% 12.7ns ± 6% -68.66% (p=0.000 n=8+9) +LogFieldType/Bool-32 12.0ns ± 5% 10.2ns ± 2% -15.18% (p=0.000 n=10+8) +LogFieldType/Bools-32 17.2ns ± 2% 12.6ns ± 4% -26.63% (p=0.000 n=10+10) +LogFieldType/Int-32 12.3ns ± 2% 11.2ns ± 4% -9.27% (p=0.000 n=9+10) +LogFieldType/Float-32 16.7ns ± 1% 12.6ns ± 2% -24.42% (p=0.000 n=7+9) +LogFieldType/Str-32 12.7ns ± 7% 11.3ns ± 7% -10.88% (p=0.000 n=10+9) +LogFieldType/Strs-32 20.3ns ± 3% 18.2ns ± 3% -10.25% (p=0.000 n=9+10) +LogFieldType/Interface-32 183ns ±12% 175ns ± 9% ~ (p=0.078 n=10+10) +``` + +Log message size savings is greatly dependent on the number and type of fields in the log message. +Assuming this log message (with an Integer, timestamp and string, in addition to level). + +`{"level":"error","Fault":41650,"time":"2018-04-01T15:18:19-07:00","message":"Some Message"}` + +Two measurements were done for the log file sizes - one without any compression, second +using [compress/zlib](https://golang.org/pkg/compress/zlib/). + +Results for 10,000 log messages: + +| Log Format | Plain File Size (in KB) | Compressed File Size (in KB) | +| :--- | :---: | :---: | +| JSON | 920 | 28 | +| CBOR | 550 | 28 | + +The example used to calculate the above data is available in [Examples](examples). diff --git a/vendor/github.com/rs/zerolog/internal/cbor/base.go b/vendor/github.com/rs/zerolog/internal/cbor/base.go new file mode 100644 index 0000000..51fe86c --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/base.go @@ -0,0 +1,19 @@ +package cbor + +// JSONMarshalFunc is used to marshal interface to JSON encoded byte slice. +// Making it package level instead of embedded in Encoder brings +// some extra efforts at importing, but avoids value copy when the functions +// of Encoder being invoked. +// DO REMEMBER to set this variable at importing, or +// you might get a nil pointer dereference panic at runtime. +var JSONMarshalFunc func(v interface{}) ([]byte, error) + +type Encoder struct{} + +// AppendKey adds a key (string) to the binary encoded log message +func (e Encoder) AppendKey(dst []byte, key string) []byte { + if len(dst) < 1 { + dst = e.AppendBeginMarker(dst) + } + return e.AppendString(dst, key) +} diff --git a/vendor/github.com/rs/zerolog/internal/cbor/cbor.go b/vendor/github.com/rs/zerolog/internal/cbor/cbor.go new file mode 100644 index 0000000..969f591 --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/cbor.go @@ -0,0 +1,100 @@ +// Package cbor provides primitives for storing different data +// in the CBOR (binary) format. CBOR is defined in RFC7049. +package cbor + +import "time" + +const ( + majorOffset = 5 + additionalMax = 23 + + // Non Values. + additionalTypeBoolFalse byte = 20 + additionalTypeBoolTrue byte = 21 + additionalTypeNull byte = 22 + + // Integer (+ve and -ve) Sub-types. + additionalTypeIntUint8 byte = 24 + additionalTypeIntUint16 byte = 25 + additionalTypeIntUint32 byte = 26 + additionalTypeIntUint64 byte = 27 + + // Float Sub-types. + additionalTypeFloat16 byte = 25 + additionalTypeFloat32 byte = 26 + additionalTypeFloat64 byte = 27 + additionalTypeBreak byte = 31 + + // Tag Sub-types. + additionalTypeTimestamp byte = 01 + + // Extended Tags - from https://www.iana.org/assignments/cbor-tags/cbor-tags.xhtml + additionalTypeTagNetworkAddr uint16 = 260 + additionalTypeTagNetworkPrefix uint16 = 261 + additionalTypeEmbeddedJSON uint16 = 262 + additionalTypeTagHexString uint16 = 263 + + // Unspecified number of elements. + additionalTypeInfiniteCount byte = 31 +) +const ( + majorTypeUnsignedInt byte = iota << majorOffset // Major type 0 + majorTypeNegativeInt // Major type 1 + majorTypeByteString // Major type 2 + majorTypeUtf8String // Major type 3 + majorTypeArray // Major type 4 + majorTypeMap // Major type 5 + majorTypeTags // Major type 6 + majorTypeSimpleAndFloat // Major type 7 +) + +const ( + maskOutAdditionalType byte = (7 << majorOffset) + maskOutMajorType byte = 31 +) + +const ( + float32Nan = "\xfa\x7f\xc0\x00\x00" + float32PosInfinity = "\xfa\x7f\x80\x00\x00" + float32NegInfinity = "\xfa\xff\x80\x00\x00" + float64Nan = "\xfb\x7f\xf8\x00\x00\x00\x00\x00\x00" + float64PosInfinity = "\xfb\x7f\xf0\x00\x00\x00\x00\x00\x00" + float64NegInfinity = "\xfb\xff\xf0\x00\x00\x00\x00\x00\x00" +) + +// IntegerTimeFieldFormat indicates the format of timestamp decoded +// from an integer (time in seconds). +var IntegerTimeFieldFormat = time.RFC3339 + +// NanoTimeFieldFormat indicates the format of timestamp decoded +// from a float value (time in seconds and nano seconds). +var NanoTimeFieldFormat = time.RFC3339Nano + +func appendCborTypePrefix(dst []byte, major byte, number uint64) []byte { + byteCount := 8 + var minor byte + switch { + case number < 256: + byteCount = 1 + minor = additionalTypeIntUint8 + + case number < 65536: + byteCount = 2 + minor = additionalTypeIntUint16 + + case number < 4294967296: + byteCount = 4 + minor = additionalTypeIntUint32 + + default: + byteCount = 8 + minor = additionalTypeIntUint64 + + } + dst = append(dst, byte(major|minor)) + byteCount-- + for ; byteCount >= 0; byteCount-- { + dst = append(dst, byte(number>>(uint(byteCount)*8))) + } + return dst +} diff --git a/vendor/github.com/rs/zerolog/internal/cbor/decode_stream.go b/vendor/github.com/rs/zerolog/internal/cbor/decode_stream.go new file mode 100644 index 0000000..e3cf3b7 --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/decode_stream.go @@ -0,0 +1,614 @@ +package cbor + +// This file contains code to decode a stream of CBOR Data into JSON. + +import ( + "bufio" + "bytes" + "fmt" + "io" + "math" + "net" + "runtime" + "strconv" + "strings" + "time" + "unicode/utf8" +) + +var decodeTimeZone *time.Location + +const hexTable = "0123456789abcdef" + +const isFloat32 = 4 +const isFloat64 = 8 + +func readNBytes(src *bufio.Reader, n int) []byte { + ret := make([]byte, n) + for i := 0; i < n; i++ { + ch, e := src.ReadByte() + if e != nil { + panic(fmt.Errorf("Tried to Read %d Bytes.. But hit end of file", n)) + } + ret[i] = ch + } + return ret +} + +func readByte(src *bufio.Reader) byte { + b, e := src.ReadByte() + if e != nil { + panic(fmt.Errorf("Tried to Read 1 Byte.. But hit end of file")) + } + return b +} + +func decodeIntAdditonalType(src *bufio.Reader, minor byte) int64 { + val := int64(0) + if minor <= 23 { + val = int64(minor) + } else { + bytesToRead := 0 + switch minor { + case additionalTypeIntUint8: + bytesToRead = 1 + case additionalTypeIntUint16: + bytesToRead = 2 + case additionalTypeIntUint32: + bytesToRead = 4 + case additionalTypeIntUint64: + bytesToRead = 8 + default: + panic(fmt.Errorf("Invalid Additional Type: %d in decodeInteger (expected <28)", minor)) + } + pb := readNBytes(src, bytesToRead) + for i := 0; i < bytesToRead; i++ { + val = val * 256 + val += int64(pb[i]) + } + } + return val +} + +func decodeInteger(src *bufio.Reader) int64 { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeUnsignedInt && major != majorTypeNegativeInt { + panic(fmt.Errorf("Major type is: %d in decodeInteger!! (expected 0 or 1)", major)) + } + val := decodeIntAdditonalType(src, minor) + if major == 0 { + return val + } + return (-1 - val) +} + +func decodeFloat(src *bufio.Reader) (float64, int) { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeSimpleAndFloat { + panic(fmt.Errorf("Incorrect Major type is: %d in decodeFloat", major)) + } + + switch minor { + case additionalTypeFloat16: + panic(fmt.Errorf("float16 is not suppported in decodeFloat")) + + case additionalTypeFloat32: + pb := readNBytes(src, 4) + switch string(pb) { + case float32Nan: + return math.NaN(), isFloat32 + case float32PosInfinity: + return math.Inf(0), isFloat32 + case float32NegInfinity: + return math.Inf(-1), isFloat32 + } + n := uint32(0) + for i := 0; i < 4; i++ { + n = n * 256 + n += uint32(pb[i]) + } + val := math.Float32frombits(n) + return float64(val), isFloat32 + case additionalTypeFloat64: + pb := readNBytes(src, 8) + switch string(pb) { + case float64Nan: + return math.NaN(), isFloat64 + case float64PosInfinity: + return math.Inf(0), isFloat64 + case float64NegInfinity: + return math.Inf(-1), isFloat64 + } + n := uint64(0) + for i := 0; i < 8; i++ { + n = n * 256 + n += uint64(pb[i]) + } + val := math.Float64frombits(n) + return val, isFloat64 + } + panic(fmt.Errorf("Invalid Additional Type: %d in decodeFloat", minor)) +} + +func decodeStringComplex(dst []byte, s string, pos uint) []byte { + i := int(pos) + start := 0 + + for i < len(s) { + b := s[i] + if b >= utf8.RuneSelf { + r, size := utf8.DecodeRuneInString(s[i:]) + if r == utf8.RuneError && size == 1 { + // In case of error, first append previous simple characters to + // the byte slice if any and append a replacement character code + // in place of the invalid sequence. + if start < i { + dst = append(dst, s[start:i]...) + } + dst = append(dst, `\ufffd`...) + i += size + start = i + continue + } + i += size + continue + } + if b >= 0x20 && b <= 0x7e && b != '\\' && b != '"' { + i++ + continue + } + // We encountered a character that needs to be encoded. + // Let's append the previous simple characters to the byte slice + // and switch our operation to read and encode the remainder + // characters byte-by-byte. + if start < i { + dst = append(dst, s[start:i]...) + } + switch b { + case '"', '\\': + dst = append(dst, '\\', b) + case '\b': + dst = append(dst, '\\', 'b') + case '\f': + dst = append(dst, '\\', 'f') + case '\n': + dst = append(dst, '\\', 'n') + case '\r': + dst = append(dst, '\\', 'r') + case '\t': + dst = append(dst, '\\', 't') + default: + dst = append(dst, '\\', 'u', '0', '0', hexTable[b>>4], hexTable[b&0xF]) + } + i++ + start = i + } + if start < len(s) { + dst = append(dst, s[start:]...) + } + return dst +} + +func decodeString(src *bufio.Reader, noQuotes bool) []byte { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeByteString { + panic(fmt.Errorf("Major type is: %d in decodeString", major)) + } + result := []byte{} + if !noQuotes { + result = append(result, '"') + } + length := decodeIntAdditonalType(src, minor) + len := int(length) + pbs := readNBytes(src, len) + result = append(result, pbs...) + if noQuotes { + return result + } + return append(result, '"') +} + +func decodeUTF8String(src *bufio.Reader) []byte { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeUtf8String { + panic(fmt.Errorf("Major type is: %d in decodeUTF8String", major)) + } + result := []byte{'"'} + length := decodeIntAdditonalType(src, minor) + len := int(length) + pbs := readNBytes(src, len) + + for i := 0; i < len; i++ { + // Check if the character needs encoding. Control characters, slashes, + // and the double quote need json encoding. Bytes above the ascii + // boundary needs utf8 encoding. + if pbs[i] < 0x20 || pbs[i] > 0x7e || pbs[i] == '\\' || pbs[i] == '"' { + // We encountered a character that needs to be encoded. Switch + // to complex version of the algorithm. + dst := []byte{'"'} + dst = decodeStringComplex(dst, string(pbs), uint(i)) + return append(dst, '"') + } + } + // The string has no need for encoding an therefore is directly + // appended to the byte slice. + result = append(result, pbs...) + return append(result, '"') +} + +func array2Json(src *bufio.Reader, dst io.Writer) { + dst.Write([]byte{'['}) + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeArray { + panic(fmt.Errorf("Major type is: %d in array2Json", major)) + } + len := 0 + unSpecifiedCount := false + if minor == additionalTypeInfiniteCount { + unSpecifiedCount = true + } else { + length := decodeIntAdditonalType(src, minor) + len = int(length) + } + for i := 0; unSpecifiedCount || i < len; i++ { + if unSpecifiedCount { + pb, e := src.Peek(1) + if e != nil { + panic(e) + } + if pb[0] == byte(majorTypeSimpleAndFloat|additionalTypeBreak) { + readByte(src) + break + } + } + cbor2JsonOneObject(src, dst) + if unSpecifiedCount { + pb, e := src.Peek(1) + if e != nil { + panic(e) + } + if pb[0] == byte(majorTypeSimpleAndFloat|additionalTypeBreak) { + readByte(src) + break + } + dst.Write([]byte{','}) + } else if i+1 < len { + dst.Write([]byte{','}) + } + } + dst.Write([]byte{']'}) +} + +func map2Json(src *bufio.Reader, dst io.Writer) { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeMap { + panic(fmt.Errorf("Major type is: %d in map2Json", major)) + } + len := 0 + unSpecifiedCount := false + if minor == additionalTypeInfiniteCount { + unSpecifiedCount = true + } else { + length := decodeIntAdditonalType(src, minor) + len = int(length) + } + dst.Write([]byte{'{'}) + for i := 0; unSpecifiedCount || i < len; i++ { + if unSpecifiedCount { + pb, e := src.Peek(1) + if e != nil { + panic(e) + } + if pb[0] == byte(majorTypeSimpleAndFloat|additionalTypeBreak) { + readByte(src) + break + } + } + cbor2JsonOneObject(src, dst) + if i%2 == 0 { + // Even position values are keys. + dst.Write([]byte{':'}) + } else { + if unSpecifiedCount { + pb, e := src.Peek(1) + if e != nil { + panic(e) + } + if pb[0] == byte(majorTypeSimpleAndFloat|additionalTypeBreak) { + readByte(src) + break + } + dst.Write([]byte{','}) + } else if i+1 < len { + dst.Write([]byte{','}) + } + } + } + dst.Write([]byte{'}'}) +} + +func decodeTagData(src *bufio.Reader) []byte { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeTags { + panic(fmt.Errorf("Major type is: %d in decodeTagData", major)) + } + switch minor { + case additionalTypeTimestamp: + return decodeTimeStamp(src) + + // Tag value is larger than 256 (so uint16). + case additionalTypeIntUint16: + val := decodeIntAdditonalType(src, minor) + + switch uint16(val) { + case additionalTypeEmbeddedJSON: + pb := readByte(src) + dataMajor := pb & maskOutAdditionalType + if dataMajor != majorTypeByteString { + panic(fmt.Errorf("Unsupported embedded Type: %d in decodeEmbeddedJSON", dataMajor)) + } + src.UnreadByte() + return decodeString(src, true) + + case additionalTypeTagNetworkAddr: + octets := decodeString(src, true) + ss := []byte{'"'} + switch len(octets) { + case 6: // MAC address. + ha := net.HardwareAddr(octets) + ss = append(append(ss, ha.String()...), '"') + case 4: // IPv4 address. + fallthrough + case 16: // IPv6 address. + ip := net.IP(octets) + ss = append(append(ss, ip.String()...), '"') + default: + panic(fmt.Errorf("Unexpected Network Address length: %d (expected 4,6,16)", len(octets))) + } + return ss + + case additionalTypeTagNetworkPrefix: + pb := readByte(src) + if pb != byte(majorTypeMap|0x1) { + panic(fmt.Errorf("IP Prefix is NOT of MAP of 1 elements as expected")) + } + octets := decodeString(src, true) + val := decodeInteger(src) + ip := net.IP(octets) + var mask net.IPMask + pfxLen := int(val) + if len(octets) == 4 { + mask = net.CIDRMask(pfxLen, 32) + } else { + mask = net.CIDRMask(pfxLen, 128) + } + ipPfx := net.IPNet{IP: ip, Mask: mask} + ss := []byte{'"'} + ss = append(append(ss, ipPfx.String()...), '"') + return ss + + case additionalTypeTagHexString: + octets := decodeString(src, true) + ss := []byte{'"'} + for _, v := range octets { + ss = append(ss, hexTable[v>>4], hexTable[v&0x0f]) + } + return append(ss, '"') + + default: + panic(fmt.Errorf("Unsupported Additional Tag Type: %d in decodeTagData", val)) + } + } + panic(fmt.Errorf("Unsupported Additional Type: %d in decodeTagData", minor)) +} + +func decodeTimeStamp(src *bufio.Reader) []byte { + pb := readByte(src) + src.UnreadByte() + tsMajor := pb & maskOutAdditionalType + if tsMajor == majorTypeUnsignedInt || tsMajor == majorTypeNegativeInt { + n := decodeInteger(src) + t := time.Unix(n, 0) + if decodeTimeZone != nil { + t = t.In(decodeTimeZone) + } else { + t = t.In(time.UTC) + } + tsb := []byte{} + tsb = append(tsb, '"') + tsb = t.AppendFormat(tsb, IntegerTimeFieldFormat) + tsb = append(tsb, '"') + return tsb + } else if tsMajor == majorTypeSimpleAndFloat { + n, _ := decodeFloat(src) + secs := int64(n) + n -= float64(secs) + n *= float64(1e9) + t := time.Unix(secs, int64(n)) + if decodeTimeZone != nil { + t = t.In(decodeTimeZone) + } else { + t = t.In(time.UTC) + } + tsb := []byte{} + tsb = append(tsb, '"') + tsb = t.AppendFormat(tsb, NanoTimeFieldFormat) + tsb = append(tsb, '"') + return tsb + } + panic(fmt.Errorf("TS format is neigther int nor float: %d", tsMajor)) +} + +func decodeSimpleFloat(src *bufio.Reader) []byte { + pb := readByte(src) + major := pb & maskOutAdditionalType + minor := pb & maskOutMajorType + if major != majorTypeSimpleAndFloat { + panic(fmt.Errorf("Major type is: %d in decodeSimpleFloat", major)) + } + switch minor { + case additionalTypeBoolTrue: + return []byte("true") + case additionalTypeBoolFalse: + return []byte("false") + case additionalTypeNull: + return []byte("null") + case additionalTypeFloat16: + fallthrough + case additionalTypeFloat32: + fallthrough + case additionalTypeFloat64: + src.UnreadByte() + v, bc := decodeFloat(src) + ba := []byte{} + switch { + case math.IsNaN(v): + return []byte("\"NaN\"") + case math.IsInf(v, 1): + return []byte("\"+Inf\"") + case math.IsInf(v, -1): + return []byte("\"-Inf\"") + } + if bc == isFloat32 { + ba = strconv.AppendFloat(ba, v, 'f', -1, 32) + } else if bc == isFloat64 { + ba = strconv.AppendFloat(ba, v, 'f', -1, 64) + } else { + panic(fmt.Errorf("Invalid Float precision from decodeFloat: %d", bc)) + } + return ba + default: + panic(fmt.Errorf("Invalid Additional Type: %d in decodeSimpleFloat", minor)) + } +} + +func cbor2JsonOneObject(src *bufio.Reader, dst io.Writer) { + pb, e := src.Peek(1) + if e != nil { + panic(e) + } + major := (pb[0] & maskOutAdditionalType) + + switch major { + case majorTypeUnsignedInt: + fallthrough + case majorTypeNegativeInt: + n := decodeInteger(src) + dst.Write([]byte(strconv.Itoa(int(n)))) + + case majorTypeByteString: + s := decodeString(src, false) + dst.Write(s) + + case majorTypeUtf8String: + s := decodeUTF8String(src) + dst.Write(s) + + case majorTypeArray: + array2Json(src, dst) + + case majorTypeMap: + map2Json(src, dst) + + case majorTypeTags: + s := decodeTagData(src) + dst.Write(s) + + case majorTypeSimpleAndFloat: + s := decodeSimpleFloat(src) + dst.Write(s) + } +} + +func moreBytesToRead(src *bufio.Reader) bool { + _, e := src.ReadByte() + if e == nil { + src.UnreadByte() + return true + } + return false +} + +// Cbor2JsonManyObjects decodes all the CBOR Objects read from src +// reader. It keeps on decoding until reader returns EOF (error when reading). +// Decoded string is written to the dst. At the end of every CBOR Object +// newline is written to the output stream. +// +// Returns error (if any) that was encountered during decode. +// The child functions will generate a panic when error is encountered and +// this function will recover non-runtime Errors and return the reason as error. +func Cbor2JsonManyObjects(src io.Reader, dst io.Writer) (err error) { + defer func() { + if r := recover(); r != nil { + if _, ok := r.(runtime.Error); ok { + panic(r) + } + err = r.(error) + } + }() + bufRdr := bufio.NewReader(src) + for moreBytesToRead(bufRdr) { + cbor2JsonOneObject(bufRdr, dst) + dst.Write([]byte("\n")) + } + return nil +} + +// Detect if the bytes to be printed is Binary or not. +func binaryFmt(p []byte) bool { + if len(p) > 0 && p[0] > 0x7F { + return true + } + return false +} + +func getReader(str string) *bufio.Reader { + return bufio.NewReader(strings.NewReader(str)) +} + +// DecodeIfBinaryToString converts a binary formatted log msg to a +// JSON formatted String Log message - suitable for printing to Console/Syslog. +func DecodeIfBinaryToString(in []byte) string { + if binaryFmt(in) { + var b bytes.Buffer + Cbor2JsonManyObjects(strings.NewReader(string(in)), &b) + return b.String() + } + return string(in) +} + +// DecodeObjectToStr checks if the input is a binary format, if so, +// it will decode a single Object and return the decoded string. +func DecodeObjectToStr(in []byte) string { + if binaryFmt(in) { + var b bytes.Buffer + cbor2JsonOneObject(getReader(string(in)), &b) + return b.String() + } + return string(in) +} + +// DecodeIfBinaryToBytes checks if the input is a binary format, if so, +// it will decode all Objects and return the decoded string as byte array. +func DecodeIfBinaryToBytes(in []byte) []byte { + if binaryFmt(in) { + var b bytes.Buffer + Cbor2JsonManyObjects(bytes.NewReader(in), &b) + return b.Bytes() + } + return in +} diff --git a/vendor/github.com/rs/zerolog/internal/cbor/string.go b/vendor/github.com/rs/zerolog/internal/cbor/string.go new file mode 100644 index 0000000..e7f90df --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/string.go @@ -0,0 +1,95 @@ +package cbor + +import "fmt" + +// AppendStrings encodes and adds an array of strings to the dst byte array. +func (e Encoder) AppendStrings(dst []byte, vals []string) []byte { + major := majorTypeArray + l := len(vals) + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendString(dst, v) + } + return dst +} + +// AppendString encodes and adds a string to the dst byte array. +func (Encoder) AppendString(dst []byte, s string) []byte { + major := majorTypeUtf8String + + l := len(s) + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, majorTypeUtf8String, uint64(l)) + } + return append(dst, s...) +} + +// AppendStringers encodes and adds an array of Stringer values +// to the dst byte array. +func (e Encoder) AppendStringers(dst []byte, vals []fmt.Stringer) []byte { + if len(vals) == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + dst = e.AppendArrayStart(dst) + dst = e.AppendStringer(dst, vals[0]) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = e.AppendStringer(dst, val) + } + } + return e.AppendArrayEnd(dst) +} + +// AppendStringer encodes and adds the Stringer value to the dst +// byte array. +func (e Encoder) AppendStringer(dst []byte, val fmt.Stringer) []byte { + if val == nil { + return e.AppendNil(dst) + } + return e.AppendString(dst, val.String()) +} + +// AppendBytes encodes and adds an array of bytes to the dst byte array. +func (Encoder) AppendBytes(dst, s []byte) []byte { + major := majorTypeByteString + + l := len(s) + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + return append(dst, s...) +} + +// AppendEmbeddedJSON adds a tag and embeds input JSON as such. +func AppendEmbeddedJSON(dst, s []byte) []byte { + major := majorTypeTags + minor := additionalTypeEmbeddedJSON + + // Append the TAG to indicate this is Embedded JSON. + dst = append(dst, byte(major|additionalTypeIntUint16)) + dst = append(dst, byte(minor>>8)) + dst = append(dst, byte(minor&0xff)) + + // Append the JSON Object as Byte String. + major = majorTypeByteString + + l := len(s) + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + return append(dst, s...) +} diff --git a/vendor/github.com/rs/zerolog/internal/cbor/time.go b/vendor/github.com/rs/zerolog/internal/cbor/time.go new file mode 100644 index 0000000..12f6a1d --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/time.go @@ -0,0 +1,93 @@ +package cbor + +import ( + "time" +) + +func appendIntegerTimestamp(dst []byte, t time.Time) []byte { + major := majorTypeTags + minor := additionalTypeTimestamp + dst = append(dst, byte(major|minor)) + secs := t.Unix() + var val uint64 + if secs < 0 { + major = majorTypeNegativeInt + val = uint64(-secs - 1) + } else { + major = majorTypeUnsignedInt + val = uint64(secs) + } + dst = appendCborTypePrefix(dst, major, uint64(val)) + return dst +} + +func (e Encoder) appendFloatTimestamp(dst []byte, t time.Time) []byte { + major := majorTypeTags + minor := additionalTypeTimestamp + dst = append(dst, byte(major|minor)) + secs := t.Unix() + nanos := t.Nanosecond() + var val float64 + val = float64(secs)*1.0 + float64(nanos)*1E-9 + return e.AppendFloat64(dst, val) +} + +// AppendTime encodes and adds a timestamp to the dst byte array. +func (e Encoder) AppendTime(dst []byte, t time.Time, unused string) []byte { + utc := t.UTC() + if utc.Nanosecond() == 0 { + return appendIntegerTimestamp(dst, utc) + } + return e.appendFloatTimestamp(dst, utc) +} + +// AppendTimes encodes and adds an array of timestamps to the dst byte array. +func (e Encoder) AppendTimes(dst []byte, vals []time.Time, unused string) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + + for _, t := range vals { + dst = e.AppendTime(dst, t, unused) + } + return dst +} + +// AppendDuration encodes and adds a duration to the dst byte array. +// useInt field indicates whether to store the duration as seconds (integer) or +// as seconds+nanoseconds (float). +func (e Encoder) AppendDuration(dst []byte, d time.Duration, unit time.Duration, useInt bool) []byte { + if useInt { + return e.AppendInt64(dst, int64(d/unit)) + } + return e.AppendFloat64(dst, float64(d)/float64(unit)) +} + +// AppendDurations encodes and adds an array of durations to the dst byte array. +// useInt field indicates whether to store the duration as seconds (integer) or +// as seconds+nanoseconds (float). +func (e Encoder) AppendDurations(dst []byte, vals []time.Duration, unit time.Duration, useInt bool) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, d := range vals { + dst = e.AppendDuration(dst, d, unit, useInt) + } + return dst +} diff --git a/vendor/github.com/rs/zerolog/internal/cbor/types.go b/vendor/github.com/rs/zerolog/internal/cbor/types.go new file mode 100644 index 0000000..a41c797 --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/cbor/types.go @@ -0,0 +1,477 @@ +package cbor + +import ( + "fmt" + "math" + "net" +) + +// AppendNil inserts a 'Nil' object into the dst byte array. +func (Encoder) AppendNil(dst []byte) []byte { + return append(dst, byte(majorTypeSimpleAndFloat|additionalTypeNull)) +} + +// AppendBeginMarker inserts a map start into the dst byte array. +func (Encoder) AppendBeginMarker(dst []byte) []byte { + return append(dst, byte(majorTypeMap|additionalTypeInfiniteCount)) +} + +// AppendEndMarker inserts a map end into the dst byte array. +func (Encoder) AppendEndMarker(dst []byte) []byte { + return append(dst, byte(majorTypeSimpleAndFloat|additionalTypeBreak)) +} + +// AppendObjectData takes an object in form of a byte array and appends to dst. +func (Encoder) AppendObjectData(dst []byte, o []byte) []byte { + // BeginMarker is present in the dst, which + // should not be copied when appending to existing data. + return append(dst, o[1:]...) +} + +// AppendArrayStart adds markers to indicate the start of an array. +func (Encoder) AppendArrayStart(dst []byte) []byte { + return append(dst, byte(majorTypeArray|additionalTypeInfiniteCount)) +} + +// AppendArrayEnd adds markers to indicate the end of an array. +func (Encoder) AppendArrayEnd(dst []byte) []byte { + return append(dst, byte(majorTypeSimpleAndFloat|additionalTypeBreak)) +} + +// AppendArrayDelim adds markers to indicate end of a particular array element. +func (Encoder) AppendArrayDelim(dst []byte) []byte { + //No delimiters needed in cbor + return dst +} + +// AppendLineBreak is a noop that keep API compat with json encoder. +func (Encoder) AppendLineBreak(dst []byte) []byte { + // No line breaks needed in binary format. + return dst +} + +// AppendBool encodes and inserts a boolean value into the dst byte array. +func (Encoder) AppendBool(dst []byte, val bool) []byte { + b := additionalTypeBoolFalse + if val { + b = additionalTypeBoolTrue + } + return append(dst, byte(majorTypeSimpleAndFloat|b)) +} + +// AppendBools encodes and inserts an array of boolean values into the dst byte array. +func (e Encoder) AppendBools(dst []byte, vals []bool) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendBool(dst, v) + } + return dst +} + +// AppendInt encodes and inserts an integer value into the dst byte array. +func (Encoder) AppendInt(dst []byte, val int) []byte { + major := majorTypeUnsignedInt + contentVal := val + if val < 0 { + major = majorTypeNegativeInt + contentVal = -val - 1 + } + if contentVal <= additionalMax { + lb := byte(contentVal) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(contentVal)) + } + return dst +} + +// AppendInts encodes and inserts an array of integer values into the dst byte array. +func (e Encoder) AppendInts(dst []byte, vals []int) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendInt(dst, v) + } + return dst +} + +// AppendInt8 encodes and inserts an int8 value into the dst byte array. +func (e Encoder) AppendInt8(dst []byte, val int8) []byte { + return e.AppendInt(dst, int(val)) +} + +// AppendInts8 encodes and inserts an array of integer values into the dst byte array. +func (e Encoder) AppendInts8(dst []byte, vals []int8) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendInt(dst, int(v)) + } + return dst +} + +// AppendInt16 encodes and inserts a int16 value into the dst byte array. +func (e Encoder) AppendInt16(dst []byte, val int16) []byte { + return e.AppendInt(dst, int(val)) +} + +// AppendInts16 encodes and inserts an array of int16 values into the dst byte array. +func (e Encoder) AppendInts16(dst []byte, vals []int16) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendInt(dst, int(v)) + } + return dst +} + +// AppendInt32 encodes and inserts a int32 value into the dst byte array. +func (e Encoder) AppendInt32(dst []byte, val int32) []byte { + return e.AppendInt(dst, int(val)) +} + +// AppendInts32 encodes and inserts an array of int32 values into the dst byte array. +func (e Encoder) AppendInts32(dst []byte, vals []int32) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendInt(dst, int(v)) + } + return dst +} + +// AppendInt64 encodes and inserts a int64 value into the dst byte array. +func (Encoder) AppendInt64(dst []byte, val int64) []byte { + major := majorTypeUnsignedInt + contentVal := val + if val < 0 { + major = majorTypeNegativeInt + contentVal = -val - 1 + } + if contentVal <= additionalMax { + lb := byte(contentVal) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(contentVal)) + } + return dst +} + +// AppendInts64 encodes and inserts an array of int64 values into the dst byte array. +func (e Encoder) AppendInts64(dst []byte, vals []int64) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendInt64(dst, v) + } + return dst +} + +// AppendUint encodes and inserts an unsigned integer value into the dst byte array. +func (e Encoder) AppendUint(dst []byte, val uint) []byte { + return e.AppendInt64(dst, int64(val)) +} + +// AppendUints encodes and inserts an array of unsigned integer values into the dst byte array. +func (e Encoder) AppendUints(dst []byte, vals []uint) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendUint(dst, v) + } + return dst +} + +// AppendUint8 encodes and inserts a unsigned int8 value into the dst byte array. +func (e Encoder) AppendUint8(dst []byte, val uint8) []byte { + return e.AppendUint(dst, uint(val)) +} + +// AppendUints8 encodes and inserts an array of uint8 values into the dst byte array. +func (e Encoder) AppendUints8(dst []byte, vals []uint8) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendUint8(dst, v) + } + return dst +} + +// AppendUint16 encodes and inserts a uint16 value into the dst byte array. +func (e Encoder) AppendUint16(dst []byte, val uint16) []byte { + return e.AppendUint(dst, uint(val)) +} + +// AppendUints16 encodes and inserts an array of uint16 values into the dst byte array. +func (e Encoder) AppendUints16(dst []byte, vals []uint16) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendUint16(dst, v) + } + return dst +} + +// AppendUint32 encodes and inserts a uint32 value into the dst byte array. +func (e Encoder) AppendUint32(dst []byte, val uint32) []byte { + return e.AppendUint(dst, uint(val)) +} + +// AppendUints32 encodes and inserts an array of uint32 values into the dst byte array. +func (e Encoder) AppendUints32(dst []byte, vals []uint32) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendUint32(dst, v) + } + return dst +} + +// AppendUint64 encodes and inserts a uint64 value into the dst byte array. +func (Encoder) AppendUint64(dst []byte, val uint64) []byte { + major := majorTypeUnsignedInt + contentVal := val + if contentVal <= additionalMax { + lb := byte(contentVal) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(contentVal)) + } + return dst +} + +// AppendUints64 encodes and inserts an array of uint64 values into the dst byte array. +func (e Encoder) AppendUints64(dst []byte, vals []uint64) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendUint64(dst, v) + } + return dst +} + +// AppendFloat32 encodes and inserts a single precision float value into the dst byte array. +func (Encoder) AppendFloat32(dst []byte, val float32) []byte { + switch { + case math.IsNaN(float64(val)): + return append(dst, "\xfa\x7f\xc0\x00\x00"...) + case math.IsInf(float64(val), 1): + return append(dst, "\xfa\x7f\x80\x00\x00"...) + case math.IsInf(float64(val), -1): + return append(dst, "\xfa\xff\x80\x00\x00"...) + } + major := majorTypeSimpleAndFloat + subType := additionalTypeFloat32 + n := math.Float32bits(val) + var buf [4]byte + for i := uint(0); i < 4; i++ { + buf[i] = byte(n >> ((3 - i) * 8)) + } + return append(append(dst, byte(major|subType)), buf[0], buf[1], buf[2], buf[3]) +} + +// AppendFloats32 encodes and inserts an array of single precision float value into the dst byte array. +func (e Encoder) AppendFloats32(dst []byte, vals []float32) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendFloat32(dst, v) + } + return dst +} + +// AppendFloat64 encodes and inserts a double precision float value into the dst byte array. +func (Encoder) AppendFloat64(dst []byte, val float64) []byte { + switch { + case math.IsNaN(val): + return append(dst, "\xfb\x7f\xf8\x00\x00\x00\x00\x00\x00"...) + case math.IsInf(val, 1): + return append(dst, "\xfb\x7f\xf0\x00\x00\x00\x00\x00\x00"...) + case math.IsInf(val, -1): + return append(dst, "\xfb\xff\xf0\x00\x00\x00\x00\x00\x00"...) + } + major := majorTypeSimpleAndFloat + subType := additionalTypeFloat64 + n := math.Float64bits(val) + dst = append(dst, byte(major|subType)) + for i := uint(1); i <= 8; i++ { + b := byte(n >> ((8 - i) * 8)) + dst = append(dst, b) + } + return dst +} + +// AppendFloats64 encodes and inserts an array of double precision float values into the dst byte array. +func (e Encoder) AppendFloats64(dst []byte, vals []float64) []byte { + major := majorTypeArray + l := len(vals) + if l == 0 { + return e.AppendArrayEnd(e.AppendArrayStart(dst)) + } + if l <= additionalMax { + lb := byte(l) + dst = append(dst, byte(major|lb)) + } else { + dst = appendCborTypePrefix(dst, major, uint64(l)) + } + for _, v := range vals { + dst = e.AppendFloat64(dst, v) + } + return dst +} + +// AppendInterface takes an arbitrary object and converts it to JSON and embeds it dst. +func (e Encoder) AppendInterface(dst []byte, i interface{}) []byte { + marshaled, err := JSONMarshalFunc(i) + if err != nil { + return e.AppendString(dst, fmt.Sprintf("marshaling error: %v", err)) + } + return AppendEmbeddedJSON(dst, marshaled) +} + +// AppendIPAddr encodes and inserts an IP Address (IPv4 or IPv6). +func (e Encoder) AppendIPAddr(dst []byte, ip net.IP) []byte { + dst = append(dst, byte(majorTypeTags|additionalTypeIntUint16)) + dst = append(dst, byte(additionalTypeTagNetworkAddr>>8)) + dst = append(dst, byte(additionalTypeTagNetworkAddr&0xff)) + return e.AppendBytes(dst, ip) +} + +// AppendIPPrefix encodes and inserts an IP Address Prefix (Address + Mask Length). +func (e Encoder) AppendIPPrefix(dst []byte, pfx net.IPNet) []byte { + dst = append(dst, byte(majorTypeTags|additionalTypeIntUint16)) + dst = append(dst, byte(additionalTypeTagNetworkPrefix>>8)) + dst = append(dst, byte(additionalTypeTagNetworkPrefix&0xff)) + + // Prefix is a tuple (aka MAP of 1 pair of elements) - + // first element is prefix, second is mask length. + dst = append(dst, byte(majorTypeMap|0x1)) + dst = e.AppendBytes(dst, pfx.IP) + maskLen, _ := pfx.Mask.Size() + return e.AppendUint8(dst, uint8(maskLen)) +} + +// AppendMACAddr encodes and inserts an Hardware (MAC) address. +func (e Encoder) AppendMACAddr(dst []byte, ha net.HardwareAddr) []byte { + dst = append(dst, byte(majorTypeTags|additionalTypeIntUint16)) + dst = append(dst, byte(additionalTypeTagNetworkAddr>>8)) + dst = append(dst, byte(additionalTypeTagNetworkAddr&0xff)) + return e.AppendBytes(dst, ha) +} + +// AppendHex adds a TAG and inserts a hex bytes as a string. +func (e Encoder) AppendHex(dst []byte, val []byte) []byte { + dst = append(dst, byte(majorTypeTags|additionalTypeIntUint16)) + dst = append(dst, byte(additionalTypeTagHexString>>8)) + dst = append(dst, byte(additionalTypeTagHexString&0xff)) + return e.AppendBytes(dst, val) +} diff --git a/vendor/github.com/rs/zerolog/internal/json/base.go b/vendor/github.com/rs/zerolog/internal/json/base.go new file mode 100644 index 0000000..09ec59f --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/json/base.go @@ -0,0 +1,19 @@ +package json + +// JSONMarshalFunc is used to marshal interface to JSON encoded byte slice. +// Making it package level instead of embedded in Encoder brings +// some extra efforts at importing, but avoids value copy when the functions +// of Encoder being invoked. +// DO REMEMBER to set this variable at importing, or +// you might get a nil pointer dereference panic at runtime. +var JSONMarshalFunc func(v interface{}) ([]byte, error) + +type Encoder struct{} + +// AppendKey appends a new key to the output JSON. +func (e Encoder) AppendKey(dst []byte, key string) []byte { + if dst[len(dst)-1] != '{' { + dst = append(dst, ',') + } + return append(e.AppendString(dst, key), ':') +} diff --git a/vendor/github.com/rs/zerolog/internal/json/bytes.go b/vendor/github.com/rs/zerolog/internal/json/bytes.go new file mode 100644 index 0000000..de64120 --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/json/bytes.go @@ -0,0 +1,85 @@ +package json + +import "unicode/utf8" + +// AppendBytes is a mirror of appendString with []byte arg +func (Encoder) AppendBytes(dst, s []byte) []byte { + dst = append(dst, '"') + for i := 0; i < len(s); i++ { + if !noEscapeTable[s[i]] { + dst = appendBytesComplex(dst, s, i) + return append(dst, '"') + } + } + dst = append(dst, s...) + return append(dst, '"') +} + +// AppendHex encodes the input bytes to a hex string and appends +// the encoded string to the input byte slice. +// +// The operation loops though each byte and encodes it as hex using +// the hex lookup table. +func (Encoder) AppendHex(dst, s []byte) []byte { + dst = append(dst, '"') + for _, v := range s { + dst = append(dst, hex[v>>4], hex[v&0x0f]) + } + return append(dst, '"') +} + +// appendBytesComplex is a mirror of the appendStringComplex +// with []byte arg +func appendBytesComplex(dst, s []byte, i int) []byte { + start := 0 + for i < len(s) { + b := s[i] + if b >= utf8.RuneSelf { + r, size := utf8.DecodeRune(s[i:]) + if r == utf8.RuneError && size == 1 { + if start < i { + dst = append(dst, s[start:i]...) + } + dst = append(dst, `\ufffd`...) + i += size + start = i + continue + } + i += size + continue + } + if noEscapeTable[b] { + i++ + continue + } + // We encountered a character that needs to be encoded. + // Let's append the previous simple characters to the byte slice + // and switch our operation to read and encode the remainder + // characters byte-by-byte. + if start < i { + dst = append(dst, s[start:i]...) + } + switch b { + case '"', '\\': + dst = append(dst, '\\', b) + case '\b': + dst = append(dst, '\\', 'b') + case '\f': + dst = append(dst, '\\', 'f') + case '\n': + dst = append(dst, '\\', 'n') + case '\r': + dst = append(dst, '\\', 'r') + case '\t': + dst = append(dst, '\\', 't') + default: + dst = append(dst, '\\', 'u', '0', '0', hex[b>>4], hex[b&0xF]) + } + i++ + start = i + } + if start < len(s) { + dst = append(dst, s[start:]...) + } + return dst +} diff --git a/vendor/github.com/rs/zerolog/internal/json/string.go b/vendor/github.com/rs/zerolog/internal/json/string.go new file mode 100644 index 0000000..46698b5 --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/json/string.go @@ -0,0 +1,149 @@ +package json + +import ( + "fmt" + "unicode/utf8" +) + +const hex = "0123456789abcdef" + +var noEscapeTable = [256]bool{} + +func init() { + for i := 0; i <= 0x7e; i++ { + noEscapeTable[i] = i >= 0x20 && i != '\\' && i != '"' + } +} + +// AppendStrings encodes the input strings to json and +// appends the encoded string list to the input byte slice. +func (e Encoder) AppendStrings(dst []byte, vals []string) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = e.AppendString(dst, vals[0]) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = e.AppendString(append(dst, ','), val) + } + } + dst = append(dst, ']') + return dst +} + +// AppendString encodes the input string to json and appends +// the encoded string to the input byte slice. +// +// The operation loops though each byte in the string looking +// for characters that need json or utf8 encoding. If the string +// does not need encoding, then the string is appended in it's +// entirety to the byte slice. +// If we encounter a byte that does need encoding, switch up +// the operation and perform a byte-by-byte read-encode-append. +func (Encoder) AppendString(dst []byte, s string) []byte { + // Start with a double quote. + dst = append(dst, '"') + // Loop through each character in the string. + for i := 0; i < len(s); i++ { + // Check if the character needs encoding. Control characters, slashes, + // and the double quote need json encoding. Bytes above the ascii + // boundary needs utf8 encoding. + if !noEscapeTable[s[i]] { + // We encountered a character that needs to be encoded. Switch + // to complex version of the algorithm. + dst = appendStringComplex(dst, s, i) + return append(dst, '"') + } + } + // The string has no need for encoding an therefore is directly + // appended to the byte slice. + dst = append(dst, s...) + // End with a double quote + return append(dst, '"') +} + +// AppendStringers encodes the provided Stringer list to json and +// appends the encoded Stringer list to the input byte slice. +func (e Encoder) AppendStringers(dst []byte, vals []fmt.Stringer) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = e.AppendStringer(dst, vals[0]) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = e.AppendStringer(append(dst, ','), val) + } + } + return append(dst, ']') +} + +// AppendStringer encodes the input Stringer to json and appends the +// encoded Stringer value to the input byte slice. +func (e Encoder) AppendStringer(dst []byte, val fmt.Stringer) []byte { + if val == nil { + return e.AppendInterface(dst, nil) + } + return e.AppendString(dst, val.String()) +} + +//// appendStringComplex is used by appendString to take over an in +// progress JSON string encoding that encountered a character that needs +// to be encoded. +func appendStringComplex(dst []byte, s string, i int) []byte { + start := 0 + for i < len(s) { + b := s[i] + if b >= utf8.RuneSelf { + r, size := utf8.DecodeRuneInString(s[i:]) + if r == utf8.RuneError && size == 1 { + // In case of error, first append previous simple characters to + // the byte slice if any and append a remplacement character code + // in place of the invalid sequence. + if start < i { + dst = append(dst, s[start:i]...) + } + dst = append(dst, `\ufffd`...) + i += size + start = i + continue + } + i += size + continue + } + if noEscapeTable[b] { + i++ + continue + } + // We encountered a character that needs to be encoded. + // Let's append the previous simple characters to the byte slice + // and switch our operation to read and encode the remainder + // characters byte-by-byte. + if start < i { + dst = append(dst, s[start:i]...) + } + switch b { + case '"', '\\': + dst = append(dst, '\\', b) + case '\b': + dst = append(dst, '\\', 'b') + case '\f': + dst = append(dst, '\\', 'f') + case '\n': + dst = append(dst, '\\', 'n') + case '\r': + dst = append(dst, '\\', 'r') + case '\t': + dst = append(dst, '\\', 't') + default: + dst = append(dst, '\\', 'u', '0', '0', hex[b>>4], hex[b&0xF]) + } + i++ + start = i + } + if start < len(s) { + dst = append(dst, s[start:]...) + } + return dst +} diff --git a/vendor/github.com/rs/zerolog/internal/json/time.go b/vendor/github.com/rs/zerolog/internal/json/time.go new file mode 100644 index 0000000..5aff6be --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/json/time.go @@ -0,0 +1,106 @@ +package json + +import ( + "strconv" + "time" +) + +const ( + // Import from zerolog/global.go + timeFormatUnix = "" + timeFormatUnixMs = "UNIXMS" + timeFormatUnixMicro = "UNIXMICRO" +) + +// AppendTime formats the input time with the given format +// and appends the encoded string to the input byte slice. +func (e Encoder) AppendTime(dst []byte, t time.Time, format string) []byte { + switch format { + case timeFormatUnix: + return e.AppendInt64(dst, t.Unix()) + case timeFormatUnixMs: + return e.AppendInt64(dst, t.UnixNano()/1000000) + case timeFormatUnixMicro: + return e.AppendInt64(dst, t.UnixNano()/1000) + } + return append(t.AppendFormat(append(dst, '"'), format), '"') +} + +// AppendTimes converts the input times with the given format +// and appends the encoded string list to the input byte slice. +func (Encoder) AppendTimes(dst []byte, vals []time.Time, format string) []byte { + switch format { + case timeFormatUnix: + return appendUnixTimes(dst, vals) + case timeFormatUnixMs: + return appendUnixMsTimes(dst, vals) + } + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = append(vals[0].AppendFormat(append(dst, '"'), format), '"') + if len(vals) > 1 { + for _, t := range vals[1:] { + dst = append(t.AppendFormat(append(dst, ',', '"'), format), '"') + } + } + dst = append(dst, ']') + return dst +} + +func appendUnixTimes(dst []byte, vals []time.Time) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, vals[0].Unix(), 10) + if len(vals) > 1 { + for _, t := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), t.Unix(), 10) + } + } + dst = append(dst, ']') + return dst +} + +func appendUnixMsTimes(dst []byte, vals []time.Time) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, vals[0].UnixNano()/1000000, 10) + if len(vals) > 1 { + for _, t := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), t.UnixNano()/1000000, 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendDuration formats the input duration with the given unit & format +// and appends the encoded string to the input byte slice. +func (e Encoder) AppendDuration(dst []byte, d time.Duration, unit time.Duration, useInt bool) []byte { + if useInt { + return strconv.AppendInt(dst, int64(d/unit), 10) + } + return e.AppendFloat64(dst, float64(d)/float64(unit)) +} + +// AppendDurations formats the input durations with the given unit & format +// and appends the encoded string list to the input byte slice. +func (e Encoder) AppendDurations(dst []byte, vals []time.Duration, unit time.Duration, useInt bool) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = e.AppendDuration(dst, vals[0], unit, useInt) + if len(vals) > 1 { + for _, d := range vals[1:] { + dst = e.AppendDuration(append(dst, ','), d, unit, useInt) + } + } + dst = append(dst, ']') + return dst +} diff --git a/vendor/github.com/rs/zerolog/internal/json/types.go b/vendor/github.com/rs/zerolog/internal/json/types.go new file mode 100644 index 0000000..9e35266 --- /dev/null +++ b/vendor/github.com/rs/zerolog/internal/json/types.go @@ -0,0 +1,405 @@ +package json + +import ( + "fmt" + "math" + "net" + "strconv" +) + +// AppendNil inserts a 'Nil' object into the dst byte array. +func (Encoder) AppendNil(dst []byte) []byte { + return append(dst, "null"...) +} + +// AppendBeginMarker inserts a map start into the dst byte array. +func (Encoder) AppendBeginMarker(dst []byte) []byte { + return append(dst, '{') +} + +// AppendEndMarker inserts a map end into the dst byte array. +func (Encoder) AppendEndMarker(dst []byte) []byte { + return append(dst, '}') +} + +// AppendLineBreak appends a line break. +func (Encoder) AppendLineBreak(dst []byte) []byte { + return append(dst, '\n') +} + +// AppendArrayStart adds markers to indicate the start of an array. +func (Encoder) AppendArrayStart(dst []byte) []byte { + return append(dst, '[') +} + +// AppendArrayEnd adds markers to indicate the end of an array. +func (Encoder) AppendArrayEnd(dst []byte) []byte { + return append(dst, ']') +} + +// AppendArrayDelim adds markers to indicate end of a particular array element. +func (Encoder) AppendArrayDelim(dst []byte) []byte { + if len(dst) > 0 { + return append(dst, ',') + } + return dst +} + +// AppendBool converts the input bool to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendBool(dst []byte, val bool) []byte { + return strconv.AppendBool(dst, val) +} + +// AppendBools encodes the input bools to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendBools(dst []byte, vals []bool) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendBool(dst, vals[0]) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendBool(append(dst, ','), val) + } + } + dst = append(dst, ']') + return dst +} + +// AppendInt converts the input int to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendInt(dst []byte, val int) []byte { + return strconv.AppendInt(dst, int64(val), 10) +} + +// AppendInts encodes the input ints to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendInts(dst []byte, vals []int) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, int64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), int64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendInt8 converts the input []int8 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendInt8(dst []byte, val int8) []byte { + return strconv.AppendInt(dst, int64(val), 10) +} + +// AppendInts8 encodes the input int8s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendInts8(dst []byte, vals []int8) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, int64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), int64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendInt16 converts the input int16 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendInt16(dst []byte, val int16) []byte { + return strconv.AppendInt(dst, int64(val), 10) +} + +// AppendInts16 encodes the input int16s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendInts16(dst []byte, vals []int16) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, int64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), int64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendInt32 converts the input int32 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendInt32(dst []byte, val int32) []byte { + return strconv.AppendInt(dst, int64(val), 10) +} + +// AppendInts32 encodes the input int32s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendInts32(dst []byte, vals []int32) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, int64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), int64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendInt64 converts the input int64 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendInt64(dst []byte, val int64) []byte { + return strconv.AppendInt(dst, val, 10) +} + +// AppendInts64 encodes the input int64s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendInts64(dst []byte, vals []int64) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendInt(dst, vals[0], 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendInt(append(dst, ','), val, 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendUint converts the input uint to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendUint(dst []byte, val uint) []byte { + return strconv.AppendUint(dst, uint64(val), 10) +} + +// AppendUints encodes the input uints to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendUints(dst []byte, vals []uint) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendUint(dst, uint64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendUint(append(dst, ','), uint64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendUint8 converts the input uint8 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendUint8(dst []byte, val uint8) []byte { + return strconv.AppendUint(dst, uint64(val), 10) +} + +// AppendUints8 encodes the input uint8s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendUints8(dst []byte, vals []uint8) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendUint(dst, uint64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendUint(append(dst, ','), uint64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendUint16 converts the input uint16 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendUint16(dst []byte, val uint16) []byte { + return strconv.AppendUint(dst, uint64(val), 10) +} + +// AppendUints16 encodes the input uint16s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendUints16(dst []byte, vals []uint16) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendUint(dst, uint64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendUint(append(dst, ','), uint64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendUint32 converts the input uint32 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendUint32(dst []byte, val uint32) []byte { + return strconv.AppendUint(dst, uint64(val), 10) +} + +// AppendUints32 encodes the input uint32s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendUints32(dst []byte, vals []uint32) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendUint(dst, uint64(vals[0]), 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendUint(append(dst, ','), uint64(val), 10) + } + } + dst = append(dst, ']') + return dst +} + +// AppendUint64 converts the input uint64 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendUint64(dst []byte, val uint64) []byte { + return strconv.AppendUint(dst, uint64(val), 10) +} + +// AppendUints64 encodes the input uint64s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendUints64(dst []byte, vals []uint64) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = strconv.AppendUint(dst, vals[0], 10) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = strconv.AppendUint(append(dst, ','), val, 10) + } + } + dst = append(dst, ']') + return dst +} + +func appendFloat(dst []byte, val float64, bitSize int) []byte { + // JSON does not permit NaN or Infinity. A typical JSON encoder would fail + // with an error, but a logging library wants the data to get thru so we + // make a tradeoff and store those types as string. + switch { + case math.IsNaN(val): + return append(dst, `"NaN"`...) + case math.IsInf(val, 1): + return append(dst, `"+Inf"`...) + case math.IsInf(val, -1): + return append(dst, `"-Inf"`...) + } + return strconv.AppendFloat(dst, val, 'f', -1, bitSize) +} + +// AppendFloat32 converts the input float32 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendFloat32(dst []byte, val float32) []byte { + return appendFloat(dst, float64(val), 32) +} + +// AppendFloats32 encodes the input float32s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendFloats32(dst []byte, vals []float32) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = appendFloat(dst, float64(vals[0]), 32) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = appendFloat(append(dst, ','), float64(val), 32) + } + } + dst = append(dst, ']') + return dst +} + +// AppendFloat64 converts the input float64 to a string and +// appends the encoded string to the input byte slice. +func (Encoder) AppendFloat64(dst []byte, val float64) []byte { + return appendFloat(dst, val, 64) +} + +// AppendFloats64 encodes the input float64s to json and +// appends the encoded string list to the input byte slice. +func (Encoder) AppendFloats64(dst []byte, vals []float64) []byte { + if len(vals) == 0 { + return append(dst, '[', ']') + } + dst = append(dst, '[') + dst = appendFloat(dst, vals[0], 64) + if len(vals) > 1 { + for _, val := range vals[1:] { + dst = appendFloat(append(dst, ','), val, 64) + } + } + dst = append(dst, ']') + return dst +} + +// AppendInterface marshals the input interface to a string and +// appends the encoded string to the input byte slice. +func (e Encoder) AppendInterface(dst []byte, i interface{}) []byte { + marshaled, err := JSONMarshalFunc(i) + if err != nil { + return e.AppendString(dst, fmt.Sprintf("marshaling error: %v", err)) + } + return append(dst, marshaled...) +} + +// AppendObjectData takes in an object that is already in a byte array +// and adds it to the dst. +func (Encoder) AppendObjectData(dst []byte, o []byte) []byte { + // Three conditions apply here: + // 1. new content starts with '{' - which should be dropped OR + // 2. new content starts with '{' - which should be replaced with ',' + // to separate with existing content OR + // 3. existing content has already other fields + if o[0] == '{' { + if len(dst) > 1 { + dst = append(dst, ',') + } + o = o[1:] + } else if len(dst) > 1 { + dst = append(dst, ',') + } + return append(dst, o...) +} + +// AppendIPAddr adds IPv4 or IPv6 address to dst. +func (e Encoder) AppendIPAddr(dst []byte, ip net.IP) []byte { + return e.AppendString(dst, ip.String()) +} + +// AppendIPPrefix adds IPv4 or IPv6 Prefix (address & mask) to dst. +func (e Encoder) AppendIPPrefix(dst []byte, pfx net.IPNet) []byte { + return e.AppendString(dst, pfx.String()) + +} + +// AppendMACAddr adds MAC address to dst. +func (e Encoder) AppendMACAddr(dst []byte, ha net.HardwareAddr) []byte { + return e.AppendString(dst, ha.String()) +} diff --git a/vendor/github.com/rs/zerolog/log.go b/vendor/github.com/rs/zerolog/log.go new file mode 100644 index 0000000..6227386 --- /dev/null +++ b/vendor/github.com/rs/zerolog/log.go @@ -0,0 +1,457 @@ +// Package zerolog provides a lightweight logging library dedicated to JSON logging. +// +// A global Logger can be use for simple logging: +// +// import "github.com/rs/zerolog/log" +// +// log.Info().Msg("hello world") +// // Output: {"time":1494567715,"level":"info","message":"hello world"} +// +// NOTE: To import the global logger, import the "log" subpackage "github.com/rs/zerolog/log". +// +// Fields can be added to log messages: +// +// log.Info().Str("foo", "bar").Msg("hello world") +// // Output: {"time":1494567715,"level":"info","message":"hello world","foo":"bar"} +// +// Create logger instance to manage different outputs: +// +// logger := zerolog.New(os.Stderr).With().Timestamp().Logger() +// logger.Info(). +// Str("foo", "bar"). +// Msg("hello world") +// // Output: {"time":1494567715,"level":"info","message":"hello world","foo":"bar"} +// +// Sub-loggers let you chain loggers with additional context: +// +// sublogger := log.With().Str("component": "foo").Logger() +// sublogger.Info().Msg("hello world") +// // Output: {"time":1494567715,"level":"info","message":"hello world","component":"foo"} +// +// Level logging +// +// zerolog.SetGlobalLevel(zerolog.InfoLevel) +// +// log.Debug().Msg("filtered out message") +// log.Info().Msg("routed message") +// +// if e := log.Debug(); e.Enabled() { +// // Compute log output only if enabled. +// value := compute() +// e.Str("foo": value).Msg("some debug message") +// } +// // Output: {"level":"info","time":1494567715,"routed message"} +// +// Customize automatic field names: +// +// log.TimestampFieldName = "t" +// log.LevelFieldName = "p" +// log.MessageFieldName = "m" +// +// log.Info().Msg("hello world") +// // Output: {"t":1494567715,"p":"info","m":"hello world"} +// +// Log with no level and message: +// +// log.Log().Str("foo","bar").Msg("") +// // Output: {"time":1494567715,"foo":"bar"} +// +// Add contextual fields to global Logger: +// +// log.Logger = log.With().Str("foo", "bar").Logger() +// +// Sample logs: +// +// sampled := log.Sample(&zerolog.BasicSampler{N: 10}) +// sampled.Info().Msg("will be logged every 10 messages") +// +// Log with contextual hooks: +// +// // Create the hook: +// type SeverityHook struct{} +// +// func (h SeverityHook) Run(e *zerolog.Event, level zerolog.Level, msg string) { +// if level != zerolog.NoLevel { +// e.Str("severity", level.String()) +// } +// } +// +// // And use it: +// var h SeverityHook +// log := zerolog.New(os.Stdout).Hook(h) +// log.Warn().Msg("") +// // Output: {"level":"warn","severity":"warn"} +// +// +// Caveats +// +// There is no fields deduplication out-of-the-box. +// Using the same key multiple times creates new key in final JSON each time. +// +// logger := zerolog.New(os.Stderr).With().Timestamp().Logger() +// logger.Info(). +// Timestamp(). +// Msg("dup") +// // Output: {"level":"info","time":1494567715,"time":1494567715,"message":"dup"} +// +// In this case, many consumers will take the last value, +// but this is not guaranteed; check yours if in doubt. +package zerolog + +import ( + "fmt" + "io" + "io/ioutil" + "os" + "strconv" +) + +// Level defines log levels. +type Level int8 + +const ( + // DebugLevel defines debug log level. + DebugLevel Level = iota + // InfoLevel defines info log level. + InfoLevel + // WarnLevel defines warn log level. + WarnLevel + // ErrorLevel defines error log level. + ErrorLevel + // FatalLevel defines fatal log level. + FatalLevel + // PanicLevel defines panic log level. + PanicLevel + // NoLevel defines an absent log level. + NoLevel + // Disabled disables the logger. + Disabled + + // TraceLevel defines trace log level. + TraceLevel Level = -1 + // Values less than TraceLevel are handled as numbers. +) + +func (l Level) String() string { + switch l { + case TraceLevel: + return LevelTraceValue + case DebugLevel: + return LevelDebugValue + case InfoLevel: + return LevelInfoValue + case WarnLevel: + return LevelWarnValue + case ErrorLevel: + return LevelErrorValue + case FatalLevel: + return LevelFatalValue + case PanicLevel: + return LevelPanicValue + case Disabled: + return "disabled" + case NoLevel: + return "" + } + return strconv.Itoa(int(l)) +} + +// ParseLevel converts a level string into a zerolog Level value. +// returns an error if the input string does not match known values. +func ParseLevel(levelStr string) (Level, error) { + switch levelStr { + case LevelFieldMarshalFunc(TraceLevel): + return TraceLevel, nil + case LevelFieldMarshalFunc(DebugLevel): + return DebugLevel, nil + case LevelFieldMarshalFunc(InfoLevel): + return InfoLevel, nil + case LevelFieldMarshalFunc(WarnLevel): + return WarnLevel, nil + case LevelFieldMarshalFunc(ErrorLevel): + return ErrorLevel, nil + case LevelFieldMarshalFunc(FatalLevel): + return FatalLevel, nil + case LevelFieldMarshalFunc(PanicLevel): + return PanicLevel, nil + case LevelFieldMarshalFunc(Disabled): + return Disabled, nil + case LevelFieldMarshalFunc(NoLevel): + return NoLevel, nil + } + i, err := strconv.Atoi(levelStr) + if err != nil { + return NoLevel, fmt.Errorf("Unknown Level String: '%s', defaulting to NoLevel", levelStr) + } + if i > 127 || i < -128 { + return NoLevel, fmt.Errorf("Out-Of-Bounds Level: '%d', defaulting to NoLevel", i) + } + return Level(i), nil +} + +// A Logger represents an active logging object that generates lines +// of JSON output to an io.Writer. Each logging operation makes a single +// call to the Writer's Write method. There is no guarantee on access +// serialization to the Writer. If your Writer is not thread safe, +// you may consider a sync wrapper. +type Logger struct { + w LevelWriter + level Level + sampler Sampler + context []byte + hooks []Hook + stack bool +} + +// New creates a root logger with given output writer. If the output writer implements +// the LevelWriter interface, the WriteLevel method will be called instead of the Write +// one. +// +// Each logging operation makes a single call to the Writer's Write method. There is no +// guarantee on access serialization to the Writer. If your Writer is not thread safe, +// you may consider using sync wrapper. +func New(w io.Writer) Logger { + if w == nil { + w = ioutil.Discard + } + lw, ok := w.(LevelWriter) + if !ok { + lw = levelWriterAdapter{w} + } + return Logger{w: lw, level: TraceLevel} +} + +// Nop returns a disabled logger for which all operation are no-op. +func Nop() Logger { + return New(nil).Level(Disabled) +} + +// Output duplicates the current logger and sets w as its output. +func (l Logger) Output(w io.Writer) Logger { + l2 := New(w) + l2.level = l.level + l2.sampler = l.sampler + l2.stack = l.stack + if len(l.hooks) > 0 { + l2.hooks = append(l2.hooks, l.hooks...) + } + if l.context != nil { + l2.context = make([]byte, len(l.context), cap(l.context)) + copy(l2.context, l.context) + } + return l2 +} + +// With creates a child logger with the field added to its context. +func (l Logger) With() Context { + context := l.context + l.context = make([]byte, 0, 500) + if context != nil { + l.context = append(l.context, context...) + } else { + // This is needed for AppendKey to not check len of input + // thus making it inlinable + l.context = enc.AppendBeginMarker(l.context) + } + return Context{l} +} + +// UpdateContext updates the internal logger's context. +// +// Use this method with caution. If unsure, prefer the With method. +func (l *Logger) UpdateContext(update func(c Context) Context) { + if l == disabledLogger { + return + } + if cap(l.context) == 0 { + l.context = make([]byte, 0, 500) + } + if len(l.context) == 0 { + l.context = enc.AppendBeginMarker(l.context) + } + c := update(Context{*l}) + l.context = c.l.context +} + +// Level creates a child logger with the minimum accepted level set to level. +func (l Logger) Level(lvl Level) Logger { + l.level = lvl + return l +} + +// GetLevel returns the current Level of l. +func (l Logger) GetLevel() Level { + return l.level +} + +// Sample returns a logger with the s sampler. +func (l Logger) Sample(s Sampler) Logger { + l.sampler = s + return l +} + +// Hook returns a logger with the h Hook. +func (l Logger) Hook(h Hook) Logger { + l.hooks = append(l.hooks, h) + return l +} + +// Trace starts a new message with trace level. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Trace() *Event { + return l.newEvent(TraceLevel, nil) +} + +// Debug starts a new message with debug level. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Debug() *Event { + return l.newEvent(DebugLevel, nil) +} + +// Info starts a new message with info level. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Info() *Event { + return l.newEvent(InfoLevel, nil) +} + +// Warn starts a new message with warn level. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Warn() *Event { + return l.newEvent(WarnLevel, nil) +} + +// Error starts a new message with error level. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Error() *Event { + return l.newEvent(ErrorLevel, nil) +} + +// Err starts a new message with error level with err as a field if not nil or +// with info level if err is nil. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Err(err error) *Event { + if err != nil { + return l.Error().Err(err) + } + + return l.Info() +} + +// Fatal starts a new message with fatal level. The os.Exit(1) function +// is called by the Msg method, which terminates the program immediately. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Fatal() *Event { + return l.newEvent(FatalLevel, func(msg string) { os.Exit(1) }) +} + +// Panic starts a new message with panic level. The panic() function +// is called by the Msg method, which stops the ordinary flow of a goroutine. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Panic() *Event { + return l.newEvent(PanicLevel, func(msg string) { panic(msg) }) +} + +// WithLevel starts a new message with level. Unlike Fatal and Panic +// methods, WithLevel does not terminate the program or stop the ordinary +// flow of a gourotine when used with their respective levels. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) WithLevel(level Level) *Event { + switch level { + case TraceLevel: + return l.Trace() + case DebugLevel: + return l.Debug() + case InfoLevel: + return l.Info() + case WarnLevel: + return l.Warn() + case ErrorLevel: + return l.Error() + case FatalLevel: + return l.newEvent(FatalLevel, nil) + case PanicLevel: + return l.newEvent(PanicLevel, nil) + case NoLevel: + return l.Log() + case Disabled: + return nil + default: + return l.newEvent(level, nil) + } +} + +// Log starts a new message with no level. Setting GlobalLevel to Disabled +// will still disable events produced by this method. +// +// You must call Msg on the returned event in order to send the event. +func (l *Logger) Log() *Event { + return l.newEvent(NoLevel, nil) +} + +// Print sends a log event using debug level and no extra field. +// Arguments are handled in the manner of fmt.Print. +func (l *Logger) Print(v ...interface{}) { + if e := l.Debug(); e.Enabled() { + e.CallerSkipFrame(1).Msg(fmt.Sprint(v...)) + } +} + +// Printf sends a log event using debug level and no extra field. +// Arguments are handled in the manner of fmt.Printf. +func (l *Logger) Printf(format string, v ...interface{}) { + if e := l.Debug(); e.Enabled() { + e.CallerSkipFrame(1).Msg(fmt.Sprintf(format, v...)) + } +} + +// Write implements the io.Writer interface. This is useful to set as a writer +// for the standard library log. +func (l Logger) Write(p []byte) (n int, err error) { + n = len(p) + if n > 0 && p[n-1] == '\n' { + // Trim CR added by stdlog. + p = p[0 : n-1] + } + l.Log().CallerSkipFrame(1).Msg(string(p)) + return +} + +func (l *Logger) newEvent(level Level, done func(string)) *Event { + enabled := l.should(level) + if !enabled { + return nil + } + e := newEvent(l.w, level) + e.done = done + e.ch = l.hooks + if level != NoLevel && LevelFieldName != "" { + e.Str(LevelFieldName, LevelFieldMarshalFunc(level)) + } + if l.context != nil && len(l.context) > 1 { + e.buf = enc.AppendObjectData(e.buf, l.context) + } + if l.stack { + e.Stack() + } + return e +} + +// should returns true if the log event should be logged. +func (l *Logger) should(lvl Level) bool { + if lvl < l.level || lvl < GlobalLevel() { + return false + } + if l.sampler != nil && !samplingDisabled() { + return l.sampler.Sample(lvl) + } + return true +} diff --git a/vendor/github.com/rs/zerolog/log/log.go b/vendor/github.com/rs/zerolog/log/log.go new file mode 100644 index 0000000..a96ec50 --- /dev/null +++ b/vendor/github.com/rs/zerolog/log/log.go @@ -0,0 +1,131 @@ +// Package log provides a global logger for zerolog. +package log + +import ( + "context" + "fmt" + "io" + "os" + + "github.com/rs/zerolog" +) + +// Logger is the global logger. +var Logger = zerolog.New(os.Stderr).With().Timestamp().Logger() + +// Output duplicates the global logger and sets w as its output. +func Output(w io.Writer) zerolog.Logger { + return Logger.Output(w) +} + +// With creates a child logger with the field added to its context. +func With() zerolog.Context { + return Logger.With() +} + +// Level creates a child logger with the minimum accepted level set to level. +func Level(level zerolog.Level) zerolog.Logger { + return Logger.Level(level) +} + +// Sample returns a logger with the s sampler. +func Sample(s zerolog.Sampler) zerolog.Logger { + return Logger.Sample(s) +} + +// Hook returns a logger with the h Hook. +func Hook(h zerolog.Hook) zerolog.Logger { + return Logger.Hook(h) +} + +// Err starts a new message with error level with err as a field if not nil or +// with info level if err is nil. +// +// You must call Msg on the returned event in order to send the event. +func Err(err error) *zerolog.Event { + return Logger.Err(err) +} + +// Trace starts a new message with trace level. +// +// You must call Msg on the returned event in order to send the event. +func Trace() *zerolog.Event { + return Logger.Trace() +} + +// Debug starts a new message with debug level. +// +// You must call Msg on the returned event in order to send the event. +func Debug() *zerolog.Event { + return Logger.Debug() +} + +// Info starts a new message with info level. +// +// You must call Msg on the returned event in order to send the event. +func Info() *zerolog.Event { + return Logger.Info() +} + +// Warn starts a new message with warn level. +// +// You must call Msg on the returned event in order to send the event. +func Warn() *zerolog.Event { + return Logger.Warn() +} + +// Error starts a new message with error level. +// +// You must call Msg on the returned event in order to send the event. +func Error() *zerolog.Event { + return Logger.Error() +} + +// Fatal starts a new message with fatal level. The os.Exit(1) function +// is called by the Msg method. +// +// You must call Msg on the returned event in order to send the event. +func Fatal() *zerolog.Event { + return Logger.Fatal() +} + +// Panic starts a new message with panic level. The message is also sent +// to the panic function. +// +// You must call Msg on the returned event in order to send the event. +func Panic() *zerolog.Event { + return Logger.Panic() +} + +// WithLevel starts a new message with level. +// +// You must call Msg on the returned event in order to send the event. +func WithLevel(level zerolog.Level) *zerolog.Event { + return Logger.WithLevel(level) +} + +// Log starts a new message with no level. Setting zerolog.GlobalLevel to +// zerolog.Disabled will still disable events produced by this method. +// +// You must call Msg on the returned event in order to send the event. +func Log() *zerolog.Event { + return Logger.Log() +} + +// Print sends a log event using debug level and no extra field. +// Arguments are handled in the manner of fmt.Print. +func Print(v ...interface{}) { + Logger.Debug().CallerSkipFrame(1).Msg(fmt.Sprint(v...)) +} + +// Printf sends a log event using debug level and no extra field. +// Arguments are handled in the manner of fmt.Printf. +func Printf(format string, v ...interface{}) { + Logger.Debug().CallerSkipFrame(1).Msgf(format, v...) +} + +// Ctx returns the Logger associated with the ctx. If no logger +// is associated, a disabled logger is returned. +func Ctx(ctx context.Context) *zerolog.Logger { + return zerolog.Ctx(ctx) +} diff --git a/vendor/github.com/rs/zerolog/not_go112.go b/vendor/github.com/rs/zerolog/not_go112.go new file mode 100644 index 0000000..4c43c9e --- /dev/null +++ b/vendor/github.com/rs/zerolog/not_go112.go @@ -0,0 +1,5 @@ +// +build !go1.12 + +package zerolog + +const contextCallerSkipFrameCount = 3 diff --git a/vendor/github.com/rs/zerolog/pretty.png b/vendor/github.com/rs/zerolog/pretty.png new file mode 100644 index 0000000..2420336 Binary files /dev/null and b/vendor/github.com/rs/zerolog/pretty.png differ diff --git a/vendor/github.com/rs/zerolog/sampler.go b/vendor/github.com/rs/zerolog/sampler.go new file mode 100644 index 0000000..1be98c4 --- /dev/null +++ b/vendor/github.com/rs/zerolog/sampler.go @@ -0,0 +1,134 @@ +package zerolog + +import ( + "math/rand" + "sync/atomic" + "time" +) + +var ( + // Often samples log every ~ 10 events. + Often = RandomSampler(10) + // Sometimes samples log every ~ 100 events. + Sometimes = RandomSampler(100) + // Rarely samples log every ~ 1000 events. + Rarely = RandomSampler(1000) +) + +// Sampler defines an interface to a log sampler. +type Sampler interface { + // Sample returns true if the event should be part of the sample, false if + // the event should be dropped. + Sample(lvl Level) bool +} + +// RandomSampler use a PRNG to randomly sample an event out of N events, +// regardless of their level. +type RandomSampler uint32 + +// Sample implements the Sampler interface. +func (s RandomSampler) Sample(lvl Level) bool { + if s <= 0 { + return false + } + if rand.Intn(int(s)) != 0 { + return false + } + return true +} + +// BasicSampler is a sampler that will send every Nth events, regardless of +// their level. +type BasicSampler struct { + N uint32 + counter uint32 +} + +// Sample implements the Sampler interface. +func (s *BasicSampler) Sample(lvl Level) bool { + n := s.N + if n == 1 { + return true + } + c := atomic.AddUint32(&s.counter, 1) + return c%n == 1 +} + +// BurstSampler lets Burst events pass per Period then pass the decision to +// NextSampler. If Sampler is not set, all subsequent events are rejected. +type BurstSampler struct { + // Burst is the maximum number of event per period allowed before calling + // NextSampler. + Burst uint32 + // Period defines the burst period. If 0, NextSampler is always called. + Period time.Duration + // NextSampler is the sampler used after the burst is reached. If nil, + // events are always rejected after the burst. + NextSampler Sampler + + counter uint32 + resetAt int64 +} + +// Sample implements the Sampler interface. +func (s *BurstSampler) Sample(lvl Level) bool { + if s.Burst > 0 && s.Period > 0 { + if s.inc() <= s.Burst { + return true + } + } + if s.NextSampler == nil { + return false + } + return s.NextSampler.Sample(lvl) +} + +func (s *BurstSampler) inc() uint32 { + now := time.Now().UnixNano() + resetAt := atomic.LoadInt64(&s.resetAt) + var c uint32 + if now > resetAt { + c = 1 + atomic.StoreUint32(&s.counter, c) + newResetAt := now + s.Period.Nanoseconds() + reset := atomic.CompareAndSwapInt64(&s.resetAt, resetAt, newResetAt) + if !reset { + // Lost the race with another goroutine trying to reset. + c = atomic.AddUint32(&s.counter, 1) + } + } else { + c = atomic.AddUint32(&s.counter, 1) + } + return c +} + +// LevelSampler applies a different sampler for each level. +type LevelSampler struct { + TraceSampler, DebugSampler, InfoSampler, WarnSampler, ErrorSampler Sampler +} + +func (s LevelSampler) Sample(lvl Level) bool { + switch lvl { + case TraceLevel: + if s.TraceSampler != nil { + return s.TraceSampler.Sample(lvl) + } + case DebugLevel: + if s.DebugSampler != nil { + return s.DebugSampler.Sample(lvl) + } + case InfoLevel: + if s.InfoSampler != nil { + return s.InfoSampler.Sample(lvl) + } + case WarnLevel: + if s.WarnSampler != nil { + return s.WarnSampler.Sample(lvl) + } + case ErrorLevel: + if s.ErrorSampler != nil { + return s.ErrorSampler.Sample(lvl) + } + } + return true +} diff --git a/vendor/github.com/rs/zerolog/syslog.go b/vendor/github.com/rs/zerolog/syslog.go new file mode 100644 index 0000000..c408283 --- /dev/null +++ b/vendor/github.com/rs/zerolog/syslog.go @@ -0,0 +1,80 @@ +// +build !windows +// +build !binary_log + +package zerolog + +import ( + "io" +) + +// See http://cee.mitre.org/language/1.0-beta1/clt.html#syslog +// or https://www.rsyslog.com/json-elasticsearch/ +const ceePrefix = "@cee:" + +// SyslogWriter is an interface matching a syslog.Writer struct. +type SyslogWriter interface { + io.Writer + Debug(m string) error + Info(m string) error + Warning(m string) error + Err(m string) error + Emerg(m string) error + Crit(m string) error +} + +type syslogWriter struct { + w SyslogWriter + prefix string +} + +// SyslogLevelWriter wraps a SyslogWriter and call the right syslog level +// method matching the zerolog level. +func SyslogLevelWriter(w SyslogWriter) LevelWriter { + return syslogWriter{w, ""} +} + +// SyslogCEEWriter wraps a SyslogWriter with a SyslogLevelWriter that adds a +// MITRE CEE prefix for JSON syslog entries, compatible with rsyslog +// and syslog-ng JSON logging support. +// See https://www.rsyslog.com/json-elasticsearch/ +func SyslogCEEWriter(w SyslogWriter) LevelWriter { + return syslogWriter{w, ceePrefix} +} + +func (sw syslogWriter) Write(p []byte) (n int, err error) { + var pn int + if sw.prefix != "" { + pn, err = sw.w.Write([]byte(sw.prefix)) + if err != nil { + return pn, err + } + } + n, err = sw.w.Write(p) + return pn + n, err +} + +// WriteLevel implements LevelWriter interface. +func (sw syslogWriter) WriteLevel(level Level, p []byte) (n int, err error) { + switch level { + case TraceLevel: + case DebugLevel: + err = sw.w.Debug(sw.prefix + string(p)) + case InfoLevel: + err = sw.w.Info(sw.prefix + string(p)) + case WarnLevel: + err = sw.w.Warning(sw.prefix + string(p)) + case ErrorLevel: + err = sw.w.Err(sw.prefix + string(p)) + case FatalLevel: + err = sw.w.Emerg(sw.prefix + string(p)) + case PanicLevel: + err = sw.w.Crit(sw.prefix + string(p)) + case NoLevel: + err = sw.w.Info(sw.prefix + string(p)) + default: + panic("invalid level") + } + // Any CEE prefix is not part of the message, so we don't include its length + n = len(p) + return +} diff --git a/vendor/github.com/rs/zerolog/writer.go b/vendor/github.com/rs/zerolog/writer.go new file mode 100644 index 0000000..26f5e63 --- /dev/null +++ b/vendor/github.com/rs/zerolog/writer.go @@ -0,0 +1,154 @@ +package zerolog + +import ( + "bytes" + "io" + "path" + "runtime" + "strconv" + "strings" + "sync" +) + +// LevelWriter defines as interface a writer may implement in order +// to receive level information with payload. +type LevelWriter interface { + io.Writer + WriteLevel(level Level, p []byte) (n int, err error) +} + +type levelWriterAdapter struct { + io.Writer +} + +func (lw levelWriterAdapter) WriteLevel(l Level, p []byte) (n int, err error) { + return lw.Write(p) +} + +type syncWriter struct { + mu sync.Mutex + lw LevelWriter +} + +// SyncWriter wraps w so that each call to Write is synchronized with a mutex. +// This syncer can be used to wrap the call to writer's Write method if it is +// not thread safe. Note that you do not need this wrapper for os.File Write +// operations on POSIX and Windows systems as they are already thread-safe. +func SyncWriter(w io.Writer) io.Writer { + if lw, ok := w.(LevelWriter); ok { + return &syncWriter{lw: lw} + } + return &syncWriter{lw: levelWriterAdapter{w}} +} + +// Write implements the io.Writer interface. +func (s *syncWriter) Write(p []byte) (n int, err error) { + s.mu.Lock() + defer s.mu.Unlock() + return s.lw.Write(p) +} + +// WriteLevel implements the LevelWriter interface. +func (s *syncWriter) WriteLevel(l Level, p []byte) (n int, err error) { + s.mu.Lock() + defer s.mu.Unlock() + return s.lw.WriteLevel(l, p) +} + +type multiLevelWriter struct { + writers []LevelWriter +} + +func (t multiLevelWriter) Write(p []byte) (n int, err error) { + for _, w := range t.writers { + if _n, _err := w.Write(p); err == nil { + n = _n + if _err != nil { + err = _err + } else if _n != len(p) { + err = io.ErrShortWrite + } + } + } + return n, err +} + +func (t multiLevelWriter) WriteLevel(l Level, p []byte) (n int, err error) { + for _, w := range t.writers { + if _n, _err := w.WriteLevel(l, p); err == nil { + n = _n + if _err != nil { + err = _err + } else if _n != len(p) { + err = io.ErrShortWrite + } + } + } + return n, err +} + +// MultiLevelWriter creates a writer that duplicates its writes to all the +// provided writers, similar to the Unix tee(1) command. If some writers +// implement LevelWriter, their WriteLevel method will be used instead of Write. +func MultiLevelWriter(writers ...io.Writer) LevelWriter { + lwriters := make([]LevelWriter, 0, len(writers)) + for _, w := range writers { + if lw, ok := w.(LevelWriter); ok { + lwriters = append(lwriters, lw) + } else { + lwriters = append(lwriters, levelWriterAdapter{w}) + } + } + return multiLevelWriter{lwriters} +} + +// TestingLog is the logging interface of testing.TB. +type TestingLog interface { + Log(args ...interface{}) + Logf(format string, args ...interface{}) + Helper() +} + +// TestWriter is a writer that writes to testing.TB. +type TestWriter struct { + T TestingLog + + // Frame skips caller frames to capture the original file and line numbers. + Frame int +} + +// NewTestWriter creates a writer that logs to the testing.TB. +func NewTestWriter(t TestingLog) TestWriter { + return TestWriter{T: t} +} + +// Write to testing.TB. +func (t TestWriter) Write(p []byte) (n int, err error) { + t.T.Helper() + + n = len(p) + + // Strip trailing newline because t.Log always adds one. + p = bytes.TrimRight(p, "\n") + + // Try to correct the log file and line number to the caller. + if t.Frame > 0 { + _, origFile, origLine, _ := runtime.Caller(1) + _, frameFile, frameLine, ok := runtime.Caller(1 + t.Frame) + if ok { + erase := strings.Repeat("\b", len(path.Base(origFile))+len(strconv.Itoa(origLine))+3) + t.T.Logf("%s%s:%d: %s", erase, path.Base(frameFile), frameLine, p) + return n, err + } + } + t.T.Log(string(p)) + + return n, err +} + +// ConsoleTestWriter creates an option that correctly sets the file frame depth for testing.TB log. +func ConsoleTestWriter(t TestingLog) func(w *ConsoleWriter) { + return func(w *ConsoleWriter) { + w.Out = TestWriter{T: t, Frame: 6} + } +} diff --git a/vendor/github.com/xdg-go/pbkdf2/.gitignore b/vendor/github.com/xdg-go/pbkdf2/.gitignore new file mode 100644 index 0000000..f1c181e --- /dev/null +++ b/vendor/github.com/xdg-go/pbkdf2/.gitignore @@ -0,0 +1,12 @@ +# Binaries for programs and plugins +*.exe +*.exe~ +*.dll +*.so +*.dylib + +# Test binary, build with `go test -c` +*.test + +# Output of the go coverage tool, specifically when used with LiteIDE +*.out diff --git a/vendor/github.com/xdg-go/pbkdf2/LICENSE b/vendor/github.com/xdg-go/pbkdf2/LICENSE new file mode 100644 index 0000000..67db858 --- /dev/null +++ b/vendor/github.com/xdg-go/pbkdf2/LICENSE @@ -0,0 +1,175 @@ + + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. diff --git a/vendor/github.com/xdg-go/pbkdf2/README.md b/vendor/github.com/xdg-go/pbkdf2/README.md new file mode 100644 index 0000000..d2824e4 --- /dev/null +++ b/vendor/github.com/xdg-go/pbkdf2/README.md @@ -0,0 +1,17 @@ +[![Go Reference](https://pkg.go.dev/badge/github.com/xdg-go/pbkdf2.svg)](https://pkg.go.dev/github.com/xdg-go/pbkdf2) +[![Go Report Card](https://goreportcard.com/badge/github.com/xdg-go/pbkdf2)](https://goreportcard.com/report/github.com/xdg-go/pbkdf2) +[![Github Actions](https://github.com/xdg-go/pbkdf2/actions/workflows/test.yml/badge.svg)](https://github.com/xdg-go/pbkdf2/actions/workflows/test.yml) + +# pbkdf2 – Go implementation of PBKDF2 + +## Description + +Package pbkdf2 provides password-based key derivation based on +[RFC 8018](https://tools.ietf.org/html/rfc8018). + +## Copyright and License + +Copyright 2021 by David A. Golden. All rights reserved. + +Licensed under the Apache License, Version 2.0 (the "License"). You may +obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 diff --git a/vendor/github.com/xdg-go/pbkdf2/pbkdf2.go b/vendor/github.com/xdg-go/pbkdf2/pbkdf2.go new file mode 100644 index 0000000..029945c --- /dev/null +++ b/vendor/github.com/xdg-go/pbkdf2/pbkdf2.go @@ -0,0 +1,76 @@ +// Copyright 2021 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +// Package pbkdf2 implements password-based key derivation using the PBKDF2 +// algorithm described in RFC 2898 and RFC 8018. +// +// It provides a drop-in replacement for `golang.org/x/crypto/pbkdf2`, with +// the following benefits: +// +// - Released as a module with semantic versioning +// +// - Does not pull in dependencies for unrelated `x/crypto/*` packages +// +// - Supports Go 1.9+ +// +// See https://tools.ietf.org/html/rfc8018#section-4 for security considerations +// in the selection of a salt and iteration count. +package pbkdf2 + +import ( + "crypto/hmac" + "encoding/binary" + "hash" +) + +// Key generates a derived key from a password using the PBKDF2 algorithm. The +// inputs include salt bytes, the iteration count, desired key length, and a +// constructor for a hashing function. For example, for a 32-byte key using +// SHA-256: +// +// key := Key([]byte("trustNo1"), salt, 10000, 32, sha256.New) +func Key(password, salt []byte, iterCount, keyLen int, h func() hash.Hash) []byte { + prf := hmac.New(h, password) + hLen := prf.Size() + numBlocks := keyLen / hLen + // Get an extra block if keyLen is not an even number of hLen blocks. + if keyLen%hLen > 0 { + numBlocks++ + } + + Ti := make([]byte, hLen) + Uj := make([]byte, hLen) + dk := make([]byte, 0, hLen*numBlocks) + buf := make([]byte, 4) + + for i := uint32(1); i <= uint32(numBlocks); i++ { + // Initialize Uj for j == 1 from salt and block index. + // Initialize Ti = U1. + binary.BigEndian.PutUint32(buf, i) + prf.Reset() + prf.Write(salt) + prf.Write(buf) + Uj = Uj[:0] + Uj = prf.Sum(Uj) + + // Ti = U1 ^ U2 ^ ... ^ Ux + copy(Ti, Uj) + for j := 2; j <= iterCount; j++ { + prf.Reset() + prf.Write(Uj) + Uj = Uj[:0] + Uj = prf.Sum(Uj) + for k := range Uj { + Ti[k] ^= Uj[k] + } + } + + // DK = concat(T1, T2, ... Tn) + dk = append(dk, Ti...) + } + + return dk[0:keyLen] +} diff --git a/vendor/github.com/xdg-go/scram/.gitignore b/vendor/github.com/xdg-go/scram/.gitignore new file mode 100644 index 0000000..e69de29 diff --git a/vendor/github.com/xdg-go/scram/CHANGELOG.md b/vendor/github.com/xdg-go/scram/CHANGELOG.md new file mode 100644 index 0000000..21828db --- /dev/null +++ b/vendor/github.com/xdg-go/scram/CHANGELOG.md @@ -0,0 +1,22 @@ +# CHANGELOG + +## v1.1.1 - 2022-03-03 + +- Bump stringprep dependency to v1.0.3 for upstream CVE fix. + +## v1.1.0 - 2022-01-16 + +- Add SHA-512 hash generator function for convenience. + +## v1.0.2 - 2021-03-28 + +- Switch PBKDF2 dependency to github.com/xdg-go/pbkdf2 to + minimize transitive dependencies and support Go 1.9+. + +## v1.0.1 - 2021-03-27 + +- Bump stringprep dependency to v1.0.2 for Go 1.11 support. + +## v1.0.0 - 2021-03-27 + +- First release as a Go module diff --git a/vendor/github.com/xdg-go/scram/LICENSE b/vendor/github.com/xdg-go/scram/LICENSE new file mode 100644 index 0000000..67db858 --- /dev/null +++ b/vendor/github.com/xdg-go/scram/LICENSE @@ -0,0 +1,175 @@ + + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. diff --git a/vendor/github.com/xdg-go/scram/README.md b/vendor/github.com/xdg-go/scram/README.md new file mode 100644 index 0000000..3a46f5c --- /dev/null +++ b/vendor/github.com/xdg-go/scram/README.md @@ -0,0 +1,72 @@ +[![Go Reference](https://pkg.go.dev/badge/github.com/xdg-go/scram.svg)](https://pkg.go.dev/github.com/xdg-go/scram) +[![Go Report Card](https://goreportcard.com/badge/github.com/xdg-go/scram)](https://goreportcard.com/report/github.com/xdg-go/scram) +[![Github Actions](https://github.com/xdg-go/scram/actions/workflows/test.yml/badge.svg)](https://github.com/xdg-go/scram/actions/workflows/test.yml) + +# scram – Go implementation of RFC-5802 + +## Description + +Package scram provides client and server implementations of the Salted +Challenge Response Authentication Mechanism (SCRAM) described in +[RFC-5802](https://tools.ietf.org/html/rfc5802) and +[RFC-7677](https://tools.ietf.org/html/rfc7677). + +It includes both client and server side support. + +Channel binding and extensions are not (yet) supported. + +## Examples + +### Client side + + package main + + import "github.com/xdg-go/scram" + + func main() { + // Get Client with username, password and (optional) authorization ID. + clientSHA1, err := scram.SHA1.NewClient("mulder", "trustno1", "") + if err != nil { + panic(err) + } + + // Prepare the authentication conversation. Use the empty string as the + // initial server message argument to start the conversation. + conv := clientSHA1.NewConversation() + var serverMsg string + + // Get the first message, send it and read the response. + firstMsg, err := conv.Step(serverMsg) + if err != nil { + panic(err) + } + serverMsg = sendClientMsg(firstMsg) + + // Get the second message, send it, and read the response. + secondMsg, err := conv.Step(serverMsg) + if err != nil { + panic(err) + } + serverMsg = sendClientMsg(secondMsg) + + // Validate the server's final message. We have no further message to + // send so ignore that return value. + _, err = conv.Step(serverMsg) + if err != nil { + panic(err) + } + + return + } + + func sendClientMsg(s string) string { + // A real implementation would send this to a server and read a reply. + return "" + } + +## Copyright and License + +Copyright 2018 by David A. Golden. All rights reserved. + +Licensed under the Apache License, Version 2.0 (the "License"). You may +obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 diff --git a/vendor/github.com/xdg-go/scram/client.go b/vendor/github.com/xdg-go/scram/client.go new file mode 100644 index 0000000..5b53021 --- /dev/null +++ b/vendor/github.com/xdg-go/scram/client.go @@ -0,0 +1,130 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import ( + "sync" + + "github.com/xdg-go/pbkdf2" +) + +// Client implements the client side of SCRAM authentication. It holds +// configuration values needed to initialize new client-side conversations for +// a specific username, password and authorization ID tuple. Client caches +// the computationally-expensive parts of a SCRAM conversation as described in +// RFC-5802. If repeated authentication conversations may be required for a +// user (e.g. disconnect/reconnect), the user's Client should be preserved. +// +// For security reasons, Clients have a default minimum PBKDF2 iteration count +// of 4096. If a server requests a smaller iteration count, an authentication +// conversation will error. +// +// A Client can also be used by a server application to construct the hashed +// authentication values to be stored for a new user. See StoredCredentials() +// for more. +type Client struct { + sync.RWMutex + username string + password string + authzID string + minIters int + nonceGen NonceGeneratorFcn + hashGen HashGeneratorFcn + cache map[KeyFactors]derivedKeys +} + +func newClient(username, password, authzID string, fcn HashGeneratorFcn) *Client { + return &Client{ + username: username, + password: password, + authzID: authzID, + minIters: 4096, + nonceGen: defaultNonceGenerator, + hashGen: fcn, + cache: make(map[KeyFactors]derivedKeys), + } +} + +// WithMinIterations changes minimum required PBKDF2 iteration count. +func (c *Client) WithMinIterations(n int) *Client { + c.Lock() + defer c.Unlock() + c.minIters = n + return c +} + +// WithNonceGenerator replaces the default nonce generator (base64 encoding of +// 24 bytes from crypto/rand) with a custom generator. This is provided for +// testing or for users with custom nonce requirements. +func (c *Client) WithNonceGenerator(ng NonceGeneratorFcn) *Client { + c.Lock() + defer c.Unlock() + c.nonceGen = ng + return c +} + +// NewConversation constructs a client-side authentication conversation. +// Conversations cannot be reused, so this must be called for each new +// authentication attempt. +func (c *Client) NewConversation() *ClientConversation { + c.RLock() + defer c.RUnlock() + return &ClientConversation{ + client: c, + nonceGen: c.nonceGen, + hashGen: c.hashGen, + minIters: c.minIters, + } +} + +func (c *Client) getDerivedKeys(kf KeyFactors) derivedKeys { + dk, ok := c.getCache(kf) + if !ok { + dk = c.computeKeys(kf) + c.setCache(kf, dk) + } + return dk +} + +// GetStoredCredentials takes a salt and iteration count structure and +// provides the values that must be stored by a server to authentication a +// user. These values are what the Server credential lookup function must +// return for a given username. +func (c *Client) GetStoredCredentials(kf KeyFactors) StoredCredentials { + dk := c.getDerivedKeys(kf) + return StoredCredentials{ + KeyFactors: kf, + StoredKey: dk.StoredKey, + ServerKey: dk.ServerKey, + } +} + +func (c *Client) computeKeys(kf KeyFactors) derivedKeys { + h := c.hashGen() + saltedPassword := pbkdf2.Key([]byte(c.password), []byte(kf.Salt), kf.Iters, h.Size(), c.hashGen) + clientKey := computeHMAC(c.hashGen, saltedPassword, []byte("Client Key")) + + return derivedKeys{ + ClientKey: clientKey, + StoredKey: computeHash(c.hashGen, clientKey), + ServerKey: computeHMAC(c.hashGen, saltedPassword, []byte("Server Key")), + } +} + +func (c *Client) getCache(kf KeyFactors) (derivedKeys, bool) { + c.RLock() + defer c.RUnlock() + dk, ok := c.cache[kf] + return dk, ok +} + +func (c *Client) setCache(kf KeyFactors, dk derivedKeys) { + c.Lock() + defer c.Unlock() + c.cache[kf] = dk + return +} diff --git a/vendor/github.com/xdg-go/scram/client_conv.go b/vendor/github.com/xdg-go/scram/client_conv.go new file mode 100644 index 0000000..8340568 --- /dev/null +++ b/vendor/github.com/xdg-go/scram/client_conv.go @@ -0,0 +1,149 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import ( + "crypto/hmac" + "encoding/base64" + "errors" + "fmt" + "strings" +) + +type clientState int + +const ( + clientStarting clientState = iota + clientFirst + clientFinal + clientDone +) + +// ClientConversation implements the client-side of an authentication +// conversation with a server. A new conversation must be created for +// each authentication attempt. +type ClientConversation struct { + client *Client + nonceGen NonceGeneratorFcn + hashGen HashGeneratorFcn + minIters int + state clientState + valid bool + gs2 string + nonce string + c1b string + serveSig []byte +} + +// Step takes a string provided from a server (or just an empty string for the +// very first conversation step) and attempts to move the authentication +// conversation forward. It returns a string to be sent to the server or an +// error if the server message is invalid. Calling Step after a conversation +// completes is also an error. +func (cc *ClientConversation) Step(challenge string) (response string, err error) { + switch cc.state { + case clientStarting: + cc.state = clientFirst + response, err = cc.firstMsg() + case clientFirst: + cc.state = clientFinal + response, err = cc.finalMsg(challenge) + case clientFinal: + cc.state = clientDone + response, err = cc.validateServer(challenge) + default: + response, err = "", errors.New("Conversation already completed") + } + return +} + +// Done returns true if the conversation is completed or has errored. +func (cc *ClientConversation) Done() bool { + return cc.state == clientDone +} + +// Valid returns true if the conversation successfully authenticated with the +// server, including counter-validation that the server actually has the +// user's stored credentials. +func (cc *ClientConversation) Valid() bool { + return cc.valid +} + +func (cc *ClientConversation) firstMsg() (string, error) { + // Values are cached for use in final message parameters + cc.gs2 = cc.gs2Header() + cc.nonce = cc.client.nonceGen() + cc.c1b = fmt.Sprintf("n=%s,r=%s", encodeName(cc.client.username), cc.nonce) + + return cc.gs2 + cc.c1b, nil +} + +func (cc *ClientConversation) finalMsg(s1 string) (string, error) { + msg, err := parseServerFirst(s1) + if err != nil { + return "", err + } + + // Check nonce prefix and update + if !strings.HasPrefix(msg.nonce, cc.nonce) { + return "", errors.New("server nonce did not extend client nonce") + } + cc.nonce = msg.nonce + + // Check iteration count vs minimum + if msg.iters < cc.minIters { + return "", fmt.Errorf("server requested too few iterations (%d)", msg.iters) + } + + // Create client-final-message-without-proof + c2wop := fmt.Sprintf( + "c=%s,r=%s", + base64.StdEncoding.EncodeToString([]byte(cc.gs2)), + cc.nonce, + ) + + // Create auth message + authMsg := cc.c1b + "," + s1 + "," + c2wop + + // Get derived keys from client cache + dk := cc.client.getDerivedKeys(KeyFactors{Salt: string(msg.salt), Iters: msg.iters}) + + // Create proof as clientkey XOR clientsignature + clientSignature := computeHMAC(cc.hashGen, dk.StoredKey, []byte(authMsg)) + clientProof := xorBytes(dk.ClientKey, clientSignature) + proof := base64.StdEncoding.EncodeToString(clientProof) + + // Cache ServerSignature for later validation + cc.serveSig = computeHMAC(cc.hashGen, dk.ServerKey, []byte(authMsg)) + + return fmt.Sprintf("%s,p=%s", c2wop, proof), nil +} + +func (cc *ClientConversation) validateServer(s2 string) (string, error) { + msg, err := parseServerFinal(s2) + if err != nil { + return "", err + } + + if len(msg.err) > 0 { + return "", fmt.Errorf("server error: %s", msg.err) + } + + if !hmac.Equal(msg.verifier, cc.serveSig) { + return "", errors.New("server validation failed") + } + + cc.valid = true + return "", nil +} + +func (cc *ClientConversation) gs2Header() string { + if cc.client.authzID == "" { + return "n,," + } + return fmt.Sprintf("n,%s,", encodeName(cc.client.authzID)) +} diff --git a/vendor/github.com/xdg-go/scram/common.go b/vendor/github.com/xdg-go/scram/common.go new file mode 100644 index 0000000..cb705cb --- /dev/null +++ b/vendor/github.com/xdg-go/scram/common.go @@ -0,0 +1,97 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import ( + "crypto/hmac" + "crypto/rand" + "encoding/base64" + "strings" +) + +// NonceGeneratorFcn defines a function that returns a string of high-quality +// random printable ASCII characters EXCLUDING the comma (',') character. The +// default nonce generator provides Base64 encoding of 24 bytes from +// crypto/rand. +type NonceGeneratorFcn func() string + +// derivedKeys collects the three cryptographically derived values +// into one struct for caching. +type derivedKeys struct { + ClientKey []byte + StoredKey []byte + ServerKey []byte +} + +// KeyFactors represent the two server-provided factors needed to compute +// client credentials for authentication. Salt is decoded bytes (i.e. not +// base64), but in string form so that KeyFactors can be used as a map key for +// cached credentials. +type KeyFactors struct { + Salt string + Iters int +} + +// StoredCredentials are the values that a server must store for a given +// username to allow authentication. They include the salt and iteration +// count, plus the derived values to authenticate a client and for the server +// to authenticate itself back to the client. +// +// NOTE: these are specific to a given hash function. To allow a user to +// authenticate with either SCRAM-SHA-1 or SCRAM-SHA-256, two sets of +// StoredCredentials must be created and stored, one for each hash function. +type StoredCredentials struct { + KeyFactors + StoredKey []byte + ServerKey []byte +} + +// CredentialLookup is a callback to provide StoredCredentials for a given +// username. This is used to configure Server objects. +// +// NOTE: these are specific to a given hash function. The callback provided +// to a Server with a given hash function must provide the corresponding +// StoredCredentials. +type CredentialLookup func(string) (StoredCredentials, error) + +func defaultNonceGenerator() string { + raw := make([]byte, 24) + nonce := make([]byte, base64.StdEncoding.EncodedLen(len(raw))) + rand.Read(raw) + base64.StdEncoding.Encode(nonce, raw) + return string(nonce) +} + +func encodeName(s string) string { + return strings.Replace(strings.Replace(s, "=", "=3D", -1), ",", "=2C", -1) +} + +func decodeName(s string) (string, error) { + // TODO Check for = not followed by 2C or 3D + return strings.Replace(strings.Replace(s, "=2C", ",", -1), "=3D", "=", -1), nil +} + +func computeHash(hg HashGeneratorFcn, b []byte) []byte { + h := hg() + h.Write(b) + return h.Sum(nil) +} + +func computeHMAC(hg HashGeneratorFcn, key, data []byte) []byte { + mac := hmac.New(hg, key) + mac.Write(data) + return mac.Sum(nil) +} + +func xorBytes(a, b []byte) []byte { + // TODO check a & b are same length, or just xor to smallest + xor := make([]byte, len(a)) + for i := range a { + xor[i] = a[i] ^ b[i] + } + return xor +} diff --git a/vendor/github.com/xdg-go/scram/doc.go b/vendor/github.com/xdg-go/scram/doc.go new file mode 100644 index 0000000..82e8aee --- /dev/null +++ b/vendor/github.com/xdg-go/scram/doc.go @@ -0,0 +1,26 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +// Package scram provides client and server implementations of the Salted +// Challenge Response Authentication Mechanism (SCRAM) described in RFC-5802 +// and RFC-7677. +// +// Usage +// +// The scram package provides variables, `SHA1`, `SHA256`, and `SHA512`, that +// are used to construct Client or Server objects. +// +// clientSHA1, err := scram.SHA1.NewClient(username, password, authID) +// clientSHA256, err := scram.SHA256.NewClient(username, password, authID) +// clientSHA512, err := scram.SHA512.NewClient(username, password, authID) +// +// serverSHA1, err := scram.SHA1.NewServer(credentialLookupFcn) +// serverSHA256, err := scram.SHA256.NewServer(credentialLookupFcn) +// serverSHA512, err := scram.SHA512.NewServer(credentialLookupFcn) +// +// These objects are used to construct ClientConversation or +// ServerConversation objects that are used to carry out authentication. +package scram diff --git a/vendor/github.com/xdg-go/scram/parse.go b/vendor/github.com/xdg-go/scram/parse.go new file mode 100644 index 0000000..722f604 --- /dev/null +++ b/vendor/github.com/xdg-go/scram/parse.go @@ -0,0 +1,205 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import ( + "encoding/base64" + "errors" + "fmt" + "strconv" + "strings" +) + +type c1Msg struct { + gs2Header string + authzID string + username string + nonce string + c1b string +} + +type c2Msg struct { + cbind []byte + nonce string + proof []byte + c2wop string +} + +type s1Msg struct { + nonce string + salt []byte + iters int +} + +type s2Msg struct { + verifier []byte + err string +} + +func parseField(s, k string) (string, error) { + t := strings.TrimPrefix(s, k+"=") + if t == s { + return "", fmt.Errorf("error parsing '%s' for field '%s'", s, k) + } + return t, nil +} + +func parseGS2Flag(s string) (string, error) { + if s[0] == 'p' { + return "", fmt.Errorf("channel binding requested but not supported") + } + + if s == "n" || s == "y" { + return s, nil + } + + return "", fmt.Errorf("error parsing '%s' for gs2 flag", s) +} + +func parseFieldBase64(s, k string) ([]byte, error) { + raw, err := parseField(s, k) + if err != nil { + return nil, err + } + + dec, err := base64.StdEncoding.DecodeString(raw) + if err != nil { + return nil, err + } + + return dec, nil +} + +func parseFieldInt(s, k string) (int, error) { + raw, err := parseField(s, k) + if err != nil { + return 0, err + } + + num, err := strconv.Atoi(raw) + if err != nil { + return 0, fmt.Errorf("error parsing field '%s': %v", k, err) + } + + return num, nil +} + +func parseClientFirst(c1 string) (msg c1Msg, err error) { + + fields := strings.Split(c1, ",") + if len(fields) < 4 { + err = errors.New("not enough fields in first server message") + return + } + + gs2flag, err := parseGS2Flag(fields[0]) + if err != nil { + return + } + + // 'a' field is optional + if len(fields[1]) > 0 { + msg.authzID, err = parseField(fields[1], "a") + if err != nil { + return + } + } + + // Recombine and save the gs2 header + msg.gs2Header = gs2flag + "," + msg.authzID + "," + + // Check for unsupported extensions field "m". + if strings.HasPrefix(fields[2], "m=") { + err = errors.New("SCRAM message extensions are not supported") + return + } + + msg.username, err = parseField(fields[2], "n") + if err != nil { + return + } + + msg.nonce, err = parseField(fields[3], "r") + if err != nil { + return + } + + msg.c1b = strings.Join(fields[2:], ",") + + return +} + +func parseClientFinal(c2 string) (msg c2Msg, err error) { + fields := strings.Split(c2, ",") + if len(fields) < 3 { + err = errors.New("not enough fields in first server message") + return + } + + msg.cbind, err = parseFieldBase64(fields[0], "c") + if err != nil { + return + } + + msg.nonce, err = parseField(fields[1], "r") + if err != nil { + return + } + + // Extension fields may come between nonce and proof, so we + // grab the *last* fields as proof. + msg.proof, err = parseFieldBase64(fields[len(fields)-1], "p") + if err != nil { + return + } + + msg.c2wop = c2[:strings.LastIndex(c2, ",")] + + return +} + +func parseServerFirst(s1 string) (msg s1Msg, err error) { + + // Check for unsupported extensions field "m". + if strings.HasPrefix(s1, "m=") { + err = errors.New("SCRAM message extensions are not supported") + return + } + + fields := strings.Split(s1, ",") + if len(fields) < 3 { + err = errors.New("not enough fields in first server message") + return + } + + msg.nonce, err = parseField(fields[0], "r") + if err != nil { + return + } + + msg.salt, err = parseFieldBase64(fields[1], "s") + if err != nil { + return + } + + msg.iters, err = parseFieldInt(fields[2], "i") + + return +} + +func parseServerFinal(s2 string) (msg s2Msg, err error) { + fields := strings.Split(s2, ",") + + msg.verifier, err = parseFieldBase64(fields[0], "v") + if err == nil { + return + } + + msg.err, err = parseField(fields[0], "e") + + return +} diff --git a/vendor/github.com/xdg-go/scram/scram.go b/vendor/github.com/xdg-go/scram/scram.go new file mode 100644 index 0000000..a7b3660 --- /dev/null +++ b/vendor/github.com/xdg-go/scram/scram.go @@ -0,0 +1,71 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import ( + "crypto/sha1" + "crypto/sha256" + "crypto/sha512" + "fmt" + "hash" + + "github.com/xdg-go/stringprep" +) + +// HashGeneratorFcn abstracts a factory function that returns a hash.Hash +// value to be used for SCRAM operations. Generally, one would use the +// provided package variables, `scram.SHA1` and `scram.SHA256`, for the most +// common forms of SCRAM. +type HashGeneratorFcn func() hash.Hash + +// SHA1 is a function that returns a crypto/sha1 hasher and should be used to +// create Client objects configured for SHA-1 hashing. +var SHA1 HashGeneratorFcn = func() hash.Hash { return sha1.New() } + +// SHA256 is a function that returns a crypto/sha256 hasher and should be used +// to create Client objects configured for SHA-256 hashing. +var SHA256 HashGeneratorFcn = func() hash.Hash { return sha256.New() } + +// SHA512 is a function that returns a crypto/sha512 hasher and should be used +// to create Client objects configured for SHA-512 hashing. +var SHA512 HashGeneratorFcn = func() hash.Hash { return sha512.New() } + +// NewClient constructs a SCRAM client component based on a given hash.Hash +// factory receiver. This constructor will normalize the username, password +// and authzID via the SASLprep algorithm, as recommended by RFC-5802. If +// SASLprep fails, the method returns an error. +func (f HashGeneratorFcn) NewClient(username, password, authzID string) (*Client, error) { + var userprep, passprep, authprep string + var err error + + if userprep, err = stringprep.SASLprep.Prepare(username); err != nil { + return nil, fmt.Errorf("Error SASLprepping username '%s': %v", username, err) + } + if passprep, err = stringprep.SASLprep.Prepare(password); err != nil { + return nil, fmt.Errorf("Error SASLprepping password '%s': %v", password, err) + } + if authprep, err = stringprep.SASLprep.Prepare(authzID); err != nil { + return nil, fmt.Errorf("Error SASLprepping authzID '%s': %v", authzID, err) + } + + return newClient(userprep, passprep, authprep, f), nil +} + +// NewClientUnprepped acts like NewClient, except none of the arguments will +// be normalized via SASLprep. This is not generally recommended, but is +// provided for users that may have custom normalization needs. +func (f HashGeneratorFcn) NewClientUnprepped(username, password, authzID string) (*Client, error) { + return newClient(username, password, authzID, f), nil +} + +// NewServer constructs a SCRAM server component based on a given hash.Hash +// factory receiver. To be maximally generic, it uses dependency injection to +// handle credential lookup, which is the process of turning a username string +// into a struct with stored credentials for authentication. +func (f HashGeneratorFcn) NewServer(cl CredentialLookup) (*Server, error) { + return newServer(cl, f) +} diff --git a/vendor/github.com/xdg-go/scram/server.go b/vendor/github.com/xdg-go/scram/server.go new file mode 100644 index 0000000..b119b36 --- /dev/null +++ b/vendor/github.com/xdg-go/scram/server.go @@ -0,0 +1,50 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import "sync" + +// Server implements the server side of SCRAM authentication. It holds +// configuration values needed to initialize new server-side conversations. +// Generally, this can be persistent within an application. +type Server struct { + sync.RWMutex + credentialCB CredentialLookup + nonceGen NonceGeneratorFcn + hashGen HashGeneratorFcn +} + +func newServer(cl CredentialLookup, fcn HashGeneratorFcn) (*Server, error) { + return &Server{ + credentialCB: cl, + nonceGen: defaultNonceGenerator, + hashGen: fcn, + }, nil +} + +// WithNonceGenerator replaces the default nonce generator (base64 encoding of +// 24 bytes from crypto/rand) with a custom generator. This is provided for +// testing or for users with custom nonce requirements. +func (s *Server) WithNonceGenerator(ng NonceGeneratorFcn) *Server { + s.Lock() + defer s.Unlock() + s.nonceGen = ng + return s +} + +// NewConversation constructs a server-side authentication conversation. +// Conversations cannot be reused, so this must be called for each new +// authentication attempt. +func (s *Server) NewConversation() *ServerConversation { + s.RLock() + defer s.RUnlock() + return &ServerConversation{ + nonceGen: s.nonceGen, + hashGen: s.hashGen, + credentialCB: s.credentialCB, + } +} diff --git a/vendor/github.com/xdg-go/scram/server_conv.go b/vendor/github.com/xdg-go/scram/server_conv.go new file mode 100644 index 0000000..9c8838c --- /dev/null +++ b/vendor/github.com/xdg-go/scram/server_conv.go @@ -0,0 +1,151 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package scram + +import ( + "crypto/hmac" + "encoding/base64" + "errors" + "fmt" +) + +type serverState int + +const ( + serverFirst serverState = iota + serverFinal + serverDone +) + +// ServerConversation implements the server-side of an authentication +// conversation with a client. A new conversation must be created for +// each authentication attempt. +type ServerConversation struct { + nonceGen NonceGeneratorFcn + hashGen HashGeneratorFcn + credentialCB CredentialLookup + state serverState + credential StoredCredentials + valid bool + gs2Header string + username string + authzID string + nonce string + c1b string + s1 string +} + +// Step takes a string provided from a client and attempts to move the +// authentication conversation forward. It returns a string to be sent to the +// client or an error if the client message is invalid. Calling Step after a +// conversation completes is also an error. +func (sc *ServerConversation) Step(challenge string) (response string, err error) { + switch sc.state { + case serverFirst: + sc.state = serverFinal + response, err = sc.firstMsg(challenge) + case serverFinal: + sc.state = serverDone + response, err = sc.finalMsg(challenge) + default: + response, err = "", errors.New("Conversation already completed") + } + return +} + +// Done returns true if the conversation is completed or has errored. +func (sc *ServerConversation) Done() bool { + return sc.state == serverDone +} + +// Valid returns true if the conversation successfully authenticated the +// client. +func (sc *ServerConversation) Valid() bool { + return sc.valid +} + +// Username returns the client-provided username. This is valid to call +// if the first conversation Step() is successful. +func (sc *ServerConversation) Username() string { + return sc.username +} + +// AuthzID returns the (optional) client-provided authorization identity, if +// any. If one was not provided, it returns the empty string. This is valid +// to call if the first conversation Step() is successful. +func (sc *ServerConversation) AuthzID() string { + return sc.authzID +} + +func (sc *ServerConversation) firstMsg(c1 string) (string, error) { + msg, err := parseClientFirst(c1) + if err != nil { + sc.state = serverDone + return "", err + } + + sc.gs2Header = msg.gs2Header + sc.username = msg.username + sc.authzID = msg.authzID + + sc.credential, err = sc.credentialCB(msg.username) + if err != nil { + sc.state = serverDone + return "e=unknown-user", err + } + + sc.nonce = msg.nonce + sc.nonceGen() + sc.c1b = msg.c1b + sc.s1 = fmt.Sprintf("r=%s,s=%s,i=%d", + sc.nonce, + base64.StdEncoding.EncodeToString([]byte(sc.credential.Salt)), + sc.credential.Iters, + ) + + return sc.s1, nil +} + +// For errors, returns server error message as well as non-nil error. Callers +// can choose whether to send server error or not. +func (sc *ServerConversation) finalMsg(c2 string) (string, error) { + msg, err := parseClientFinal(c2) + if err != nil { + return "", err + } + + // Check channel binding matches what we expect; in this case, we expect + // just the gs2 header we received as we don't support channel binding + // with a data payload. If we add binding, we need to independently + // compute the header to match here. + if string(msg.cbind) != sc.gs2Header { + return "e=channel-bindings-dont-match", fmt.Errorf("channel binding received '%s' doesn't match expected '%s'", msg.cbind, sc.gs2Header) + } + + // Check nonce received matches what we sent + if msg.nonce != sc.nonce { + return "e=other-error", errors.New("nonce received did not match nonce sent") + } + + // Create auth message + authMsg := sc.c1b + "," + sc.s1 + "," + msg.c2wop + + // Retrieve ClientKey from proof and verify it + clientSignature := computeHMAC(sc.hashGen, sc.credential.StoredKey, []byte(authMsg)) + clientKey := xorBytes([]byte(msg.proof), clientSignature) + storedKey := computeHash(sc.hashGen, clientKey) + + // Compare with constant-time function + if !hmac.Equal(storedKey, sc.credential.StoredKey) { + return "e=invalid-proof", errors.New("challenge proof invalid") + } + + sc.valid = true + + // Compute and return server verifier + serverSignature := computeHMAC(sc.hashGen, sc.credential.ServerKey, []byte(authMsg)) + return "v=" + base64.StdEncoding.EncodeToString(serverSignature), nil +} diff --git a/vendor/github.com/xdg-go/stringprep/.gitignore b/vendor/github.com/xdg-go/stringprep/.gitignore new file mode 100644 index 0000000..e69de29 diff --git a/vendor/github.com/xdg-go/stringprep/CHANGELOG.md b/vendor/github.com/xdg-go/stringprep/CHANGELOG.md new file mode 100644 index 0000000..e06787f --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/CHANGELOG.md @@ -0,0 +1,29 @@ +# CHANGELOG + + +## [v1.0.3] - 2022-03-01 + +### Maintenance + +- Bump golang.org/x/text to v0.3.7 due to CVE-2021-38561 + + +## [v1.0.2] - 2021-03-27 + +### Maintenance + +- Change minimum Go version to 1.11 + + +## [v1.0.1] - 2021-03-24 + +### Bug Fixes + +- Add go.mod file + + +## [v1.0.0] - 2018-02-21 + +[v1.0.2]: https://github.com/xdg-go/stringprep/releases/tag/v1.0.2 +[v1.0.1]: https://github.com/xdg-go/stringprep/releases/tag/v1.0.1 +[v1.0.0]: https://github.com/xdg-go/stringprep/releases/tag/v1.0.0 diff --git a/vendor/github.com/xdg-go/stringprep/LICENSE b/vendor/github.com/xdg-go/stringprep/LICENSE new file mode 100644 index 0000000..67db858 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/LICENSE @@ -0,0 +1,175 @@ + + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. diff --git a/vendor/github.com/xdg-go/stringprep/README.md b/vendor/github.com/xdg-go/stringprep/README.md new file mode 100644 index 0000000..83ea534 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/README.md @@ -0,0 +1,28 @@ +[![Go Reference](https://pkg.go.dev/badge/github.com/xdg-go/stringprep.svg)](https://pkg.go.dev/github.com/xdg-go/stringprep) +[![Go Report Card](https://goreportcard.com/badge/github.com/xdg-go/stringprep)](https://goreportcard.com/report/github.com/xdg-go/stringprep) +[![Github Actions](https://github.com/xdg-go/stringprep/actions/workflows/test.yml/badge.svg)](https://github.com/xdg-go/stringprep/actions/workflows/test.yml) + +# stringprep – Go implementation of RFC-3454 stringprep and RFC-4013 SASLprep + +## Synopsis + +``` + import "github.com/xdg-go/stringprep" + + prepped := stringprep.SASLprep.Prepare("TrustNô1") + +``` + +## Description + +This library provides an implementation of the stringprep algorithm +(RFC-3454) in Go, including all data tables. + +A pre-built SASLprep (RFC-4013) profile is provided as well. + +## Copyright and License + +Copyright 2018 by David A. Golden. All rights reserved. + +Licensed under the Apache License, Version 2.0 (the "License"). You may +obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 diff --git a/vendor/github.com/xdg-go/stringprep/bidi.go b/vendor/github.com/xdg-go/stringprep/bidi.go new file mode 100644 index 0000000..6f6d321 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/bidi.go @@ -0,0 +1,73 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package stringprep + +var errHasLCat = "BiDi string can't have runes from category L" +var errFirstRune = "BiDi string first rune must have category R or AL" +var errLastRune = "BiDi string last rune must have category R or AL" + +// Check for prohibited characters from table C.8 +func checkBiDiProhibitedRune(s string) error { + for _, r := range s { + if TableC8.Contains(r) { + return Error{Msg: errProhibited, Rune: r} + } + } + return nil +} + +// Check for LCat characters from table D.2 +func checkBiDiLCat(s string) error { + for _, r := range s { + if TableD2.Contains(r) { + return Error{Msg: errHasLCat, Rune: r} + } + } + return nil +} + +// Check first and last characters are in table D.1; requires non-empty string +func checkBadFirstAndLastRandALCat(s string) error { + rs := []rune(s) + if !TableD1.Contains(rs[0]) { + return Error{Msg: errFirstRune, Rune: rs[0]} + } + n := len(rs) - 1 + if !TableD1.Contains(rs[n]) { + return Error{Msg: errLastRune, Rune: rs[n]} + } + return nil +} + +// Look for RandALCat characters from table D.1 +func hasBiDiRandALCat(s string) bool { + for _, r := range s { + if TableD1.Contains(r) { + return true + } + } + return false +} + +// Check that BiDi rules are satisfied ; let empty string pass this rule +func passesBiDiRules(s string) error { + if len(s) == 0 { + return nil + } + if err := checkBiDiProhibitedRune(s); err != nil { + return err + } + if hasBiDiRandALCat(s) { + if err := checkBiDiLCat(s); err != nil { + return err + } + if err := checkBadFirstAndLastRandALCat(s); err != nil { + return err + } + } + return nil +} diff --git a/vendor/github.com/xdg-go/stringprep/doc.go b/vendor/github.com/xdg-go/stringprep/doc.go new file mode 100644 index 0000000..b319e08 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/doc.go @@ -0,0 +1,10 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +// Package stringprep provides data tables and algorithms for RFC-3454, +// including errata (as of 2018-02). It also provides a profile for +// SASLprep as defined in RFC-4013. +package stringprep diff --git a/vendor/github.com/xdg-go/stringprep/error.go b/vendor/github.com/xdg-go/stringprep/error.go new file mode 100644 index 0000000..7403e49 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/error.go @@ -0,0 +1,14 @@ +package stringprep + +import "fmt" + +// Error describes problems encountered during stringprep, including what rune +// was problematic. +type Error struct { + Msg string + Rune rune +} + +func (e Error) Error() string { + return fmt.Sprintf("%s (rune: '\\u%04x')", e.Msg, e.Rune) +} diff --git a/vendor/github.com/xdg-go/stringprep/map.go b/vendor/github.com/xdg-go/stringprep/map.go new file mode 100644 index 0000000..e56a0dd --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/map.go @@ -0,0 +1,21 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package stringprep + +// Mapping represents a stringprep mapping, from a single rune to zero or more +// runes. +type Mapping map[rune][]rune + +// Map maps a rune to a (possibly empty) rune slice via a stringprep Mapping. +// The ok return value is false if the rune was not found. +func (m Mapping) Map(r rune) (replacement []rune, ok bool) { + rs, ok := m[r] + if !ok { + return nil, false + } + return rs, true +} diff --git a/vendor/github.com/xdg-go/stringprep/profile.go b/vendor/github.com/xdg-go/stringprep/profile.go new file mode 100644 index 0000000..5a73be9 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/profile.go @@ -0,0 +1,75 @@ +package stringprep + +import ( + "golang.org/x/text/unicode/norm" +) + +// Profile represents a stringprep profile. +type Profile struct { + Mappings []Mapping + Normalize bool + Prohibits []Set + CheckBiDi bool +} + +var errProhibited = "prohibited character" + +// Prepare transforms an input string to an output string following +// the rules defined in the profile as defined by RFC-3454. +func (p Profile) Prepare(s string) (string, error) { + // Optimistically, assume output will be same length as input + temp := make([]rune, 0, len(s)) + + // Apply maps + for _, r := range s { + rs, ok := p.applyMaps(r) + if ok { + temp = append(temp, rs...) + } else { + temp = append(temp, r) + } + } + + // Normalize + var out string + if p.Normalize { + out = norm.NFKC.String(string(temp)) + } else { + out = string(temp) + } + + // Check prohibited + for _, r := range out { + if p.runeIsProhibited(r) { + return "", Error{Msg: errProhibited, Rune: r} + } + } + + // Check BiDi allowed + if p.CheckBiDi { + if err := passesBiDiRules(out); err != nil { + return "", err + } + } + + return out, nil +} + +func (p Profile) applyMaps(r rune) ([]rune, bool) { + for _, m := range p.Mappings { + rs, ok := m.Map(r) + if ok { + return rs, true + } + } + return nil, false +} + +func (p Profile) runeIsProhibited(r rune) bool { + for _, s := range p.Prohibits { + if s.Contains(r) { + return true + } + } + return false +} diff --git a/vendor/github.com/xdg-go/stringprep/saslprep.go b/vendor/github.com/xdg-go/stringprep/saslprep.go new file mode 100644 index 0000000..4001348 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/saslprep.go @@ -0,0 +1,52 @@ +package stringprep + +var mapNonASCIISpaceToASCIISpace = Mapping{ + 0x00A0: []rune{0x0020}, + 0x1680: []rune{0x0020}, + 0x2000: []rune{0x0020}, + 0x2001: []rune{0x0020}, + 0x2002: []rune{0x0020}, + 0x2003: []rune{0x0020}, + 0x2004: []rune{0x0020}, + 0x2005: []rune{0x0020}, + 0x2006: []rune{0x0020}, + 0x2007: []rune{0x0020}, + 0x2008: []rune{0x0020}, + 0x2009: []rune{0x0020}, + 0x200A: []rune{0x0020}, + 0x200B: []rune{0x0020}, + 0x202F: []rune{0x0020}, + 0x205F: []rune{0x0020}, + 0x3000: []rune{0x0020}, +} + +// SASLprep is a pre-defined stringprep profile for user names and passwords +// as described in RFC-4013. +// +// Because the stringprep distinction between query and stored strings was +// intended for compatibility across profile versions, but SASLprep was never +// updated and is now deprecated, this profile only operates in stored +// strings mode, prohibiting unassigned code points. +var SASLprep Profile = saslprep + +var saslprep = Profile{ + Mappings: []Mapping{ + TableB1, + mapNonASCIISpaceToASCIISpace, + }, + Normalize: true, + Prohibits: []Set{ + TableA1, + TableC1_2, + TableC2_1, + TableC2_2, + TableC3, + TableC4, + TableC5, + TableC6, + TableC7, + TableC8, + TableC9, + }, + CheckBiDi: true, +} diff --git a/vendor/github.com/xdg-go/stringprep/set.go b/vendor/github.com/xdg-go/stringprep/set.go new file mode 100644 index 0000000..c837e28 --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/set.go @@ -0,0 +1,36 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package stringprep + +import "sort" + +// RuneRange represents a close-ended range of runes: [N,M]. For a range +// consisting of a single rune, N and M will be equal. +type RuneRange [2]rune + +// Contains returns true if a rune is within the bounds of the RuneRange. +func (rr RuneRange) Contains(r rune) bool { + return rr[0] <= r && r <= rr[1] +} + +func (rr RuneRange) isAbove(r rune) bool { + return r <= rr[0] +} + +// Set represents a stringprep data table used to identify runes of a +// particular type. +type Set []RuneRange + +// Contains returns true if a rune is within any of the RuneRanges in the +// Set. +func (s Set) Contains(r rune) bool { + i := sort.Search(len(s), func(i int) bool { return s[i].Contains(r) || s[i].isAbove(r) }) + if i < len(s) && s[i].Contains(r) { + return true + } + return false +} diff --git a/vendor/github.com/xdg-go/stringprep/tables.go b/vendor/github.com/xdg-go/stringprep/tables.go new file mode 100644 index 0000000..c3fc1fa --- /dev/null +++ b/vendor/github.com/xdg-go/stringprep/tables.go @@ -0,0 +1,3215 @@ +// Copyright 2018 by David A. Golden. All rights reserved. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package stringprep + +var tableA1 = Set{ + RuneRange{0x0221, 0x0221}, + RuneRange{0x0234, 0x024F}, + RuneRange{0x02AE, 0x02AF}, + RuneRange{0x02EF, 0x02FF}, + RuneRange{0x0350, 0x035F}, + RuneRange{0x0370, 0x0373}, + RuneRange{0x0376, 0x0379}, + RuneRange{0x037B, 0x037D}, + RuneRange{0x037F, 0x0383}, + RuneRange{0x038B, 0x038B}, + RuneRange{0x038D, 0x038D}, + RuneRange{0x03A2, 0x03A2}, + RuneRange{0x03CF, 0x03CF}, + RuneRange{0x03F7, 0x03FF}, + RuneRange{0x0487, 0x0487}, + RuneRange{0x04CF, 0x04CF}, + RuneRange{0x04F6, 0x04F7}, + RuneRange{0x04FA, 0x04FF}, + RuneRange{0x0510, 0x0530}, + RuneRange{0x0557, 0x0558}, + RuneRange{0x0560, 0x0560}, + RuneRange{0x0588, 0x0588}, + RuneRange{0x058B, 0x0590}, + RuneRange{0x05A2, 0x05A2}, + RuneRange{0x05BA, 0x05BA}, + RuneRange{0x05C5, 0x05CF}, + RuneRange{0x05EB, 0x05EF}, + RuneRange{0x05F5, 0x060B}, + RuneRange{0x060D, 0x061A}, + RuneRange{0x061C, 0x061E}, + RuneRange{0x0620, 0x0620}, + RuneRange{0x063B, 0x063F}, + RuneRange{0x0656, 0x065F}, + RuneRange{0x06EE, 0x06EF}, + RuneRange{0x06FF, 0x06FF}, + RuneRange{0x070E, 0x070E}, + RuneRange{0x072D, 0x072F}, + RuneRange{0x074B, 0x077F}, + RuneRange{0x07B2, 0x0900}, + RuneRange{0x0904, 0x0904}, + RuneRange{0x093A, 0x093B}, + RuneRange{0x094E, 0x094F}, + RuneRange{0x0955, 0x0957}, + RuneRange{0x0971, 0x0980}, + RuneRange{0x0984, 0x0984}, + RuneRange{0x098D, 0x098E}, + RuneRange{0x0991, 0x0992}, + RuneRange{0x09A9, 0x09A9}, + RuneRange{0x09B1, 0x09B1}, + RuneRange{0x09B3, 0x09B5}, + RuneRange{0x09BA, 0x09BB}, + RuneRange{0x09BD, 0x09BD}, + RuneRange{0x09C5, 0x09C6}, + RuneRange{0x09C9, 0x09CA}, + RuneRange{0x09CE, 0x09D6}, + RuneRange{0x09D8, 0x09DB}, + RuneRange{0x09DE, 0x09DE}, + RuneRange{0x09E4, 0x09E5}, + RuneRange{0x09FB, 0x0A01}, + RuneRange{0x0A03, 0x0A04}, + RuneRange{0x0A0B, 0x0A0E}, + RuneRange{0x0A11, 0x0A12}, + RuneRange{0x0A29, 0x0A29}, + RuneRange{0x0A31, 0x0A31}, + RuneRange{0x0A34, 0x0A34}, + RuneRange{0x0A37, 0x0A37}, + RuneRange{0x0A3A, 0x0A3B}, + RuneRange{0x0A3D, 0x0A3D}, + RuneRange{0x0A43, 0x0A46}, + RuneRange{0x0A49, 0x0A4A}, + RuneRange{0x0A4E, 0x0A58}, + RuneRange{0x0A5D, 0x0A5D}, + RuneRange{0x0A5F, 0x0A65}, + RuneRange{0x0A75, 0x0A80}, + RuneRange{0x0A84, 0x0A84}, + RuneRange{0x0A8C, 0x0A8C}, + RuneRange{0x0A8E, 0x0A8E}, + RuneRange{0x0A92, 0x0A92}, + RuneRange{0x0AA9, 0x0AA9}, + RuneRange{0x0AB1, 0x0AB1}, + RuneRange{0x0AB4, 0x0AB4}, + RuneRange{0x0ABA, 0x0ABB}, + RuneRange{0x0AC6, 0x0AC6}, + RuneRange{0x0ACA, 0x0ACA}, + RuneRange{0x0ACE, 0x0ACF}, + RuneRange{0x0AD1, 0x0ADF}, + RuneRange{0x0AE1, 0x0AE5}, + RuneRange{0x0AF0, 0x0B00}, + RuneRange{0x0B04, 0x0B04}, + RuneRange{0x0B0D, 0x0B0E}, + RuneRange{0x0B11, 0x0B12}, + RuneRange{0x0B29, 0x0B29}, + RuneRange{0x0B31, 0x0B31}, + RuneRange{0x0B34, 0x0B35}, + RuneRange{0x0B3A, 0x0B3B}, + RuneRange{0x0B44, 0x0B46}, + RuneRange{0x0B49, 0x0B4A}, + RuneRange{0x0B4E, 0x0B55}, + RuneRange{0x0B58, 0x0B5B}, + RuneRange{0x0B5E, 0x0B5E}, + RuneRange{0x0B62, 0x0B65}, + RuneRange{0x0B71, 0x0B81}, + RuneRange{0x0B84, 0x0B84}, + RuneRange{0x0B8B, 0x0B8D}, + RuneRange{0x0B91, 0x0B91}, + RuneRange{0x0B96, 0x0B98}, + RuneRange{0x0B9B, 0x0B9B}, + RuneRange{0x0B9D, 0x0B9D}, + RuneRange{0x0BA0, 0x0BA2}, + RuneRange{0x0BA5, 0x0BA7}, + RuneRange{0x0BAB, 0x0BAD}, + RuneRange{0x0BB6, 0x0BB6}, + RuneRange{0x0BBA, 0x0BBD}, + RuneRange{0x0BC3, 0x0BC5}, + RuneRange{0x0BC9, 0x0BC9}, + RuneRange{0x0BCE, 0x0BD6}, + RuneRange{0x0BD8, 0x0BE6}, + RuneRange{0x0BF3, 0x0C00}, + RuneRange{0x0C04, 0x0C04}, + RuneRange{0x0C0D, 0x0C0D}, + RuneRange{0x0C11, 0x0C11}, + RuneRange{0x0C29, 0x0C29}, + RuneRange{0x0C34, 0x0C34}, + RuneRange{0x0C3A, 0x0C3D}, + RuneRange{0x0C45, 0x0C45}, + RuneRange{0x0C49, 0x0C49}, + RuneRange{0x0C4E, 0x0C54}, + RuneRange{0x0C57, 0x0C5F}, + RuneRange{0x0C62, 0x0C65}, + RuneRange{0x0C70, 0x0C81}, + RuneRange{0x0C84, 0x0C84}, + RuneRange{0x0C8D, 0x0C8D}, + RuneRange{0x0C91, 0x0C91}, + RuneRange{0x0CA9, 0x0CA9}, + RuneRange{0x0CB4, 0x0CB4}, + RuneRange{0x0CBA, 0x0CBD}, + RuneRange{0x0CC5, 0x0CC5}, + RuneRange{0x0CC9, 0x0CC9}, + RuneRange{0x0CCE, 0x0CD4}, + RuneRange{0x0CD7, 0x0CDD}, + RuneRange{0x0CDF, 0x0CDF}, + RuneRange{0x0CE2, 0x0CE5}, + RuneRange{0x0CF0, 0x0D01}, + RuneRange{0x0D04, 0x0D04}, + RuneRange{0x0D0D, 0x0D0D}, + RuneRange{0x0D11, 0x0D11}, + RuneRange{0x0D29, 0x0D29}, + RuneRange{0x0D3A, 0x0D3D}, + RuneRange{0x0D44, 0x0D45}, + RuneRange{0x0D49, 0x0D49}, + RuneRange{0x0D4E, 0x0D56}, + RuneRange{0x0D58, 0x0D5F}, + RuneRange{0x0D62, 0x0D65}, + RuneRange{0x0D70, 0x0D81}, + RuneRange{0x0D84, 0x0D84}, + RuneRange{0x0D97, 0x0D99}, + RuneRange{0x0DB2, 0x0DB2}, + RuneRange{0x0DBC, 0x0DBC}, + RuneRange{0x0DBE, 0x0DBF}, + RuneRange{0x0DC7, 0x0DC9}, + RuneRange{0x0DCB, 0x0DCE}, + RuneRange{0x0DD5, 0x0DD5}, + RuneRange{0x0DD7, 0x0DD7}, + RuneRange{0x0DE0, 0x0DF1}, + RuneRange{0x0DF5, 0x0E00}, + RuneRange{0x0E3B, 0x0E3E}, + RuneRange{0x0E5C, 0x0E80}, + RuneRange{0x0E83, 0x0E83}, + RuneRange{0x0E85, 0x0E86}, + RuneRange{0x0E89, 0x0E89}, + RuneRange{0x0E8B, 0x0E8C}, + RuneRange{0x0E8E, 0x0E93}, + RuneRange{0x0E98, 0x0E98}, + RuneRange{0x0EA0, 0x0EA0}, + RuneRange{0x0EA4, 0x0EA4}, + RuneRange{0x0EA6, 0x0EA6}, + RuneRange{0x0EA8, 0x0EA9}, + RuneRange{0x0EAC, 0x0EAC}, + RuneRange{0x0EBA, 0x0EBA}, + RuneRange{0x0EBE, 0x0EBF}, + RuneRange{0x0EC5, 0x0EC5}, + RuneRange{0x0EC7, 0x0EC7}, + RuneRange{0x0ECE, 0x0ECF}, + RuneRange{0x0EDA, 0x0EDB}, + RuneRange{0x0EDE, 0x0EFF}, + RuneRange{0x0F48, 0x0F48}, + RuneRange{0x0F6B, 0x0F70}, + RuneRange{0x0F8C, 0x0F8F}, + RuneRange{0x0F98, 0x0F98}, + RuneRange{0x0FBD, 0x0FBD}, + RuneRange{0x0FCD, 0x0FCE}, + RuneRange{0x0FD0, 0x0FFF}, + RuneRange{0x1022, 0x1022}, + RuneRange{0x1028, 0x1028}, + RuneRange{0x102B, 0x102B}, + RuneRange{0x1033, 0x1035}, + RuneRange{0x103A, 0x103F}, + RuneRange{0x105A, 0x109F}, + RuneRange{0x10C6, 0x10CF}, + RuneRange{0x10F9, 0x10FA}, + RuneRange{0x10FC, 0x10FF}, + RuneRange{0x115A, 0x115E}, + RuneRange{0x11A3, 0x11A7}, + RuneRange{0x11FA, 0x11FF}, + RuneRange{0x1207, 0x1207}, + RuneRange{0x1247, 0x1247}, + RuneRange{0x1249, 0x1249}, + RuneRange{0x124E, 0x124F}, + RuneRange{0x1257, 0x1257}, + RuneRange{0x1259, 0x1259}, + RuneRange{0x125E, 0x125F}, + RuneRange{0x1287, 0x1287}, + RuneRange{0x1289, 0x1289}, + RuneRange{0x128E, 0x128F}, + RuneRange{0x12AF, 0x12AF}, + RuneRange{0x12B1, 0x12B1}, + RuneRange{0x12B6, 0x12B7}, + RuneRange{0x12BF, 0x12BF}, + RuneRange{0x12C1, 0x12C1}, + RuneRange{0x12C6, 0x12C7}, + RuneRange{0x12CF, 0x12CF}, + RuneRange{0x12D7, 0x12D7}, + RuneRange{0x12EF, 0x12EF}, + RuneRange{0x130F, 0x130F}, + RuneRange{0x1311, 0x1311}, + RuneRange{0x1316, 0x1317}, + RuneRange{0x131F, 0x131F}, + RuneRange{0x1347, 0x1347}, + RuneRange{0x135B, 0x1360}, + RuneRange{0x137D, 0x139F}, + RuneRange{0x13F5, 0x1400}, + RuneRange{0x1677, 0x167F}, + RuneRange{0x169D, 0x169F}, + RuneRange{0x16F1, 0x16FF}, + RuneRange{0x170D, 0x170D}, + RuneRange{0x1715, 0x171F}, + RuneRange{0x1737, 0x173F}, + RuneRange{0x1754, 0x175F}, + RuneRange{0x176D, 0x176D}, + RuneRange{0x1771, 0x1771}, + RuneRange{0x1774, 0x177F}, + RuneRange{0x17DD, 0x17DF}, + RuneRange{0x17EA, 0x17FF}, + RuneRange{0x180F, 0x180F}, + RuneRange{0x181A, 0x181F}, + RuneRange{0x1878, 0x187F}, + RuneRange{0x18AA, 0x1DFF}, + RuneRange{0x1E9C, 0x1E9F}, + RuneRange{0x1EFA, 0x1EFF}, + RuneRange{0x1F16, 0x1F17}, + RuneRange{0x1F1E, 0x1F1F}, + RuneRange{0x1F46, 0x1F47}, + RuneRange{0x1F4E, 0x1F4F}, + RuneRange{0x1F58, 0x1F58}, + RuneRange{0x1F5A, 0x1F5A}, + RuneRange{0x1F5C, 0x1F5C}, + RuneRange{0x1F5E, 0x1F5E}, + RuneRange{0x1F7E, 0x1F7F}, + RuneRange{0x1FB5, 0x1FB5}, + RuneRange{0x1FC5, 0x1FC5}, + RuneRange{0x1FD4, 0x1FD5}, + RuneRange{0x1FDC, 0x1FDC}, + RuneRange{0x1FF0, 0x1FF1}, + RuneRange{0x1FF5, 0x1FF5}, + RuneRange{0x1FFF, 0x1FFF}, + RuneRange{0x2053, 0x2056}, + RuneRange{0x2058, 0x205E}, + RuneRange{0x2064, 0x2069}, + RuneRange{0x2072, 0x2073}, + RuneRange{0x208F, 0x209F}, + RuneRange{0x20B2, 0x20CF}, + RuneRange{0x20EB, 0x20FF}, + RuneRange{0x213B, 0x213C}, + RuneRange{0x214C, 0x2152}, + RuneRange{0x2184, 0x218F}, + RuneRange{0x23CF, 0x23FF}, + RuneRange{0x2427, 0x243F}, + RuneRange{0x244B, 0x245F}, + RuneRange{0x24FF, 0x24FF}, + RuneRange{0x2614, 0x2615}, + RuneRange{0x2618, 0x2618}, + RuneRange{0x267E, 0x267F}, + RuneRange{0x268A, 0x2700}, + RuneRange{0x2705, 0x2705}, + RuneRange{0x270A, 0x270B}, + RuneRange{0x2728, 0x2728}, + RuneRange{0x274C, 0x274C}, + RuneRange{0x274E, 0x274E}, + RuneRange{0x2753, 0x2755}, + RuneRange{0x2757, 0x2757}, + RuneRange{0x275F, 0x2760}, + RuneRange{0x2795, 0x2797}, + RuneRange{0x27B0, 0x27B0}, + RuneRange{0x27BF, 0x27CF}, + RuneRange{0x27EC, 0x27EF}, + RuneRange{0x2B00, 0x2E7F}, + RuneRange{0x2E9A, 0x2E9A}, + RuneRange{0x2EF4, 0x2EFF}, + RuneRange{0x2FD6, 0x2FEF}, + RuneRange{0x2FFC, 0x2FFF}, + RuneRange{0x3040, 0x3040}, + RuneRange{0x3097, 0x3098}, + RuneRange{0x3100, 0x3104}, + RuneRange{0x312D, 0x3130}, + RuneRange{0x318F, 0x318F}, + RuneRange{0x31B8, 0x31EF}, + RuneRange{0x321D, 0x321F}, + RuneRange{0x3244, 0x3250}, + RuneRange{0x327C, 0x327E}, + RuneRange{0x32CC, 0x32CF}, + RuneRange{0x32FF, 0x32FF}, + RuneRange{0x3377, 0x337A}, + RuneRange{0x33DE, 0x33DF}, + RuneRange{0x33FF, 0x33FF}, + RuneRange{0x4DB6, 0x4DFF}, + RuneRange{0x9FA6, 0x9FFF}, + RuneRange{0xA48D, 0xA48F}, + RuneRange{0xA4C7, 0xABFF}, + RuneRange{0xD7A4, 0xD7FF}, + RuneRange{0xFA2E, 0xFA2F}, + RuneRange{0xFA6B, 0xFAFF}, + RuneRange{0xFB07, 0xFB12}, + RuneRange{0xFB18, 0xFB1C}, + RuneRange{0xFB37, 0xFB37}, + RuneRange{0xFB3D, 0xFB3D}, + RuneRange{0xFB3F, 0xFB3F}, + RuneRange{0xFB42, 0xFB42}, + RuneRange{0xFB45, 0xFB45}, + RuneRange{0xFBB2, 0xFBD2}, + RuneRange{0xFD40, 0xFD4F}, + RuneRange{0xFD90, 0xFD91}, + RuneRange{0xFDC8, 0xFDCF}, + RuneRange{0xFDFD, 0xFDFF}, + RuneRange{0xFE10, 0xFE1F}, + RuneRange{0xFE24, 0xFE2F}, + RuneRange{0xFE47, 0xFE48}, + RuneRange{0xFE53, 0xFE53}, + RuneRange{0xFE67, 0xFE67}, + RuneRange{0xFE6C, 0xFE6F}, + RuneRange{0xFE75, 0xFE75}, + RuneRange{0xFEFD, 0xFEFE}, + RuneRange{0xFF00, 0xFF00}, + RuneRange{0xFFBF, 0xFFC1}, + RuneRange{0xFFC8, 0xFFC9}, + RuneRange{0xFFD0, 0xFFD1}, + RuneRange{0xFFD8, 0xFFD9}, + RuneRange{0xFFDD, 0xFFDF}, + RuneRange{0xFFE7, 0xFFE7}, + RuneRange{0xFFEF, 0xFFF8}, + RuneRange{0x10000, 0x102FF}, + RuneRange{0x1031F, 0x1031F}, + RuneRange{0x10324, 0x1032F}, + RuneRange{0x1034B, 0x103FF}, + RuneRange{0x10426, 0x10427}, + RuneRange{0x1044E, 0x1CFFF}, + RuneRange{0x1D0F6, 0x1D0FF}, + RuneRange{0x1D127, 0x1D129}, + RuneRange{0x1D1DE, 0x1D3FF}, + RuneRange{0x1D455, 0x1D455}, + RuneRange{0x1D49D, 0x1D49D}, + RuneRange{0x1D4A0, 0x1D4A1}, + RuneRange{0x1D4A3, 0x1D4A4}, + RuneRange{0x1D4A7, 0x1D4A8}, + RuneRange{0x1D4AD, 0x1D4AD}, + RuneRange{0x1D4BA, 0x1D4BA}, + RuneRange{0x1D4BC, 0x1D4BC}, + RuneRange{0x1D4C1, 0x1D4C1}, + RuneRange{0x1D4C4, 0x1D4C4}, + RuneRange{0x1D506, 0x1D506}, + RuneRange{0x1D50B, 0x1D50C}, + RuneRange{0x1D515, 0x1D515}, + RuneRange{0x1D51D, 0x1D51D}, + RuneRange{0x1D53A, 0x1D53A}, + RuneRange{0x1D53F, 0x1D53F}, + RuneRange{0x1D545, 0x1D545}, + RuneRange{0x1D547, 0x1D549}, + RuneRange{0x1D551, 0x1D551}, + RuneRange{0x1D6A4, 0x1D6A7}, + RuneRange{0x1D7CA, 0x1D7CD}, + RuneRange{0x1D800, 0x1FFFD}, + RuneRange{0x2A6D7, 0x2F7FF}, + RuneRange{0x2FA1E, 0x2FFFD}, + RuneRange{0x30000, 0x3FFFD}, + RuneRange{0x40000, 0x4FFFD}, + RuneRange{0x50000, 0x5FFFD}, + RuneRange{0x60000, 0x6FFFD}, + RuneRange{0x70000, 0x7FFFD}, + RuneRange{0x80000, 0x8FFFD}, + RuneRange{0x90000, 0x9FFFD}, + RuneRange{0xA0000, 0xAFFFD}, + RuneRange{0xB0000, 0xBFFFD}, + RuneRange{0xC0000, 0xCFFFD}, + RuneRange{0xD0000, 0xDFFFD}, + RuneRange{0xE0000, 0xE0000}, + RuneRange{0xE0002, 0xE001F}, + RuneRange{0xE0080, 0xEFFFD}, +} + +// TableA1 represents RFC-3454 Table A.1. +var TableA1 Set = tableA1 + +var tableB1 = Mapping{ + 0x00AD: []rune{}, // Map to nothing + 0x034F: []rune{}, // Map to nothing + 0x180B: []rune{}, // Map to nothing + 0x180C: []rune{}, // Map to nothing + 0x180D: []rune{}, // Map to nothing + 0x200B: []rune{}, // Map to nothing + 0x200C: []rune{}, // Map to nothing + 0x200D: []rune{}, // Map to nothing + 0x2060: []rune{}, // Map to nothing + 0xFE00: []rune{}, // Map to nothing + 0xFE01: []rune{}, // Map to nothing + 0xFE02: []rune{}, // Map to nothing + 0xFE03: []rune{}, // Map to nothing + 0xFE04: []rune{}, // Map to nothing + 0xFE05: []rune{}, // Map to nothing + 0xFE06: []rune{}, // Map to nothing + 0xFE07: []rune{}, // Map to nothing + 0xFE08: []rune{}, // Map to nothing + 0xFE09: []rune{}, // Map to nothing + 0xFE0A: []rune{}, // Map to nothing + 0xFE0B: []rune{}, // Map to nothing + 0xFE0C: []rune{}, // Map to nothing + 0xFE0D: []rune{}, // Map to nothing + 0xFE0E: []rune{}, // Map to nothing + 0xFE0F: []rune{}, // Map to nothing + 0xFEFF: []rune{}, // Map to nothing +} + +// TableB1 represents RFC-3454 Table B.1. +var TableB1 Mapping = tableB1 + +var tableB2 = Mapping{ + 0x0041: []rune{0x0061}, // Case map + 0x0042: []rune{0x0062}, // Case map + 0x0043: []rune{0x0063}, // Case map + 0x0044: []rune{0x0064}, // Case map + 0x0045: []rune{0x0065}, // Case map + 0x0046: []rune{0x0066}, // Case map + 0x0047: []rune{0x0067}, // Case map + 0x0048: []rune{0x0068}, // Case map + 0x0049: []rune{0x0069}, // Case map + 0x004A: []rune{0x006A}, // Case map + 0x004B: []rune{0x006B}, // Case map + 0x004C: []rune{0x006C}, // Case map + 0x004D: []rune{0x006D}, // Case map + 0x004E: []rune{0x006E}, // Case map + 0x004F: []rune{0x006F}, // Case map + 0x0050: []rune{0x0070}, // Case map + 0x0051: []rune{0x0071}, // Case map + 0x0052: []rune{0x0072}, // Case map + 0x0053: []rune{0x0073}, // Case map + 0x0054: []rune{0x0074}, // Case map + 0x0055: []rune{0x0075}, // Case map + 0x0056: []rune{0x0076}, // Case map + 0x0057: []rune{0x0077}, // Case map + 0x0058: []rune{0x0078}, // Case map + 0x0059: []rune{0x0079}, // Case map + 0x005A: []rune{0x007A}, // Case map + 0x00B5: []rune{0x03BC}, // Case map + 0x00C0: []rune{0x00E0}, // Case map + 0x00C1: []rune{0x00E1}, // Case map + 0x00C2: []rune{0x00E2}, // Case map + 0x00C3: []rune{0x00E3}, // Case map + 0x00C4: []rune{0x00E4}, // Case map + 0x00C5: []rune{0x00E5}, // Case map + 0x00C6: []rune{0x00E6}, // Case map + 0x00C7: []rune{0x00E7}, // Case map + 0x00C8: []rune{0x00E8}, // Case map + 0x00C9: []rune{0x00E9}, // Case map + 0x00CA: []rune{0x00EA}, // Case map + 0x00CB: []rune{0x00EB}, // Case map + 0x00CC: []rune{0x00EC}, // Case map + 0x00CD: []rune{0x00ED}, // Case map + 0x00CE: []rune{0x00EE}, // Case map + 0x00CF: []rune{0x00EF}, // Case map + 0x00D0: []rune{0x00F0}, // Case map + 0x00D1: []rune{0x00F1}, // Case map + 0x00D2: []rune{0x00F2}, // Case map + 0x00D3: []rune{0x00F3}, // Case map + 0x00D4: []rune{0x00F4}, // Case map + 0x00D5: []rune{0x00F5}, // Case map + 0x00D6: []rune{0x00F6}, // Case map + 0x00D8: []rune{0x00F8}, // Case map + 0x00D9: []rune{0x00F9}, // Case map + 0x00DA: []rune{0x00FA}, // Case map + 0x00DB: []rune{0x00FB}, // Case map + 0x00DC: []rune{0x00FC}, // Case map + 0x00DD: []rune{0x00FD}, // Case map + 0x00DE: []rune{0x00FE}, // Case map + 0x00DF: []rune{0x0073, 0x0073}, // Case map + 0x0100: []rune{0x0101}, // Case map + 0x0102: []rune{0x0103}, // Case map + 0x0104: []rune{0x0105}, // Case map + 0x0106: []rune{0x0107}, // Case map + 0x0108: []rune{0x0109}, // Case map + 0x010A: []rune{0x010B}, // Case map + 0x010C: []rune{0x010D}, // Case map + 0x010E: []rune{0x010F}, // Case map + 0x0110: []rune{0x0111}, // Case map + 0x0112: []rune{0x0113}, // Case map + 0x0114: []rune{0x0115}, // Case map + 0x0116: []rune{0x0117}, // Case map + 0x0118: []rune{0x0119}, // Case map + 0x011A: []rune{0x011B}, // Case map + 0x011C: []rune{0x011D}, // Case map + 0x011E: []rune{0x011F}, // Case map + 0x0120: []rune{0x0121}, // Case map + 0x0122: []rune{0x0123}, // Case map + 0x0124: []rune{0x0125}, // Case map + 0x0126: []rune{0x0127}, // Case map + 0x0128: []rune{0x0129}, // Case map + 0x012A: []rune{0x012B}, // Case map + 0x012C: []rune{0x012D}, // Case map + 0x012E: []rune{0x012F}, // Case map + 0x0130: []rune{0x0069, 0x0307}, // Case map + 0x0132: []rune{0x0133}, // Case map + 0x0134: []rune{0x0135}, // Case map + 0x0136: []rune{0x0137}, // Case map + 0x0139: []rune{0x013A}, // Case map + 0x013B: []rune{0x013C}, // Case map + 0x013D: []rune{0x013E}, // Case map + 0x013F: []rune{0x0140}, // Case map + 0x0141: []rune{0x0142}, // Case map + 0x0143: []rune{0x0144}, // Case map + 0x0145: []rune{0x0146}, // Case map + 0x0147: []rune{0x0148}, // Case map + 0x0149: []rune{0x02BC, 0x006E}, // Case map + 0x014A: []rune{0x014B}, // Case map + 0x014C: []rune{0x014D}, // Case map + 0x014E: []rune{0x014F}, // Case map + 0x0150: []rune{0x0151}, // Case map + 0x0152: []rune{0x0153}, // Case map + 0x0154: []rune{0x0155}, // Case map + 0x0156: []rune{0x0157}, // Case map + 0x0158: []rune{0x0159}, // Case map + 0x015A: []rune{0x015B}, // Case map + 0x015C: []rune{0x015D}, // Case map + 0x015E: []rune{0x015F}, // Case map + 0x0160: []rune{0x0161}, // Case map + 0x0162: []rune{0x0163}, // Case map + 0x0164: []rune{0x0165}, // Case map + 0x0166: []rune{0x0167}, // Case map + 0x0168: []rune{0x0169}, // Case map + 0x016A: []rune{0x016B}, // Case map + 0x016C: []rune{0x016D}, // Case map + 0x016E: []rune{0x016F}, // Case map + 0x0170: []rune{0x0171}, // Case map + 0x0172: []rune{0x0173}, // Case map + 0x0174: []rune{0x0175}, // Case map + 0x0176: []rune{0x0177}, // Case map + 0x0178: []rune{0x00FF}, // Case map + 0x0179: []rune{0x017A}, // Case map + 0x017B: []rune{0x017C}, // Case map + 0x017D: []rune{0x017E}, // Case map + 0x017F: []rune{0x0073}, // Case map + 0x0181: []rune{0x0253}, // Case map + 0x0182: []rune{0x0183}, // Case map + 0x0184: []rune{0x0185}, // Case map + 0x0186: []rune{0x0254}, // Case map + 0x0187: []rune{0x0188}, // Case map + 0x0189: []rune{0x0256}, // Case map + 0x018A: []rune{0x0257}, // Case map + 0x018B: []rune{0x018C}, // Case map + 0x018E: []rune{0x01DD}, // Case map + 0x018F: []rune{0x0259}, // Case map + 0x0190: []rune{0x025B}, // Case map + 0x0191: []rune{0x0192}, // Case map + 0x0193: []rune{0x0260}, // Case map + 0x0194: []rune{0x0263}, // Case map + 0x0196: []rune{0x0269}, // Case map + 0x0197: []rune{0x0268}, // Case map + 0x0198: []rune{0x0199}, // Case map + 0x019C: []rune{0x026F}, // Case map + 0x019D: []rune{0x0272}, // Case map + 0x019F: []rune{0x0275}, // Case map + 0x01A0: []rune{0x01A1}, // Case map + 0x01A2: []rune{0x01A3}, // Case map + 0x01A4: []rune{0x01A5}, // Case map + 0x01A6: []rune{0x0280}, // Case map + 0x01A7: []rune{0x01A8}, // Case map + 0x01A9: []rune{0x0283}, // Case map + 0x01AC: []rune{0x01AD}, // Case map + 0x01AE: []rune{0x0288}, // Case map + 0x01AF: []rune{0x01B0}, // Case map + 0x01B1: []rune{0x028A}, // Case map + 0x01B2: []rune{0x028B}, // Case map + 0x01B3: []rune{0x01B4}, // Case map + 0x01B5: []rune{0x01B6}, // Case map + 0x01B7: []rune{0x0292}, // Case map + 0x01B8: []rune{0x01B9}, // Case map + 0x01BC: []rune{0x01BD}, // Case map + 0x01C4: []rune{0x01C6}, // Case map + 0x01C5: []rune{0x01C6}, // Case map + 0x01C7: []rune{0x01C9}, // Case map + 0x01C8: []rune{0x01C9}, // Case map + 0x01CA: []rune{0x01CC}, // Case map + 0x01CB: []rune{0x01CC}, // Case map + 0x01CD: []rune{0x01CE}, // Case map + 0x01CF: []rune{0x01D0}, // Case map + 0x01D1: []rune{0x01D2}, // Case map + 0x01D3: []rune{0x01D4}, // Case map + 0x01D5: []rune{0x01D6}, // Case map + 0x01D7: []rune{0x01D8}, // Case map + 0x01D9: []rune{0x01DA}, // Case map + 0x01DB: []rune{0x01DC}, // Case map + 0x01DE: []rune{0x01DF}, // Case map + 0x01E0: []rune{0x01E1}, // Case map + 0x01E2: []rune{0x01E3}, // Case map + 0x01E4: []rune{0x01E5}, // Case map + 0x01E6: []rune{0x01E7}, // Case map + 0x01E8: []rune{0x01E9}, // Case map + 0x01EA: []rune{0x01EB}, // Case map + 0x01EC: []rune{0x01ED}, // Case map + 0x01EE: []rune{0x01EF}, // Case map + 0x01F0: []rune{0x006A, 0x030C}, // Case map + 0x01F1: []rune{0x01F3}, // Case map + 0x01F2: []rune{0x01F3}, // Case map + 0x01F4: []rune{0x01F5}, // Case map + 0x01F6: []rune{0x0195}, // Case map + 0x01F7: []rune{0x01BF}, // Case map + 0x01F8: []rune{0x01F9}, // Case map + 0x01FA: []rune{0x01FB}, // Case map + 0x01FC: []rune{0x01FD}, // Case map + 0x01FE: []rune{0x01FF}, // Case map + 0x0200: []rune{0x0201}, // Case map + 0x0202: []rune{0x0203}, // Case map + 0x0204: []rune{0x0205}, // Case map + 0x0206: []rune{0x0207}, // Case map + 0x0208: []rune{0x0209}, // Case map + 0x020A: []rune{0x020B}, // Case map + 0x020C: []rune{0x020D}, // Case map + 0x020E: []rune{0x020F}, // Case map + 0x0210: []rune{0x0211}, // Case map + 0x0212: []rune{0x0213}, // Case map + 0x0214: []rune{0x0215}, // Case map + 0x0216: []rune{0x0217}, // Case map + 0x0218: []rune{0x0219}, // Case map + 0x021A: []rune{0x021B}, // Case map + 0x021C: []rune{0x021D}, // Case map + 0x021E: []rune{0x021F}, // Case map + 0x0220: []rune{0x019E}, // Case map + 0x0222: []rune{0x0223}, // Case map + 0x0224: []rune{0x0225}, // Case map + 0x0226: []rune{0x0227}, // Case map + 0x0228: []rune{0x0229}, // Case map + 0x022A: []rune{0x022B}, // Case map + 0x022C: []rune{0x022D}, // Case map + 0x022E: []rune{0x022F}, // Case map + 0x0230: []rune{0x0231}, // Case map + 0x0232: []rune{0x0233}, // Case map + 0x0345: []rune{0x03B9}, // Case map + 0x037A: []rune{0x0020, 0x03B9}, // Additional folding + 0x0386: []rune{0x03AC}, // Case map + 0x0388: []rune{0x03AD}, // Case map + 0x0389: []rune{0x03AE}, // Case map + 0x038A: []rune{0x03AF}, // Case map + 0x038C: []rune{0x03CC}, // Case map + 0x038E: []rune{0x03CD}, // Case map + 0x038F: []rune{0x03CE}, // Case map + 0x0390: []rune{0x03B9, 0x0308, 0x0301}, // Case map + 0x0391: []rune{0x03B1}, // Case map + 0x0392: []rune{0x03B2}, // Case map + 0x0393: []rune{0x03B3}, // Case map + 0x0394: []rune{0x03B4}, // Case map + 0x0395: []rune{0x03B5}, // Case map + 0x0396: []rune{0x03B6}, // Case map + 0x0397: []rune{0x03B7}, // Case map + 0x0398: []rune{0x03B8}, // Case map + 0x0399: []rune{0x03B9}, // Case map + 0x039A: []rune{0x03BA}, // Case map + 0x039B: []rune{0x03BB}, // Case map + 0x039C: []rune{0x03BC}, // Case map + 0x039D: []rune{0x03BD}, // Case map + 0x039E: []rune{0x03BE}, // Case map + 0x039F: []rune{0x03BF}, // Case map + 0x03A0: []rune{0x03C0}, // Case map + 0x03A1: []rune{0x03C1}, // Case map + 0x03A3: []rune{0x03C3}, // Case map + 0x03A4: []rune{0x03C4}, // Case map + 0x03A5: []rune{0x03C5}, // Case map + 0x03A6: []rune{0x03C6}, // Case map + 0x03A7: []rune{0x03C7}, // Case map + 0x03A8: []rune{0x03C8}, // Case map + 0x03A9: []rune{0x03C9}, // Case map + 0x03AA: []rune{0x03CA}, // Case map + 0x03AB: []rune{0x03CB}, // Case map + 0x03B0: []rune{0x03C5, 0x0308, 0x0301}, // Case map + 0x03C2: []rune{0x03C3}, // Case map + 0x03D0: []rune{0x03B2}, // Case map + 0x03D1: []rune{0x03B8}, // Case map + 0x03D2: []rune{0x03C5}, // Additional folding + 0x03D3: []rune{0x03CD}, // Additional folding + 0x03D4: []rune{0x03CB}, // Additional folding + 0x03D5: []rune{0x03C6}, // Case map + 0x03D6: []rune{0x03C0}, // Case map + 0x03D8: []rune{0x03D9}, // Case map + 0x03DA: []rune{0x03DB}, // Case map + 0x03DC: []rune{0x03DD}, // Case map + 0x03DE: []rune{0x03DF}, // Case map + 0x03E0: []rune{0x03E1}, // Case map + 0x03E2: []rune{0x03E3}, // Case map + 0x03E4: []rune{0x03E5}, // Case map + 0x03E6: []rune{0x03E7}, // Case map + 0x03E8: []rune{0x03E9}, // Case map + 0x03EA: []rune{0x03EB}, // Case map + 0x03EC: []rune{0x03ED}, // Case map + 0x03EE: []rune{0x03EF}, // Case map + 0x03F0: []rune{0x03BA}, // Case map + 0x03F1: []rune{0x03C1}, // Case map + 0x03F2: []rune{0x03C3}, // Case map + 0x03F4: []rune{0x03B8}, // Case map + 0x03F5: []rune{0x03B5}, // Case map + 0x0400: []rune{0x0450}, // Case map + 0x0401: []rune{0x0451}, // Case map + 0x0402: []rune{0x0452}, // Case map + 0x0403: []rune{0x0453}, // Case map + 0x0404: []rune{0x0454}, // Case map + 0x0405: []rune{0x0455}, // Case map + 0x0406: []rune{0x0456}, // Case map + 0x0407: []rune{0x0457}, // Case map + 0x0408: []rune{0x0458}, // Case map + 0x0409: []rune{0x0459}, // Case map + 0x040A: []rune{0x045A}, // Case map + 0x040B: []rune{0x045B}, // Case map + 0x040C: []rune{0x045C}, // Case map + 0x040D: []rune{0x045D}, // Case map + 0x040E: []rune{0x045E}, // Case map + 0x040F: []rune{0x045F}, // Case map + 0x0410: []rune{0x0430}, // Case map + 0x0411: []rune{0x0431}, // Case map + 0x0412: []rune{0x0432}, // Case map + 0x0413: []rune{0x0433}, // Case map + 0x0414: []rune{0x0434}, // Case map + 0x0415: []rune{0x0435}, // Case map + 0x0416: []rune{0x0436}, // Case map + 0x0417: []rune{0x0437}, // Case map + 0x0418: []rune{0x0438}, // Case map + 0x0419: []rune{0x0439}, // Case map + 0x041A: []rune{0x043A}, // Case map + 0x041B: []rune{0x043B}, // Case map + 0x041C: []rune{0x043C}, // Case map + 0x041D: []rune{0x043D}, // Case map + 0x041E: []rune{0x043E}, // Case map + 0x041F: []rune{0x043F}, // Case map + 0x0420: []rune{0x0440}, // Case map + 0x0421: []rune{0x0441}, // Case map + 0x0422: []rune{0x0442}, // Case map + 0x0423: []rune{0x0443}, // Case map + 0x0424: []rune{0x0444}, // Case map + 0x0425: []rune{0x0445}, // Case map + 0x0426: []rune{0x0446}, // Case map + 0x0427: []rune{0x0447}, // Case map + 0x0428: []rune{0x0448}, // Case map + 0x0429: []rune{0x0449}, // Case map + 0x042A: []rune{0x044A}, // Case map + 0x042B: []rune{0x044B}, // Case map + 0x042C: []rune{0x044C}, // Case map + 0x042D: []rune{0x044D}, // Case map + 0x042E: []rune{0x044E}, // Case map + 0x042F: []rune{0x044F}, // Case map + 0x0460: []rune{0x0461}, // Case map + 0x0462: []rune{0x0463}, // Case map + 0x0464: []rune{0x0465}, // Case map + 0x0466: []rune{0x0467}, // Case map + 0x0468: []rune{0x0469}, // Case map + 0x046A: []rune{0x046B}, // Case map + 0x046C: []rune{0x046D}, // Case map + 0x046E: []rune{0x046F}, // Case map + 0x0470: []rune{0x0471}, // Case map + 0x0472: []rune{0x0473}, // Case map + 0x0474: []rune{0x0475}, // Case map + 0x0476: []rune{0x0477}, // Case map + 0x0478: []rune{0x0479}, // Case map + 0x047A: []rune{0x047B}, // Case map + 0x047C: []rune{0x047D}, // Case map + 0x047E: []rune{0x047F}, // Case map + 0x0480: []rune{0x0481}, // Case map + 0x048A: []rune{0x048B}, // Case map + 0x048C: []rune{0x048D}, // Case map + 0x048E: []rune{0x048F}, // Case map + 0x0490: []rune{0x0491}, // Case map + 0x0492: []rune{0x0493}, // Case map + 0x0494: []rune{0x0495}, // Case map + 0x0496: []rune{0x0497}, // Case map + 0x0498: []rune{0x0499}, // Case map + 0x049A: []rune{0x049B}, // Case map + 0x049C: []rune{0x049D}, // Case map + 0x049E: []rune{0x049F}, // Case map + 0x04A0: []rune{0x04A1}, // Case map + 0x04A2: []rune{0x04A3}, // Case map + 0x04A4: []rune{0x04A5}, // Case map + 0x04A6: []rune{0x04A7}, // Case map + 0x04A8: []rune{0x04A9}, // Case map + 0x04AA: []rune{0x04AB}, // Case map + 0x04AC: []rune{0x04AD}, // Case map + 0x04AE: []rune{0x04AF}, // Case map + 0x04B0: []rune{0x04B1}, // Case map + 0x04B2: []rune{0x04B3}, // Case map + 0x04B4: []rune{0x04B5}, // Case map + 0x04B6: []rune{0x04B7}, // Case map + 0x04B8: []rune{0x04B9}, // Case map + 0x04BA: []rune{0x04BB}, // Case map + 0x04BC: []rune{0x04BD}, // Case map + 0x04BE: []rune{0x04BF}, // Case map + 0x04C1: []rune{0x04C2}, // Case map + 0x04C3: []rune{0x04C4}, // Case map + 0x04C5: []rune{0x04C6}, // Case map + 0x04C7: []rune{0x04C8}, // Case map + 0x04C9: []rune{0x04CA}, // Case map + 0x04CB: []rune{0x04CC}, // Case map + 0x04CD: []rune{0x04CE}, // Case map + 0x04D0: []rune{0x04D1}, // Case map + 0x04D2: []rune{0x04D3}, // Case map + 0x04D4: []rune{0x04D5}, // Case map + 0x04D6: []rune{0x04D7}, // Case map + 0x04D8: []rune{0x04D9}, // Case map + 0x04DA: []rune{0x04DB}, // Case map + 0x04DC: []rune{0x04DD}, // Case map + 0x04DE: []rune{0x04DF}, // Case map + 0x04E0: []rune{0x04E1}, // Case map + 0x04E2: []rune{0x04E3}, // Case map + 0x04E4: []rune{0x04E5}, // Case map + 0x04E6: []rune{0x04E7}, // Case map + 0x04E8: []rune{0x04E9}, // Case map + 0x04EA: []rune{0x04EB}, // Case map + 0x04EC: []rune{0x04ED}, // Case map + 0x04EE: []rune{0x04EF}, // Case map + 0x04F0: []rune{0x04F1}, // Case map + 0x04F2: []rune{0x04F3}, // Case map + 0x04F4: []rune{0x04F5}, // Case map + 0x04F8: []rune{0x04F9}, // Case map + 0x0500: []rune{0x0501}, // Case map + 0x0502: []rune{0x0503}, // Case map + 0x0504: []rune{0x0505}, // Case map + 0x0506: []rune{0x0507}, // Case map + 0x0508: []rune{0x0509}, // Case map + 0x050A: []rune{0x050B}, // Case map + 0x050C: []rune{0x050D}, // Case map + 0x050E: []rune{0x050F}, // Case map + 0x0531: []rune{0x0561}, // Case map + 0x0532: []rune{0x0562}, // Case map + 0x0533: []rune{0x0563}, // Case map + 0x0534: []rune{0x0564}, // Case map + 0x0535: []rune{0x0565}, // Case map + 0x0536: []rune{0x0566}, // Case map + 0x0537: []rune{0x0567}, // Case map + 0x0538: []rune{0x0568}, // Case map + 0x0539: []rune{0x0569}, // Case map + 0x053A: []rune{0x056A}, // Case map + 0x053B: []rune{0x056B}, // Case map + 0x053C: []rune{0x056C}, // Case map + 0x053D: []rune{0x056D}, // Case map + 0x053E: []rune{0x056E}, // Case map + 0x053F: []rune{0x056F}, // Case map + 0x0540: []rune{0x0570}, // Case map + 0x0541: []rune{0x0571}, // Case map + 0x0542: []rune{0x0572}, // Case map + 0x0543: []rune{0x0573}, // Case map + 0x0544: []rune{0x0574}, // Case map + 0x0545: []rune{0x0575}, // Case map + 0x0546: []rune{0x0576}, // Case map + 0x0547: []rune{0x0577}, // Case map + 0x0548: []rune{0x0578}, // Case map + 0x0549: []rune{0x0579}, // Case map + 0x054A: []rune{0x057A}, // Case map + 0x054B: []rune{0x057B}, // Case map + 0x054C: []rune{0x057C}, // Case map + 0x054D: []rune{0x057D}, // Case map + 0x054E: []rune{0x057E}, // Case map + 0x054F: []rune{0x057F}, // Case map + 0x0550: []rune{0x0580}, // Case map + 0x0551: []rune{0x0581}, // Case map + 0x0552: []rune{0x0582}, // Case map + 0x0553: []rune{0x0583}, // Case map + 0x0554: []rune{0x0584}, // Case map + 0x0555: []rune{0x0585}, // Case map + 0x0556: []rune{0x0586}, // Case map + 0x0587: []rune{0x0565, 0x0582}, // Case map + 0x1E00: []rune{0x1E01}, // Case map + 0x1E02: []rune{0x1E03}, // Case map + 0x1E04: []rune{0x1E05}, // Case map + 0x1E06: []rune{0x1E07}, // Case map + 0x1E08: []rune{0x1E09}, // Case map + 0x1E0A: []rune{0x1E0B}, // Case map + 0x1E0C: []rune{0x1E0D}, // Case map + 0x1E0E: []rune{0x1E0F}, // Case map + 0x1E10: []rune{0x1E11}, // Case map + 0x1E12: []rune{0x1E13}, // Case map + 0x1E14: []rune{0x1E15}, // Case map + 0x1E16: []rune{0x1E17}, // Case map + 0x1E18: []rune{0x1E19}, // Case map + 0x1E1A: []rune{0x1E1B}, // Case map + 0x1E1C: []rune{0x1E1D}, // Case map + 0x1E1E: []rune{0x1E1F}, // Case map + 0x1E20: []rune{0x1E21}, // Case map + 0x1E22: []rune{0x1E23}, // Case map + 0x1E24: []rune{0x1E25}, // Case map + 0x1E26: []rune{0x1E27}, // Case map + 0x1E28: []rune{0x1E29}, // Case map + 0x1E2A: []rune{0x1E2B}, // Case map + 0x1E2C: []rune{0x1E2D}, // Case map + 0x1E2E: []rune{0x1E2F}, // Case map + 0x1E30: []rune{0x1E31}, // Case map + 0x1E32: []rune{0x1E33}, // Case map + 0x1E34: []rune{0x1E35}, // Case map + 0x1E36: []rune{0x1E37}, // Case map + 0x1E38: []rune{0x1E39}, // Case map + 0x1E3A: []rune{0x1E3B}, // Case map + 0x1E3C: []rune{0x1E3D}, // Case map + 0x1E3E: []rune{0x1E3F}, // Case map + 0x1E40: []rune{0x1E41}, // Case map + 0x1E42: []rune{0x1E43}, // Case map + 0x1E44: []rune{0x1E45}, // Case map + 0x1E46: []rune{0x1E47}, // Case map + 0x1E48: []rune{0x1E49}, // Case map + 0x1E4A: []rune{0x1E4B}, // Case map + 0x1E4C: []rune{0x1E4D}, // Case map + 0x1E4E: []rune{0x1E4F}, // Case map + 0x1E50: []rune{0x1E51}, // Case map + 0x1E52: []rune{0x1E53}, // Case map + 0x1E54: []rune{0x1E55}, // Case map + 0x1E56: []rune{0x1E57}, // Case map + 0x1E58: []rune{0x1E59}, // Case map + 0x1E5A: []rune{0x1E5B}, // Case map + 0x1E5C: []rune{0x1E5D}, // Case map + 0x1E5E: []rune{0x1E5F}, // Case map + 0x1E60: []rune{0x1E61}, // Case map + 0x1E62: []rune{0x1E63}, // Case map + 0x1E64: []rune{0x1E65}, // Case map + 0x1E66: []rune{0x1E67}, // Case map + 0x1E68: []rune{0x1E69}, // Case map + 0x1E6A: []rune{0x1E6B}, // Case map + 0x1E6C: []rune{0x1E6D}, // Case map + 0x1E6E: []rune{0x1E6F}, // Case map + 0x1E70: []rune{0x1E71}, // Case map + 0x1E72: []rune{0x1E73}, // Case map + 0x1E74: []rune{0x1E75}, // Case map + 0x1E76: []rune{0x1E77}, // Case map + 0x1E78: []rune{0x1E79}, // Case map + 0x1E7A: []rune{0x1E7B}, // Case map + 0x1E7C: []rune{0x1E7D}, // Case map + 0x1E7E: []rune{0x1E7F}, // Case map + 0x1E80: []rune{0x1E81}, // Case map + 0x1E82: []rune{0x1E83}, // Case map + 0x1E84: []rune{0x1E85}, // Case map + 0x1E86: []rune{0x1E87}, // Case map + 0x1E88: []rune{0x1E89}, // Case map + 0x1E8A: []rune{0x1E8B}, // Case map + 0x1E8C: []rune{0x1E8D}, // Case map + 0x1E8E: []rune{0x1E8F}, // Case map + 0x1E90: []rune{0x1E91}, // Case map + 0x1E92: []rune{0x1E93}, // Case map + 0x1E94: []rune{0x1E95}, // Case map + 0x1E96: []rune{0x0068, 0x0331}, // Case map + 0x1E97: []rune{0x0074, 0x0308}, // Case map + 0x1E98: []rune{0x0077, 0x030A}, // Case map + 0x1E99: []rune{0x0079, 0x030A}, // Case map + 0x1E9A: []rune{0x0061, 0x02BE}, // Case map + 0x1E9B: []rune{0x1E61}, // Case map + 0x1EA0: []rune{0x1EA1}, // Case map + 0x1EA2: []rune{0x1EA3}, // Case map + 0x1EA4: []rune{0x1EA5}, // Case map + 0x1EA6: []rune{0x1EA7}, // Case map + 0x1EA8: []rune{0x1EA9}, // Case map + 0x1EAA: []rune{0x1EAB}, // Case map + 0x1EAC: []rune{0x1EAD}, // Case map + 0x1EAE: []rune{0x1EAF}, // Case map + 0x1EB0: []rune{0x1EB1}, // Case map + 0x1EB2: []rune{0x1EB3}, // Case map + 0x1EB4: []rune{0x1EB5}, // Case map + 0x1EB6: []rune{0x1EB7}, // Case map + 0x1EB8: []rune{0x1EB9}, // Case map + 0x1EBA: []rune{0x1EBB}, // Case map + 0x1EBC: []rune{0x1EBD}, // Case map + 0x1EBE: []rune{0x1EBF}, // Case map + 0x1EC0: []rune{0x1EC1}, // Case map + 0x1EC2: []rune{0x1EC3}, // Case map + 0x1EC4: []rune{0x1EC5}, // Case map + 0x1EC6: []rune{0x1EC7}, // Case map + 0x1EC8: []rune{0x1EC9}, // Case map + 0x1ECA: []rune{0x1ECB}, // Case map + 0x1ECC: []rune{0x1ECD}, // Case map + 0x1ECE: []rune{0x1ECF}, // Case map + 0x1ED0: []rune{0x1ED1}, // Case map + 0x1ED2: []rune{0x1ED3}, // Case map + 0x1ED4: []rune{0x1ED5}, // Case map + 0x1ED6: []rune{0x1ED7}, // Case map + 0x1ED8: []rune{0x1ED9}, // Case map + 0x1EDA: []rune{0x1EDB}, // Case map + 0x1EDC: []rune{0x1EDD}, // Case map + 0x1EDE: []rune{0x1EDF}, // Case map + 0x1EE0: []rune{0x1EE1}, // Case map + 0x1EE2: []rune{0x1EE3}, // Case map + 0x1EE4: []rune{0x1EE5}, // Case map + 0x1EE6: []rune{0x1EE7}, // Case map + 0x1EE8: []rune{0x1EE9}, // Case map + 0x1EEA: []rune{0x1EEB}, // Case map + 0x1EEC: []rune{0x1EED}, // Case map + 0x1EEE: []rune{0x1EEF}, // Case map + 0x1EF0: []rune{0x1EF1}, // Case map + 0x1EF2: []rune{0x1EF3}, // Case map + 0x1EF4: []rune{0x1EF5}, // Case map + 0x1EF6: []rune{0x1EF7}, // Case map + 0x1EF8: []rune{0x1EF9}, // Case map + 0x1F08: []rune{0x1F00}, // Case map + 0x1F09: []rune{0x1F01}, // Case map + 0x1F0A: []rune{0x1F02}, // Case map + 0x1F0B: []rune{0x1F03}, // Case map + 0x1F0C: []rune{0x1F04}, // Case map + 0x1F0D: []rune{0x1F05}, // Case map + 0x1F0E: []rune{0x1F06}, // Case map + 0x1F0F: []rune{0x1F07}, // Case map + 0x1F18: []rune{0x1F10}, // Case map + 0x1F19: []rune{0x1F11}, // Case map + 0x1F1A: []rune{0x1F12}, // Case map + 0x1F1B: []rune{0x1F13}, // Case map + 0x1F1C: []rune{0x1F14}, // Case map + 0x1F1D: []rune{0x1F15}, // Case map + 0x1F28: []rune{0x1F20}, // Case map + 0x1F29: []rune{0x1F21}, // Case map + 0x1F2A: []rune{0x1F22}, // Case map + 0x1F2B: []rune{0x1F23}, // Case map + 0x1F2C: []rune{0x1F24}, // Case map + 0x1F2D: []rune{0x1F25}, // Case map + 0x1F2E: []rune{0x1F26}, // Case map + 0x1F2F: []rune{0x1F27}, // Case map + 0x1F38: []rune{0x1F30}, // Case map + 0x1F39: []rune{0x1F31}, // Case map + 0x1F3A: []rune{0x1F32}, // Case map + 0x1F3B: []rune{0x1F33}, // Case map + 0x1F3C: []rune{0x1F34}, // Case map + 0x1F3D: []rune{0x1F35}, // Case map + 0x1F3E: []rune{0x1F36}, // Case map + 0x1F3F: []rune{0x1F37}, // Case map + 0x1F48: []rune{0x1F40}, // Case map + 0x1F49: []rune{0x1F41}, // Case map + 0x1F4A: []rune{0x1F42}, // Case map + 0x1F4B: []rune{0x1F43}, // Case map + 0x1F4C: []rune{0x1F44}, // Case map + 0x1F4D: []rune{0x1F45}, // Case map + 0x1F50: []rune{0x03C5, 0x0313}, // Case map + 0x1F52: []rune{0x03C5, 0x0313, 0x0300}, // Case map + 0x1F54: []rune{0x03C5, 0x0313, 0x0301}, // Case map + 0x1F56: []rune{0x03C5, 0x0313, 0x0342}, // Case map + 0x1F59: []rune{0x1F51}, // Case map + 0x1F5B: []rune{0x1F53}, // Case map + 0x1F5D: []rune{0x1F55}, // Case map + 0x1F5F: []rune{0x1F57}, // Case map + 0x1F68: []rune{0x1F60}, // Case map + 0x1F69: []rune{0x1F61}, // Case map + 0x1F6A: []rune{0x1F62}, // Case map + 0x1F6B: []rune{0x1F63}, // Case map + 0x1F6C: []rune{0x1F64}, // Case map + 0x1F6D: []rune{0x1F65}, // Case map + 0x1F6E: []rune{0x1F66}, // Case map + 0x1F6F: []rune{0x1F67}, // Case map + 0x1F80: []rune{0x1F00, 0x03B9}, // Case map + 0x1F81: []rune{0x1F01, 0x03B9}, // Case map + 0x1F82: []rune{0x1F02, 0x03B9}, // Case map + 0x1F83: []rune{0x1F03, 0x03B9}, // Case map + 0x1F84: []rune{0x1F04, 0x03B9}, // Case map + 0x1F85: []rune{0x1F05, 0x03B9}, // Case map + 0x1F86: []rune{0x1F06, 0x03B9}, // Case map + 0x1F87: []rune{0x1F07, 0x03B9}, // Case map + 0x1F88: []rune{0x1F00, 0x03B9}, // Case map + 0x1F89: []rune{0x1F01, 0x03B9}, // Case map + 0x1F8A: []rune{0x1F02, 0x03B9}, // Case map + 0x1F8B: []rune{0x1F03, 0x03B9}, // Case map + 0x1F8C: []rune{0x1F04, 0x03B9}, // Case map + 0x1F8D: []rune{0x1F05, 0x03B9}, // Case map + 0x1F8E: []rune{0x1F06, 0x03B9}, // Case map + 0x1F8F: []rune{0x1F07, 0x03B9}, // Case map + 0x1F90: []rune{0x1F20, 0x03B9}, // Case map + 0x1F91: []rune{0x1F21, 0x03B9}, // Case map + 0x1F92: []rune{0x1F22, 0x03B9}, // Case map + 0x1F93: []rune{0x1F23, 0x03B9}, // Case map + 0x1F94: []rune{0x1F24, 0x03B9}, // Case map + 0x1F95: []rune{0x1F25, 0x03B9}, // Case map + 0x1F96: []rune{0x1F26, 0x03B9}, // Case map + 0x1F97: []rune{0x1F27, 0x03B9}, // Case map + 0x1F98: []rune{0x1F20, 0x03B9}, // Case map + 0x1F99: []rune{0x1F21, 0x03B9}, // Case map + 0x1F9A: []rune{0x1F22, 0x03B9}, // Case map + 0x1F9B: []rune{0x1F23, 0x03B9}, // Case map + 0x1F9C: []rune{0x1F24, 0x03B9}, // Case map + 0x1F9D: []rune{0x1F25, 0x03B9}, // Case map + 0x1F9E: []rune{0x1F26, 0x03B9}, // Case map + 0x1F9F: []rune{0x1F27, 0x03B9}, // Case map + 0x1FA0: []rune{0x1F60, 0x03B9}, // Case map + 0x1FA1: []rune{0x1F61, 0x03B9}, // Case map + 0x1FA2: []rune{0x1F62, 0x03B9}, // Case map + 0x1FA3: []rune{0x1F63, 0x03B9}, // Case map + 0x1FA4: []rune{0x1F64, 0x03B9}, // Case map + 0x1FA5: []rune{0x1F65, 0x03B9}, // Case map + 0x1FA6: []rune{0x1F66, 0x03B9}, // Case map + 0x1FA7: []rune{0x1F67, 0x03B9}, // Case map + 0x1FA8: []rune{0x1F60, 0x03B9}, // Case map + 0x1FA9: []rune{0x1F61, 0x03B9}, // Case map + 0x1FAA: []rune{0x1F62, 0x03B9}, // Case map + 0x1FAB: []rune{0x1F63, 0x03B9}, // Case map + 0x1FAC: []rune{0x1F64, 0x03B9}, // Case map + 0x1FAD: []rune{0x1F65, 0x03B9}, // Case map + 0x1FAE: []rune{0x1F66, 0x03B9}, // Case map + 0x1FAF: []rune{0x1F67, 0x03B9}, // Case map + 0x1FB2: []rune{0x1F70, 0x03B9}, // Case map + 0x1FB3: []rune{0x03B1, 0x03B9}, // Case map + 0x1FB4: []rune{0x03AC, 0x03B9}, // Case map + 0x1FB6: []rune{0x03B1, 0x0342}, // Case map + 0x1FB7: []rune{0x03B1, 0x0342, 0x03B9}, // Case map + 0x1FB8: []rune{0x1FB0}, // Case map + 0x1FB9: []rune{0x1FB1}, // Case map + 0x1FBA: []rune{0x1F70}, // Case map + 0x1FBB: []rune{0x1F71}, // Case map + 0x1FBC: []rune{0x03B1, 0x03B9}, // Case map + 0x1FBE: []rune{0x03B9}, // Case map + 0x1FC2: []rune{0x1F74, 0x03B9}, // Case map + 0x1FC3: []rune{0x03B7, 0x03B9}, // Case map + 0x1FC4: []rune{0x03AE, 0x03B9}, // Case map + 0x1FC6: []rune{0x03B7, 0x0342}, // Case map + 0x1FC7: []rune{0x03B7, 0x0342, 0x03B9}, // Case map + 0x1FC8: []rune{0x1F72}, // Case map + 0x1FC9: []rune{0x1F73}, // Case map + 0x1FCA: []rune{0x1F74}, // Case map + 0x1FCB: []rune{0x1F75}, // Case map + 0x1FCC: []rune{0x03B7, 0x03B9}, // Case map + 0x1FD2: []rune{0x03B9, 0x0308, 0x0300}, // Case map + 0x1FD3: []rune{0x03B9, 0x0308, 0x0301}, // Case map + 0x1FD6: []rune{0x03B9, 0x0342}, // Case map + 0x1FD7: []rune{0x03B9, 0x0308, 0x0342}, // Case map + 0x1FD8: []rune{0x1FD0}, // Case map + 0x1FD9: []rune{0x1FD1}, // Case map + 0x1FDA: []rune{0x1F76}, // Case map + 0x1FDB: []rune{0x1F77}, // Case map + 0x1FE2: []rune{0x03C5, 0x0308, 0x0300}, // Case map + 0x1FE3: []rune{0x03C5, 0x0308, 0x0301}, // Case map + 0x1FE4: []rune{0x03C1, 0x0313}, // Case map + 0x1FE6: []rune{0x03C5, 0x0342}, // Case map + 0x1FE7: []rune{0x03C5, 0x0308, 0x0342}, // Case map + 0x1FE8: []rune{0x1FE0}, // Case map + 0x1FE9: []rune{0x1FE1}, // Case map + 0x1FEA: []rune{0x1F7A}, // Case map + 0x1FEB: []rune{0x1F7B}, // Case map + 0x1FEC: []rune{0x1FE5}, // Case map + 0x1FF2: []rune{0x1F7C, 0x03B9}, // Case map + 0x1FF3: []rune{0x03C9, 0x03B9}, // Case map + 0x1FF4: []rune{0x03CE, 0x03B9}, // Case map + 0x1FF6: []rune{0x03C9, 0x0342}, // Case map + 0x1FF7: []rune{0x03C9, 0x0342, 0x03B9}, // Case map + 0x1FF8: []rune{0x1F78}, // Case map + 0x1FF9: []rune{0x1F79}, // Case map + 0x1FFA: []rune{0x1F7C}, // Case map + 0x1FFB: []rune{0x1F7D}, // Case map + 0x1FFC: []rune{0x03C9, 0x03B9}, // Case map + 0x20A8: []rune{0x0072, 0x0073}, // Additional folding + 0x2102: []rune{0x0063}, // Additional folding + 0x2103: []rune{0x00B0, 0x0063}, // Additional folding + 0x2107: []rune{0x025B}, // Additional folding + 0x2109: []rune{0x00B0, 0x0066}, // Additional folding + 0x210B: []rune{0x0068}, // Additional folding + 0x210C: []rune{0x0068}, // Additional folding + 0x210D: []rune{0x0068}, // Additional folding + 0x2110: []rune{0x0069}, // Additional folding + 0x2111: []rune{0x0069}, // Additional folding + 0x2112: []rune{0x006C}, // Additional folding + 0x2115: []rune{0x006E}, // Additional folding + 0x2116: []rune{0x006E, 0x006F}, // Additional folding + 0x2119: []rune{0x0070}, // Additional folding + 0x211A: []rune{0x0071}, // Additional folding + 0x211B: []rune{0x0072}, // Additional folding + 0x211C: []rune{0x0072}, // Additional folding + 0x211D: []rune{0x0072}, // Additional folding + 0x2120: []rune{0x0073, 0x006D}, // Additional folding + 0x2121: []rune{0x0074, 0x0065, 0x006C}, // Additional folding + 0x2122: []rune{0x0074, 0x006D}, // Additional folding + 0x2124: []rune{0x007A}, // Additional folding + 0x2126: []rune{0x03C9}, // Case map + 0x2128: []rune{0x007A}, // Additional folding + 0x212A: []rune{0x006B}, // Case map + 0x212B: []rune{0x00E5}, // Case map + 0x212C: []rune{0x0062}, // Additional folding + 0x212D: []rune{0x0063}, // Additional folding + 0x2130: []rune{0x0065}, // Additional folding + 0x2131: []rune{0x0066}, // Additional folding + 0x2133: []rune{0x006D}, // Additional folding + 0x213E: []rune{0x03B3}, // Additional folding + 0x213F: []rune{0x03C0}, // Additional folding + 0x2145: []rune{0x0064}, // Additional folding + 0x2160: []rune{0x2170}, // Case map + 0x2161: []rune{0x2171}, // Case map + 0x2162: []rune{0x2172}, // Case map + 0x2163: []rune{0x2173}, // Case map + 0x2164: []rune{0x2174}, // Case map + 0x2165: []rune{0x2175}, // Case map + 0x2166: []rune{0x2176}, // Case map + 0x2167: []rune{0x2177}, // Case map + 0x2168: []rune{0x2178}, // Case map + 0x2169: []rune{0x2179}, // Case map + 0x216A: []rune{0x217A}, // Case map + 0x216B: []rune{0x217B}, // Case map + 0x216C: []rune{0x217C}, // Case map + 0x216D: []rune{0x217D}, // Case map + 0x216E: []rune{0x217E}, // Case map + 0x216F: []rune{0x217F}, // Case map + 0x24B6: []rune{0x24D0}, // Case map + 0x24B7: []rune{0x24D1}, // Case map + 0x24B8: []rune{0x24D2}, // Case map + 0x24B9: []rune{0x24D3}, // Case map + 0x24BA: []rune{0x24D4}, // Case map + 0x24BB: []rune{0x24D5}, // Case map + 0x24BC: []rune{0x24D6}, // Case map + 0x24BD: []rune{0x24D7}, // Case map + 0x24BE: []rune{0x24D8}, // Case map + 0x24BF: []rune{0x24D9}, // Case map + 0x24C0: []rune{0x24DA}, // Case map + 0x24C1: []rune{0x24DB}, // Case map + 0x24C2: []rune{0x24DC}, // Case map + 0x24C3: []rune{0x24DD}, // Case map + 0x24C4: []rune{0x24DE}, // Case map + 0x24C5: []rune{0x24DF}, // Case map + 0x24C6: []rune{0x24E0}, // Case map + 0x24C7: []rune{0x24E1}, // Case map + 0x24C8: []rune{0x24E2}, // Case map + 0x24C9: []rune{0x24E3}, // Case map + 0x24CA: []rune{0x24E4}, // Case map + 0x24CB: []rune{0x24E5}, // Case map + 0x24CC: []rune{0x24E6}, // Case map + 0x24CD: []rune{0x24E7}, // Case map + 0x24CE: []rune{0x24E8}, // Case map + 0x24CF: []rune{0x24E9}, // Case map + 0x3371: []rune{0x0068, 0x0070, 0x0061}, // Additional folding + 0x3373: []rune{0x0061, 0x0075}, // Additional folding + 0x3375: []rune{0x006F, 0x0076}, // Additional folding + 0x3380: []rune{0x0070, 0x0061}, // Additional folding + 0x3381: []rune{0x006E, 0x0061}, // Additional folding + 0x3382: []rune{0x03BC, 0x0061}, // Additional folding + 0x3383: []rune{0x006D, 0x0061}, // Additional folding + 0x3384: []rune{0x006B, 0x0061}, // Additional folding + 0x3385: []rune{0x006B, 0x0062}, // Additional folding + 0x3386: []rune{0x006D, 0x0062}, // Additional folding + 0x3387: []rune{0x0067, 0x0062}, // Additional folding + 0x338A: []rune{0x0070, 0x0066}, // Additional folding + 0x338B: []rune{0x006E, 0x0066}, // Additional folding + 0x338C: []rune{0x03BC, 0x0066}, // Additional folding + 0x3390: []rune{0x0068, 0x007A}, // Additional folding + 0x3391: []rune{0x006B, 0x0068, 0x007A}, // Additional folding + 0x3392: []rune{0x006D, 0x0068, 0x007A}, // Additional folding + 0x3393: []rune{0x0067, 0x0068, 0x007A}, // Additional folding + 0x3394: []rune{0x0074, 0x0068, 0x007A}, // Additional folding + 0x33A9: []rune{0x0070, 0x0061}, // Additional folding + 0x33AA: []rune{0x006B, 0x0070, 0x0061}, // Additional folding + 0x33AB: []rune{0x006D, 0x0070, 0x0061}, // Additional folding + 0x33AC: []rune{0x0067, 0x0070, 0x0061}, // Additional folding + 0x33B4: []rune{0x0070, 0x0076}, // Additional folding + 0x33B5: []rune{0x006E, 0x0076}, // Additional folding + 0x33B6: []rune{0x03BC, 0x0076}, // Additional folding + 0x33B7: []rune{0x006D, 0x0076}, // Additional folding + 0x33B8: []rune{0x006B, 0x0076}, // Additional folding + 0x33B9: []rune{0x006D, 0x0076}, // Additional folding + 0x33BA: []rune{0x0070, 0x0077}, // Additional folding + 0x33BB: []rune{0x006E, 0x0077}, // Additional folding + 0x33BC: []rune{0x03BC, 0x0077}, // Additional folding + 0x33BD: []rune{0x006D, 0x0077}, // Additional folding + 0x33BE: []rune{0x006B, 0x0077}, // Additional folding + 0x33BF: []rune{0x006D, 0x0077}, // Additional folding + 0x33C0: []rune{0x006B, 0x03C9}, // Additional folding + 0x33C1: []rune{0x006D, 0x03C9}, // Additional folding + 0x33C3: []rune{0x0062, 0x0071}, // Additional folding + 0x33C6: []rune{0x0063, 0x2215, 0x006B, 0x0067}, // Additional folding + 0x33C7: []rune{0x0063, 0x006F, 0x002E}, // Additional folding + 0x33C8: []rune{0x0064, 0x0062}, // Additional folding + 0x33C9: []rune{0x0067, 0x0079}, // Additional folding + 0x33CB: []rune{0x0068, 0x0070}, // Additional folding + 0x33CD: []rune{0x006B, 0x006B}, // Additional folding + 0x33CE: []rune{0x006B, 0x006D}, // Additional folding + 0x33D7: []rune{0x0070, 0x0068}, // Additional folding + 0x33D9: []rune{0x0070, 0x0070, 0x006D}, // Additional folding + 0x33DA: []rune{0x0070, 0x0072}, // Additional folding + 0x33DC: []rune{0x0073, 0x0076}, // Additional folding + 0x33DD: []rune{0x0077, 0x0062}, // Additional folding + 0xFB00: []rune{0x0066, 0x0066}, // Case map + 0xFB01: []rune{0x0066, 0x0069}, // Case map + 0xFB02: []rune{0x0066, 0x006C}, // Case map + 0xFB03: []rune{0x0066, 0x0066, 0x0069}, // Case map + 0xFB04: []rune{0x0066, 0x0066, 0x006C}, // Case map + 0xFB05: []rune{0x0073, 0x0074}, // Case map + 0xFB06: []rune{0x0073, 0x0074}, // Case map + 0xFB13: []rune{0x0574, 0x0576}, // Case map + 0xFB14: []rune{0x0574, 0x0565}, // Case map + 0xFB15: []rune{0x0574, 0x056B}, // Case map + 0xFB16: []rune{0x057E, 0x0576}, // Case map + 0xFB17: []rune{0x0574, 0x056D}, // Case map + 0xFF21: []rune{0xFF41}, // Case map + 0xFF22: []rune{0xFF42}, // Case map + 0xFF23: []rune{0xFF43}, // Case map + 0xFF24: []rune{0xFF44}, // Case map + 0xFF25: []rune{0xFF45}, // Case map + 0xFF26: []rune{0xFF46}, // Case map + 0xFF27: []rune{0xFF47}, // Case map + 0xFF28: []rune{0xFF48}, // Case map + 0xFF29: []rune{0xFF49}, // Case map + 0xFF2A: []rune{0xFF4A}, // Case map + 0xFF2B: []rune{0xFF4B}, // Case map + 0xFF2C: []rune{0xFF4C}, // Case map + 0xFF2D: []rune{0xFF4D}, // Case map + 0xFF2E: []rune{0xFF4E}, // Case map + 0xFF2F: []rune{0xFF4F}, // Case map + 0xFF30: []rune{0xFF50}, // Case map + 0xFF31: []rune{0xFF51}, // Case map + 0xFF32: []rune{0xFF52}, // Case map + 0xFF33: []rune{0xFF53}, // Case map + 0xFF34: []rune{0xFF54}, // Case map + 0xFF35: []rune{0xFF55}, // Case map + 0xFF36: []rune{0xFF56}, // Case map + 0xFF37: []rune{0xFF57}, // Case map + 0xFF38: []rune{0xFF58}, // Case map + 0xFF39: []rune{0xFF59}, // Case map + 0xFF3A: []rune{0xFF5A}, // Case map + 0x10400: []rune{0x10428}, // Case map + 0x10401: []rune{0x10429}, // Case map + 0x10402: []rune{0x1042A}, // Case map + 0x10403: []rune{0x1042B}, // Case map + 0x10404: []rune{0x1042C}, // Case map + 0x10405: []rune{0x1042D}, // Case map + 0x10406: []rune{0x1042E}, // Case map + 0x10407: []rune{0x1042F}, // Case map + 0x10408: []rune{0x10430}, // Case map + 0x10409: []rune{0x10431}, // Case map + 0x1040A: []rune{0x10432}, // Case map + 0x1040B: []rune{0x10433}, // Case map + 0x1040C: []rune{0x10434}, // Case map + 0x1040D: []rune{0x10435}, // Case map + 0x1040E: []rune{0x10436}, // Case map + 0x1040F: []rune{0x10437}, // Case map + 0x10410: []rune{0x10438}, // Case map + 0x10411: []rune{0x10439}, // Case map + 0x10412: []rune{0x1043A}, // Case map + 0x10413: []rune{0x1043B}, // Case map + 0x10414: []rune{0x1043C}, // Case map + 0x10415: []rune{0x1043D}, // Case map + 0x10416: []rune{0x1043E}, // Case map + 0x10417: []rune{0x1043F}, // Case map + 0x10418: []rune{0x10440}, // Case map + 0x10419: []rune{0x10441}, // Case map + 0x1041A: []rune{0x10442}, // Case map + 0x1041B: []rune{0x10443}, // Case map + 0x1041C: []rune{0x10444}, // Case map + 0x1041D: []rune{0x10445}, // Case map + 0x1041E: []rune{0x10446}, // Case map + 0x1041F: []rune{0x10447}, // Case map + 0x10420: []rune{0x10448}, // Case map + 0x10421: []rune{0x10449}, // Case map + 0x10422: []rune{0x1044A}, // Case map + 0x10423: []rune{0x1044B}, // Case map + 0x10424: []rune{0x1044C}, // Case map + 0x10425: []rune{0x1044D}, // Case map + 0x1D400: []rune{0x0061}, // Additional folding + 0x1D401: []rune{0x0062}, // Additional folding + 0x1D402: []rune{0x0063}, // Additional folding + 0x1D403: []rune{0x0064}, // Additional folding + 0x1D404: []rune{0x0065}, // Additional folding + 0x1D405: []rune{0x0066}, // Additional folding + 0x1D406: []rune{0x0067}, // Additional folding + 0x1D407: []rune{0x0068}, // Additional folding + 0x1D408: []rune{0x0069}, // Additional folding + 0x1D409: []rune{0x006A}, // Additional folding + 0x1D40A: []rune{0x006B}, // Additional folding + 0x1D40B: []rune{0x006C}, // Additional folding + 0x1D40C: []rune{0x006D}, // Additional folding + 0x1D40D: []rune{0x006E}, // Additional folding + 0x1D40E: []rune{0x006F}, // Additional folding + 0x1D40F: []rune{0x0070}, // Additional folding + 0x1D410: []rune{0x0071}, // Additional folding + 0x1D411: []rune{0x0072}, // Additional folding + 0x1D412: []rune{0x0073}, // Additional folding + 0x1D413: []rune{0x0074}, // Additional folding + 0x1D414: []rune{0x0075}, // Additional folding + 0x1D415: []rune{0x0076}, // Additional folding + 0x1D416: []rune{0x0077}, // Additional folding + 0x1D417: []rune{0x0078}, // Additional folding + 0x1D418: []rune{0x0079}, // Additional folding + 0x1D419: []rune{0x007A}, // Additional folding + 0x1D434: []rune{0x0061}, // Additional folding + 0x1D435: []rune{0x0062}, // Additional folding + 0x1D436: []rune{0x0063}, // Additional folding + 0x1D437: []rune{0x0064}, // Additional folding + 0x1D438: []rune{0x0065}, // Additional folding + 0x1D439: []rune{0x0066}, // Additional folding + 0x1D43A: []rune{0x0067}, // Additional folding + 0x1D43B: []rune{0x0068}, // Additional folding + 0x1D43C: []rune{0x0069}, // Additional folding + 0x1D43D: []rune{0x006A}, // Additional folding + 0x1D43E: []rune{0x006B}, // Additional folding + 0x1D43F: []rune{0x006C}, // Additional folding + 0x1D440: []rune{0x006D}, // Additional folding + 0x1D441: []rune{0x006E}, // Additional folding + 0x1D442: []rune{0x006F}, // Additional folding + 0x1D443: []rune{0x0070}, // Additional folding + 0x1D444: []rune{0x0071}, // Additional folding + 0x1D445: []rune{0x0072}, // Additional folding + 0x1D446: []rune{0x0073}, // Additional folding + 0x1D447: []rune{0x0074}, // Additional folding + 0x1D448: []rune{0x0075}, // Additional folding + 0x1D449: []rune{0x0076}, // Additional folding + 0x1D44A: []rune{0x0077}, // Additional folding + 0x1D44B: []rune{0x0078}, // Additional folding + 0x1D44C: []rune{0x0079}, // Additional folding + 0x1D44D: []rune{0x007A}, // Additional folding + 0x1D468: []rune{0x0061}, // Additional folding + 0x1D469: []rune{0x0062}, // Additional folding + 0x1D46A: []rune{0x0063}, // Additional folding + 0x1D46B: []rune{0x0064}, // Additional folding + 0x1D46C: []rune{0x0065}, // Additional folding + 0x1D46D: []rune{0x0066}, // Additional folding + 0x1D46E: []rune{0x0067}, // Additional folding + 0x1D46F: []rune{0x0068}, // Additional folding + 0x1D470: []rune{0x0069}, // Additional folding + 0x1D471: []rune{0x006A}, // Additional folding + 0x1D472: []rune{0x006B}, // Additional folding + 0x1D473: []rune{0x006C}, // Additional folding + 0x1D474: []rune{0x006D}, // Additional folding + 0x1D475: []rune{0x006E}, // Additional folding + 0x1D476: []rune{0x006F}, // Additional folding + 0x1D477: []rune{0x0070}, // Additional folding + 0x1D478: []rune{0x0071}, // Additional folding + 0x1D479: []rune{0x0072}, // Additional folding + 0x1D47A: []rune{0x0073}, // Additional folding + 0x1D47B: []rune{0x0074}, // Additional folding + 0x1D47C: []rune{0x0075}, // Additional folding + 0x1D47D: []rune{0x0076}, // Additional folding + 0x1D47E: []rune{0x0077}, // Additional folding + 0x1D47F: []rune{0x0078}, // Additional folding + 0x1D480: []rune{0x0079}, // Additional folding + 0x1D481: []rune{0x007A}, // Additional folding + 0x1D49C: []rune{0x0061}, // Additional folding + 0x1D49E: []rune{0x0063}, // Additional folding + 0x1D49F: []rune{0x0064}, // Additional folding + 0x1D4A2: []rune{0x0067}, // Additional folding + 0x1D4A5: []rune{0x006A}, // Additional folding + 0x1D4A6: []rune{0x006B}, // Additional folding + 0x1D4A9: []rune{0x006E}, // Additional folding + 0x1D4AA: []rune{0x006F}, // Additional folding + 0x1D4AB: []rune{0x0070}, // Additional folding + 0x1D4AC: []rune{0x0071}, // Additional folding + 0x1D4AE: []rune{0x0073}, // Additional folding + 0x1D4AF: []rune{0x0074}, // Additional folding + 0x1D4B0: []rune{0x0075}, // Additional folding + 0x1D4B1: []rune{0x0076}, // Additional folding + 0x1D4B2: []rune{0x0077}, // Additional folding + 0x1D4B3: []rune{0x0078}, // Additional folding + 0x1D4B4: []rune{0x0079}, // Additional folding + 0x1D4B5: []rune{0x007A}, // Additional folding + 0x1D4D0: []rune{0x0061}, // Additional folding + 0x1D4D1: []rune{0x0062}, // Additional folding + 0x1D4D2: []rune{0x0063}, // Additional folding + 0x1D4D3: []rune{0x0064}, // Additional folding + 0x1D4D4: []rune{0x0065}, // Additional folding + 0x1D4D5: []rune{0x0066}, // Additional folding + 0x1D4D6: []rune{0x0067}, // Additional folding + 0x1D4D7: []rune{0x0068}, // Additional folding + 0x1D4D8: []rune{0x0069}, // Additional folding + 0x1D4D9: []rune{0x006A}, // Additional folding + 0x1D4DA: []rune{0x006B}, // Additional folding + 0x1D4DB: []rune{0x006C}, // Additional folding + 0x1D4DC: []rune{0x006D}, // Additional folding + 0x1D4DD: []rune{0x006E}, // Additional folding + 0x1D4DE: []rune{0x006F}, // Additional folding + 0x1D4DF: []rune{0x0070}, // Additional folding + 0x1D4E0: []rune{0x0071}, // Additional folding + 0x1D4E1: []rune{0x0072}, // Additional folding + 0x1D4E2: []rune{0x0073}, // Additional folding + 0x1D4E3: []rune{0x0074}, // Additional folding + 0x1D4E4: []rune{0x0075}, // Additional folding + 0x1D4E5: []rune{0x0076}, // Additional folding + 0x1D4E6: []rune{0x0077}, // Additional folding + 0x1D4E7: []rune{0x0078}, // Additional folding + 0x1D4E8: []rune{0x0079}, // Additional folding + 0x1D4E9: []rune{0x007A}, // Additional folding + 0x1D504: []rune{0x0061}, // Additional folding + 0x1D505: []rune{0x0062}, // Additional folding + 0x1D507: []rune{0x0064}, // Additional folding + 0x1D508: []rune{0x0065}, // Additional folding + 0x1D509: []rune{0x0066}, // Additional folding + 0x1D50A: []rune{0x0067}, // Additional folding + 0x1D50D: []rune{0x006A}, // Additional folding + 0x1D50E: []rune{0x006B}, // Additional folding + 0x1D50F: []rune{0x006C}, // Additional folding + 0x1D510: []rune{0x006D}, // Additional folding + 0x1D511: []rune{0x006E}, // Additional folding + 0x1D512: []rune{0x006F}, // Additional folding + 0x1D513: []rune{0x0070}, // Additional folding + 0x1D514: []rune{0x0071}, // Additional folding + 0x1D516: []rune{0x0073}, // Additional folding + 0x1D517: []rune{0x0074}, // Additional folding + 0x1D518: []rune{0x0075}, // Additional folding + 0x1D519: []rune{0x0076}, // Additional folding + 0x1D51A: []rune{0x0077}, // Additional folding + 0x1D51B: []rune{0x0078}, // Additional folding + 0x1D51C: []rune{0x0079}, // Additional folding + 0x1D538: []rune{0x0061}, // Additional folding + 0x1D539: []rune{0x0062}, // Additional folding + 0x1D53B: []rune{0x0064}, // Additional folding + 0x1D53C: []rune{0x0065}, // Additional folding + 0x1D53D: []rune{0x0066}, // Additional folding + 0x1D53E: []rune{0x0067}, // Additional folding + 0x1D540: []rune{0x0069}, // Additional folding + 0x1D541: []rune{0x006A}, // Additional folding + 0x1D542: []rune{0x006B}, // Additional folding + 0x1D543: []rune{0x006C}, // Additional folding + 0x1D544: []rune{0x006D}, // Additional folding + 0x1D546: []rune{0x006F}, // Additional folding + 0x1D54A: []rune{0x0073}, // Additional folding + 0x1D54B: []rune{0x0074}, // Additional folding + 0x1D54C: []rune{0x0075}, // Additional folding + 0x1D54D: []rune{0x0076}, // Additional folding + 0x1D54E: []rune{0x0077}, // Additional folding + 0x1D54F: []rune{0x0078}, // Additional folding + 0x1D550: []rune{0x0079}, // Additional folding + 0x1D56C: []rune{0x0061}, // Additional folding + 0x1D56D: []rune{0x0062}, // Additional folding + 0x1D56E: []rune{0x0063}, // Additional folding + 0x1D56F: []rune{0x0064}, // Additional folding + 0x1D570: []rune{0x0065}, // Additional folding + 0x1D571: []rune{0x0066}, // Additional folding + 0x1D572: []rune{0x0067}, // Additional folding + 0x1D573: []rune{0x0068}, // Additional folding + 0x1D574: []rune{0x0069}, // Additional folding + 0x1D575: []rune{0x006A}, // Additional folding + 0x1D576: []rune{0x006B}, // Additional folding + 0x1D577: []rune{0x006C}, // Additional folding + 0x1D578: []rune{0x006D}, // Additional folding + 0x1D579: []rune{0x006E}, // Additional folding + 0x1D57A: []rune{0x006F}, // Additional folding + 0x1D57B: []rune{0x0070}, // Additional folding + 0x1D57C: []rune{0x0071}, // Additional folding + 0x1D57D: []rune{0x0072}, // Additional folding + 0x1D57E: []rune{0x0073}, // Additional folding + 0x1D57F: []rune{0x0074}, // Additional folding + 0x1D580: []rune{0x0075}, // Additional folding + 0x1D581: []rune{0x0076}, // Additional folding + 0x1D582: []rune{0x0077}, // Additional folding + 0x1D583: []rune{0x0078}, // Additional folding + 0x1D584: []rune{0x0079}, // Additional folding + 0x1D585: []rune{0x007A}, // Additional folding + 0x1D5A0: []rune{0x0061}, // Additional folding + 0x1D5A1: []rune{0x0062}, // Additional folding + 0x1D5A2: []rune{0x0063}, // Additional folding + 0x1D5A3: []rune{0x0064}, // Additional folding + 0x1D5A4: []rune{0x0065}, // Additional folding + 0x1D5A5: []rune{0x0066}, // Additional folding + 0x1D5A6: []rune{0x0067}, // Additional folding + 0x1D5A7: []rune{0x0068}, // Additional folding + 0x1D5A8: []rune{0x0069}, // Additional folding + 0x1D5A9: []rune{0x006A}, // Additional folding + 0x1D5AA: []rune{0x006B}, // Additional folding + 0x1D5AB: []rune{0x006C}, // Additional folding + 0x1D5AC: []rune{0x006D}, // Additional folding + 0x1D5AD: []rune{0x006E}, // Additional folding + 0x1D5AE: []rune{0x006F}, // Additional folding + 0x1D5AF: []rune{0x0070}, // Additional folding + 0x1D5B0: []rune{0x0071}, // Additional folding + 0x1D5B1: []rune{0x0072}, // Additional folding + 0x1D5B2: []rune{0x0073}, // Additional folding + 0x1D5B3: []rune{0x0074}, // Additional folding + 0x1D5B4: []rune{0x0075}, // Additional folding + 0x1D5B5: []rune{0x0076}, // Additional folding + 0x1D5B6: []rune{0x0077}, // Additional folding + 0x1D5B7: []rune{0x0078}, // Additional folding + 0x1D5B8: []rune{0x0079}, // Additional folding + 0x1D5B9: []rune{0x007A}, // Additional folding + 0x1D5D4: []rune{0x0061}, // Additional folding + 0x1D5D5: []rune{0x0062}, // Additional folding + 0x1D5D6: []rune{0x0063}, // Additional folding + 0x1D5D7: []rune{0x0064}, // Additional folding + 0x1D5D8: []rune{0x0065}, // Additional folding + 0x1D5D9: []rune{0x0066}, // Additional folding + 0x1D5DA: []rune{0x0067}, // Additional folding + 0x1D5DB: []rune{0x0068}, // Additional folding + 0x1D5DC: []rune{0x0069}, // Additional folding + 0x1D5DD: []rune{0x006A}, // Additional folding + 0x1D5DE: []rune{0x006B}, // Additional folding + 0x1D5DF: []rune{0x006C}, // Additional folding + 0x1D5E0: []rune{0x006D}, // Additional folding + 0x1D5E1: []rune{0x006E}, // Additional folding + 0x1D5E2: []rune{0x006F}, // Additional folding + 0x1D5E3: []rune{0x0070}, // Additional folding + 0x1D5E4: []rune{0x0071}, // Additional folding + 0x1D5E5: []rune{0x0072}, // Additional folding + 0x1D5E6: []rune{0x0073}, // Additional folding + 0x1D5E7: []rune{0x0074}, // Additional folding + 0x1D5E8: []rune{0x0075}, // Additional folding + 0x1D5E9: []rune{0x0076}, // Additional folding + 0x1D5EA: []rune{0x0077}, // Additional folding + 0x1D5EB: []rune{0x0078}, // Additional folding + 0x1D5EC: []rune{0x0079}, // Additional folding + 0x1D5ED: []rune{0x007A}, // Additional folding + 0x1D608: []rune{0x0061}, // Additional folding + 0x1D609: []rune{0x0062}, // Additional folding + 0x1D60A: []rune{0x0063}, // Additional folding + 0x1D60B: []rune{0x0064}, // Additional folding + 0x1D60C: []rune{0x0065}, // Additional folding + 0x1D60D: []rune{0x0066}, // Additional folding + 0x1D60E: []rune{0x0067}, // Additional folding + 0x1D60F: []rune{0x0068}, // Additional folding + 0x1D610: []rune{0x0069}, // Additional folding + 0x1D611: []rune{0x006A}, // Additional folding + 0x1D612: []rune{0x006B}, // Additional folding + 0x1D613: []rune{0x006C}, // Additional folding + 0x1D614: []rune{0x006D}, // Additional folding + 0x1D615: []rune{0x006E}, // Additional folding + 0x1D616: []rune{0x006F}, // Additional folding + 0x1D617: []rune{0x0070}, // Additional folding + 0x1D618: []rune{0x0071}, // Additional folding + 0x1D619: []rune{0x0072}, // Additional folding + 0x1D61A: []rune{0x0073}, // Additional folding + 0x1D61B: []rune{0x0074}, // Additional folding + 0x1D61C: []rune{0x0075}, // Additional folding + 0x1D61D: []rune{0x0076}, // Additional folding + 0x1D61E: []rune{0x0077}, // Additional folding + 0x1D61F: []rune{0x0078}, // Additional folding + 0x1D620: []rune{0x0079}, // Additional folding + 0x1D621: []rune{0x007A}, // Additional folding + 0x1D63C: []rune{0x0061}, // Additional folding + 0x1D63D: []rune{0x0062}, // Additional folding + 0x1D63E: []rune{0x0063}, // Additional folding + 0x1D63F: []rune{0x0064}, // Additional folding + 0x1D640: []rune{0x0065}, // Additional folding + 0x1D641: []rune{0x0066}, // Additional folding + 0x1D642: []rune{0x0067}, // Additional folding + 0x1D643: []rune{0x0068}, // Additional folding + 0x1D644: []rune{0x0069}, // Additional folding + 0x1D645: []rune{0x006A}, // Additional folding + 0x1D646: []rune{0x006B}, // Additional folding + 0x1D647: []rune{0x006C}, // Additional folding + 0x1D648: []rune{0x006D}, // Additional folding + 0x1D649: []rune{0x006E}, // Additional folding + 0x1D64A: []rune{0x006F}, // Additional folding + 0x1D64B: []rune{0x0070}, // Additional folding + 0x1D64C: []rune{0x0071}, // Additional folding + 0x1D64D: []rune{0x0072}, // Additional folding + 0x1D64E: []rune{0x0073}, // Additional folding + 0x1D64F: []rune{0x0074}, // Additional folding + 0x1D650: []rune{0x0075}, // Additional folding + 0x1D651: []rune{0x0076}, // Additional folding + 0x1D652: []rune{0x0077}, // Additional folding + 0x1D653: []rune{0x0078}, // Additional folding + 0x1D654: []rune{0x0079}, // Additional folding + 0x1D655: []rune{0x007A}, // Additional folding + 0x1D670: []rune{0x0061}, // Additional folding + 0x1D671: []rune{0x0062}, // Additional folding + 0x1D672: []rune{0x0063}, // Additional folding + 0x1D673: []rune{0x0064}, // Additional folding + 0x1D674: []rune{0x0065}, // Additional folding + 0x1D675: []rune{0x0066}, // Additional folding + 0x1D676: []rune{0x0067}, // Additional folding + 0x1D677: []rune{0x0068}, // Additional folding + 0x1D678: []rune{0x0069}, // Additional folding + 0x1D679: []rune{0x006A}, // Additional folding + 0x1D67A: []rune{0x006B}, // Additional folding + 0x1D67B: []rune{0x006C}, // Additional folding + 0x1D67C: []rune{0x006D}, // Additional folding + 0x1D67D: []rune{0x006E}, // Additional folding + 0x1D67E: []rune{0x006F}, // Additional folding + 0x1D67F: []rune{0x0070}, // Additional folding + 0x1D680: []rune{0x0071}, // Additional folding + 0x1D681: []rune{0x0072}, // Additional folding + 0x1D682: []rune{0x0073}, // Additional folding + 0x1D683: []rune{0x0074}, // Additional folding + 0x1D684: []rune{0x0075}, // Additional folding + 0x1D685: []rune{0x0076}, // Additional folding + 0x1D686: []rune{0x0077}, // Additional folding + 0x1D687: []rune{0x0078}, // Additional folding + 0x1D688: []rune{0x0079}, // Additional folding + 0x1D689: []rune{0x007A}, // Additional folding + 0x1D6A8: []rune{0x03B1}, // Additional folding + 0x1D6A9: []rune{0x03B2}, // Additional folding + 0x1D6AA: []rune{0x03B3}, // Additional folding + 0x1D6AB: []rune{0x03B4}, // Additional folding + 0x1D6AC: []rune{0x03B5}, // Additional folding + 0x1D6AD: []rune{0x03B6}, // Additional folding + 0x1D6AE: []rune{0x03B7}, // Additional folding + 0x1D6AF: []rune{0x03B8}, // Additional folding + 0x1D6B0: []rune{0x03B9}, // Additional folding + 0x1D6B1: []rune{0x03BA}, // Additional folding + 0x1D6B2: []rune{0x03BB}, // Additional folding + 0x1D6B3: []rune{0x03BC}, // Additional folding + 0x1D6B4: []rune{0x03BD}, // Additional folding + 0x1D6B5: []rune{0x03BE}, // Additional folding + 0x1D6B6: []rune{0x03BF}, // Additional folding + 0x1D6B7: []rune{0x03C0}, // Additional folding + 0x1D6B8: []rune{0x03C1}, // Additional folding + 0x1D6B9: []rune{0x03B8}, // Additional folding + 0x1D6BA: []rune{0x03C3}, // Additional folding + 0x1D6BB: []rune{0x03C4}, // Additional folding + 0x1D6BC: []rune{0x03C5}, // Additional folding + 0x1D6BD: []rune{0x03C6}, // Additional folding + 0x1D6BE: []rune{0x03C7}, // Additional folding + 0x1D6BF: []rune{0x03C8}, // Additional folding + 0x1D6C0: []rune{0x03C9}, // Additional folding + 0x1D6D3: []rune{0x03C3}, // Additional folding + 0x1D6E2: []rune{0x03B1}, // Additional folding + 0x1D6E3: []rune{0x03B2}, // Additional folding + 0x1D6E4: []rune{0x03B3}, // Additional folding + 0x1D6E5: []rune{0x03B4}, // Additional folding + 0x1D6E6: []rune{0x03B5}, // Additional folding + 0x1D6E7: []rune{0x03B6}, // Additional folding + 0x1D6E8: []rune{0x03B7}, // Additional folding + 0x1D6E9: []rune{0x03B8}, // Additional folding + 0x1D6EA: []rune{0x03B9}, // Additional folding + 0x1D6EB: []rune{0x03BA}, // Additional folding + 0x1D6EC: []rune{0x03BB}, // Additional folding + 0x1D6ED: []rune{0x03BC}, // Additional folding + 0x1D6EE: []rune{0x03BD}, // Additional folding + 0x1D6EF: []rune{0x03BE}, // Additional folding + 0x1D6F0: []rune{0x03BF}, // Additional folding + 0x1D6F1: []rune{0x03C0}, // Additional folding + 0x1D6F2: []rune{0x03C1}, // Additional folding + 0x1D6F3: []rune{0x03B8}, // Additional folding + 0x1D6F4: []rune{0x03C3}, // Additional folding + 0x1D6F5: []rune{0x03C4}, // Additional folding + 0x1D6F6: []rune{0x03C5}, // Additional folding + 0x1D6F7: []rune{0x03C6}, // Additional folding + 0x1D6F8: []rune{0x03C7}, // Additional folding + 0x1D6F9: []rune{0x03C8}, // Additional folding + 0x1D6FA: []rune{0x03C9}, // Additional folding + 0x1D70D: []rune{0x03C3}, // Additional folding + 0x1D71C: []rune{0x03B1}, // Additional folding + 0x1D71D: []rune{0x03B2}, // Additional folding + 0x1D71E: []rune{0x03B3}, // Additional folding + 0x1D71F: []rune{0x03B4}, // Additional folding + 0x1D720: []rune{0x03B5}, // Additional folding + 0x1D721: []rune{0x03B6}, // Additional folding + 0x1D722: []rune{0x03B7}, // Additional folding + 0x1D723: []rune{0x03B8}, // Additional folding + 0x1D724: []rune{0x03B9}, // Additional folding + 0x1D725: []rune{0x03BA}, // Additional folding + 0x1D726: []rune{0x03BB}, // Additional folding + 0x1D727: []rune{0x03BC}, // Additional folding + 0x1D728: []rune{0x03BD}, // Additional folding + 0x1D729: []rune{0x03BE}, // Additional folding + 0x1D72A: []rune{0x03BF}, // Additional folding + 0x1D72B: []rune{0x03C0}, // Additional folding + 0x1D72C: []rune{0x03C1}, // Additional folding + 0x1D72D: []rune{0x03B8}, // Additional folding + 0x1D72E: []rune{0x03C3}, // Additional folding + 0x1D72F: []rune{0x03C4}, // Additional folding + 0x1D730: []rune{0x03C5}, // Additional folding + 0x1D731: []rune{0x03C6}, // Additional folding + 0x1D732: []rune{0x03C7}, // Additional folding + 0x1D733: []rune{0x03C8}, // Additional folding + 0x1D734: []rune{0x03C9}, // Additional folding + 0x1D747: []rune{0x03C3}, // Additional folding + 0x1D756: []rune{0x03B1}, // Additional folding + 0x1D757: []rune{0x03B2}, // Additional folding + 0x1D758: []rune{0x03B3}, // Additional folding + 0x1D759: []rune{0x03B4}, // Additional folding + 0x1D75A: []rune{0x03B5}, // Additional folding + 0x1D75B: []rune{0x03B6}, // Additional folding + 0x1D75C: []rune{0x03B7}, // Additional folding + 0x1D75D: []rune{0x03B8}, // Additional folding + 0x1D75E: []rune{0x03B9}, // Additional folding + 0x1D75F: []rune{0x03BA}, // Additional folding + 0x1D760: []rune{0x03BB}, // Additional folding + 0x1D761: []rune{0x03BC}, // Additional folding + 0x1D762: []rune{0x03BD}, // Additional folding + 0x1D763: []rune{0x03BE}, // Additional folding + 0x1D764: []rune{0x03BF}, // Additional folding + 0x1D765: []rune{0x03C0}, // Additional folding + 0x1D766: []rune{0x03C1}, // Additional folding + 0x1D767: []rune{0x03B8}, // Additional folding + 0x1D768: []rune{0x03C3}, // Additional folding + 0x1D769: []rune{0x03C4}, // Additional folding + 0x1D76A: []rune{0x03C5}, // Additional folding + 0x1D76B: []rune{0x03C6}, // Additional folding + 0x1D76C: []rune{0x03C7}, // Additional folding + 0x1D76D: []rune{0x03C8}, // Additional folding + 0x1D76E: []rune{0x03C9}, // Additional folding + 0x1D781: []rune{0x03C3}, // Additional folding + 0x1D790: []rune{0x03B1}, // Additional folding + 0x1D791: []rune{0x03B2}, // Additional folding + 0x1D792: []rune{0x03B3}, // Additional folding + 0x1D793: []rune{0x03B4}, // Additional folding + 0x1D794: []rune{0x03B5}, // Additional folding + 0x1D795: []rune{0x03B6}, // Additional folding + 0x1D796: []rune{0x03B7}, // Additional folding + 0x1D797: []rune{0x03B8}, // Additional folding + 0x1D798: []rune{0x03B9}, // Additional folding + 0x1D799: []rune{0x03BA}, // Additional folding + 0x1D79A: []rune{0x03BB}, // Additional folding + 0x1D79B: []rune{0x03BC}, // Additional folding + 0x1D79C: []rune{0x03BD}, // Additional folding + 0x1D79D: []rune{0x03BE}, // Additional folding + 0x1D79E: []rune{0x03BF}, // Additional folding + 0x1D79F: []rune{0x03C0}, // Additional folding + 0x1D7A0: []rune{0x03C1}, // Additional folding + 0x1D7A1: []rune{0x03B8}, // Additional folding + 0x1D7A2: []rune{0x03C3}, // Additional folding + 0x1D7A3: []rune{0x03C4}, // Additional folding + 0x1D7A4: []rune{0x03C5}, // Additional folding + 0x1D7A5: []rune{0x03C6}, // Additional folding + 0x1D7A6: []rune{0x03C7}, // Additional folding + 0x1D7A7: []rune{0x03C8}, // Additional folding + 0x1D7A8: []rune{0x03C9}, // Additional folding + 0x1D7BB: []rune{0x03C3}, // Additional folding +} + +// TableB2 represents RFC-3454 Table B.2. +var TableB2 Mapping = tableB2 + +var tableB3 = Mapping{ + 0x0041: []rune{0x0061}, // Case map + 0x0042: []rune{0x0062}, // Case map + 0x0043: []rune{0x0063}, // Case map + 0x0044: []rune{0x0064}, // Case map + 0x0045: []rune{0x0065}, // Case map + 0x0046: []rune{0x0066}, // Case map + 0x0047: []rune{0x0067}, // Case map + 0x0048: []rune{0x0068}, // Case map + 0x0049: []rune{0x0069}, // Case map + 0x004A: []rune{0x006A}, // Case map + 0x004B: []rune{0x006B}, // Case map + 0x004C: []rune{0x006C}, // Case map + 0x004D: []rune{0x006D}, // Case map + 0x004E: []rune{0x006E}, // Case map + 0x004F: []rune{0x006F}, // Case map + 0x0050: []rune{0x0070}, // Case map + 0x0051: []rune{0x0071}, // Case map + 0x0052: []rune{0x0072}, // Case map + 0x0053: []rune{0x0073}, // Case map + 0x0054: []rune{0x0074}, // Case map + 0x0055: []rune{0x0075}, // Case map + 0x0056: []rune{0x0076}, // Case map + 0x0057: []rune{0x0077}, // Case map + 0x0058: []rune{0x0078}, // Case map + 0x0059: []rune{0x0079}, // Case map + 0x005A: []rune{0x007A}, // Case map + 0x00B5: []rune{0x03BC}, // Case map + 0x00C0: []rune{0x00E0}, // Case map + 0x00C1: []rune{0x00E1}, // Case map + 0x00C2: []rune{0x00E2}, // Case map + 0x00C3: []rune{0x00E3}, // Case map + 0x00C4: []rune{0x00E4}, // Case map + 0x00C5: []rune{0x00E5}, // Case map + 0x00C6: []rune{0x00E6}, // Case map + 0x00C7: []rune{0x00E7}, // Case map + 0x00C8: []rune{0x00E8}, // Case map + 0x00C9: []rune{0x00E9}, // Case map + 0x00CA: []rune{0x00EA}, // Case map + 0x00CB: []rune{0x00EB}, // Case map + 0x00CC: []rune{0x00EC}, // Case map + 0x00CD: []rune{0x00ED}, // Case map + 0x00CE: []rune{0x00EE}, // Case map + 0x00CF: []rune{0x00EF}, // Case map + 0x00D0: []rune{0x00F0}, // Case map + 0x00D1: []rune{0x00F1}, // Case map + 0x00D2: []rune{0x00F2}, // Case map + 0x00D3: []rune{0x00F3}, // Case map + 0x00D4: []rune{0x00F4}, // Case map + 0x00D5: []rune{0x00F5}, // Case map + 0x00D6: []rune{0x00F6}, // Case map + 0x00D8: []rune{0x00F8}, // Case map + 0x00D9: []rune{0x00F9}, // Case map + 0x00DA: []rune{0x00FA}, // Case map + 0x00DB: []rune{0x00FB}, // Case map + 0x00DC: []rune{0x00FC}, // Case map + 0x00DD: []rune{0x00FD}, // Case map + 0x00DE: []rune{0x00FE}, // Case map + 0x00DF: []rune{0x0073, 0x0073}, // Case map + 0x0100: []rune{0x0101}, // Case map + 0x0102: []rune{0x0103}, // Case map + 0x0104: []rune{0x0105}, // Case map + 0x0106: []rune{0x0107}, // Case map + 0x0108: []rune{0x0109}, // Case map + 0x010A: []rune{0x010B}, // Case map + 0x010C: []rune{0x010D}, // Case map + 0x010E: []rune{0x010F}, // Case map + 0x0110: []rune{0x0111}, // Case map + 0x0112: []rune{0x0113}, // Case map + 0x0114: []rune{0x0115}, // Case map + 0x0116: []rune{0x0117}, // Case map + 0x0118: []rune{0x0119}, // Case map + 0x011A: []rune{0x011B}, // Case map + 0x011C: []rune{0x011D}, // Case map + 0x011E: []rune{0x011F}, // Case map + 0x0120: []rune{0x0121}, // Case map + 0x0122: []rune{0x0123}, // Case map + 0x0124: []rune{0x0125}, // Case map + 0x0126: []rune{0x0127}, // Case map + 0x0128: []rune{0x0129}, // Case map + 0x012A: []rune{0x012B}, // Case map + 0x012C: []rune{0x012D}, // Case map + 0x012E: []rune{0x012F}, // Case map + 0x0130: []rune{0x0069, 0x0307}, // Case map + 0x0132: []rune{0x0133}, // Case map + 0x0134: []rune{0x0135}, // Case map + 0x0136: []rune{0x0137}, // Case map + 0x0139: []rune{0x013A}, // Case map + 0x013B: []rune{0x013C}, // Case map + 0x013D: []rune{0x013E}, // Case map + 0x013F: []rune{0x0140}, // Case map + 0x0141: []rune{0x0142}, // Case map + 0x0143: []rune{0x0144}, // Case map + 0x0145: []rune{0x0146}, // Case map + 0x0147: []rune{0x0148}, // Case map + 0x0149: []rune{0x02BC, 0x006E}, // Case map + 0x014A: []rune{0x014B}, // Case map + 0x014C: []rune{0x014D}, // Case map + 0x014E: []rune{0x014F}, // Case map + 0x0150: []rune{0x0151}, // Case map + 0x0152: []rune{0x0153}, // Case map + 0x0154: []rune{0x0155}, // Case map + 0x0156: []rune{0x0157}, // Case map + 0x0158: []rune{0x0159}, // Case map + 0x015A: []rune{0x015B}, // Case map + 0x015C: []rune{0x015D}, // Case map + 0x015E: []rune{0x015F}, // Case map + 0x0160: []rune{0x0161}, // Case map + 0x0162: []rune{0x0163}, // Case map + 0x0164: []rune{0x0165}, // Case map + 0x0166: []rune{0x0167}, // Case map + 0x0168: []rune{0x0169}, // Case map + 0x016A: []rune{0x016B}, // Case map + 0x016C: []rune{0x016D}, // Case map + 0x016E: []rune{0x016F}, // Case map + 0x0170: []rune{0x0171}, // Case map + 0x0172: []rune{0x0173}, // Case map + 0x0174: []rune{0x0175}, // Case map + 0x0176: []rune{0x0177}, // Case map + 0x0178: []rune{0x00FF}, // Case map + 0x0179: []rune{0x017A}, // Case map + 0x017B: []rune{0x017C}, // Case map + 0x017D: []rune{0x017E}, // Case map + 0x017F: []rune{0x0073}, // Case map + 0x0181: []rune{0x0253}, // Case map + 0x0182: []rune{0x0183}, // Case map + 0x0184: []rune{0x0185}, // Case map + 0x0186: []rune{0x0254}, // Case map + 0x0187: []rune{0x0188}, // Case map + 0x0189: []rune{0x0256}, // Case map + 0x018A: []rune{0x0257}, // Case map + 0x018B: []rune{0x018C}, // Case map + 0x018E: []rune{0x01DD}, // Case map + 0x018F: []rune{0x0259}, // Case map + 0x0190: []rune{0x025B}, // Case map + 0x0191: []rune{0x0192}, // Case map + 0x0193: []rune{0x0260}, // Case map + 0x0194: []rune{0x0263}, // Case map + 0x0196: []rune{0x0269}, // Case map + 0x0197: []rune{0x0268}, // Case map + 0x0198: []rune{0x0199}, // Case map + 0x019C: []rune{0x026F}, // Case map + 0x019D: []rune{0x0272}, // Case map + 0x019F: []rune{0x0275}, // Case map + 0x01A0: []rune{0x01A1}, // Case map + 0x01A2: []rune{0x01A3}, // Case map + 0x01A4: []rune{0x01A5}, // Case map + 0x01A6: []rune{0x0280}, // Case map + 0x01A7: []rune{0x01A8}, // Case map + 0x01A9: []rune{0x0283}, // Case map + 0x01AC: []rune{0x01AD}, // Case map + 0x01AE: []rune{0x0288}, // Case map + 0x01AF: []rune{0x01B0}, // Case map + 0x01B1: []rune{0x028A}, // Case map + 0x01B2: []rune{0x028B}, // Case map + 0x01B3: []rune{0x01B4}, // Case map + 0x01B5: []rune{0x01B6}, // Case map + 0x01B7: []rune{0x0292}, // Case map + 0x01B8: []rune{0x01B9}, // Case map + 0x01BC: []rune{0x01BD}, // Case map + 0x01C4: []rune{0x01C6}, // Case map + 0x01C5: []rune{0x01C6}, // Case map + 0x01C7: []rune{0x01C9}, // Case map + 0x01C8: []rune{0x01C9}, // Case map + 0x01CA: []rune{0x01CC}, // Case map + 0x01CB: []rune{0x01CC}, // Case map + 0x01CD: []rune{0x01CE}, // Case map + 0x01CF: []rune{0x01D0}, // Case map + 0x01D1: []rune{0x01D2}, // Case map + 0x01D3: []rune{0x01D4}, // Case map + 0x01D5: []rune{0x01D6}, // Case map + 0x01D7: []rune{0x01D8}, // Case map + 0x01D9: []rune{0x01DA}, // Case map + 0x01DB: []rune{0x01DC}, // Case map + 0x01DE: []rune{0x01DF}, // Case map + 0x01E0: []rune{0x01E1}, // Case map + 0x01E2: []rune{0x01E3}, // Case map + 0x01E4: []rune{0x01E5}, // Case map + 0x01E6: []rune{0x01E7}, // Case map + 0x01E8: []rune{0x01E9}, // Case map + 0x01EA: []rune{0x01EB}, // Case map + 0x01EC: []rune{0x01ED}, // Case map + 0x01EE: []rune{0x01EF}, // Case map + 0x01F0: []rune{0x006A, 0x030C}, // Case map + 0x01F1: []rune{0x01F3}, // Case map + 0x01F2: []rune{0x01F3}, // Case map + 0x01F4: []rune{0x01F5}, // Case map + 0x01F6: []rune{0x0195}, // Case map + 0x01F7: []rune{0x01BF}, // Case map + 0x01F8: []rune{0x01F9}, // Case map + 0x01FA: []rune{0x01FB}, // Case map + 0x01FC: []rune{0x01FD}, // Case map + 0x01FE: []rune{0x01FF}, // Case map + 0x0200: []rune{0x0201}, // Case map + 0x0202: []rune{0x0203}, // Case map + 0x0204: []rune{0x0205}, // Case map + 0x0206: []rune{0x0207}, // Case map + 0x0208: []rune{0x0209}, // Case map + 0x020A: []rune{0x020B}, // Case map + 0x020C: []rune{0x020D}, // Case map + 0x020E: []rune{0x020F}, // Case map + 0x0210: []rune{0x0211}, // Case map + 0x0212: []rune{0x0213}, // Case map + 0x0214: []rune{0x0215}, // Case map + 0x0216: []rune{0x0217}, // Case map + 0x0218: []rune{0x0219}, // Case map + 0x021A: []rune{0x021B}, // Case map + 0x021C: []rune{0x021D}, // Case map + 0x021E: []rune{0x021F}, // Case map + 0x0220: []rune{0x019E}, // Case map + 0x0222: []rune{0x0223}, // Case map + 0x0224: []rune{0x0225}, // Case map + 0x0226: []rune{0x0227}, // Case map + 0x0228: []rune{0x0229}, // Case map + 0x022A: []rune{0x022B}, // Case map + 0x022C: []rune{0x022D}, // Case map + 0x022E: []rune{0x022F}, // Case map + 0x0230: []rune{0x0231}, // Case map + 0x0232: []rune{0x0233}, // Case map + 0x0345: []rune{0x03B9}, // Case map + 0x0386: []rune{0x03AC}, // Case map + 0x0388: []rune{0x03AD}, // Case map + 0x0389: []rune{0x03AE}, // Case map + 0x038A: []rune{0x03AF}, // Case map + 0x038C: []rune{0x03CC}, // Case map + 0x038E: []rune{0x03CD}, // Case map + 0x038F: []rune{0x03CE}, // Case map + 0x0390: []rune{0x03B9, 0x0308, 0x0301}, // Case map + 0x0391: []rune{0x03B1}, // Case map + 0x0392: []rune{0x03B2}, // Case map + 0x0393: []rune{0x03B3}, // Case map + 0x0394: []rune{0x03B4}, // Case map + 0x0395: []rune{0x03B5}, // Case map + 0x0396: []rune{0x03B6}, // Case map + 0x0397: []rune{0x03B7}, // Case map + 0x0398: []rune{0x03B8}, // Case map + 0x0399: []rune{0x03B9}, // Case map + 0x039A: []rune{0x03BA}, // Case map + 0x039B: []rune{0x03BB}, // Case map + 0x039C: []rune{0x03BC}, // Case map + 0x039D: []rune{0x03BD}, // Case map + 0x039E: []rune{0x03BE}, // Case map + 0x039F: []rune{0x03BF}, // Case map + 0x03A0: []rune{0x03C0}, // Case map + 0x03A1: []rune{0x03C1}, // Case map + 0x03A3: []rune{0x03C3}, // Case map + 0x03A4: []rune{0x03C4}, // Case map + 0x03A5: []rune{0x03C5}, // Case map + 0x03A6: []rune{0x03C6}, // Case map + 0x03A7: []rune{0x03C7}, // Case map + 0x03A8: []rune{0x03C8}, // Case map + 0x03A9: []rune{0x03C9}, // Case map + 0x03AA: []rune{0x03CA}, // Case map + 0x03AB: []rune{0x03CB}, // Case map + 0x03B0: []rune{0x03C5, 0x0308, 0x0301}, // Case map + 0x03C2: []rune{0x03C3}, // Case map + 0x03D0: []rune{0x03B2}, // Case map + 0x03D1: []rune{0x03B8}, // Case map + 0x03D5: []rune{0x03C6}, // Case map + 0x03D6: []rune{0x03C0}, // Case map + 0x03D8: []rune{0x03D9}, // Case map + 0x03DA: []rune{0x03DB}, // Case map + 0x03DC: []rune{0x03DD}, // Case map + 0x03DE: []rune{0x03DF}, // Case map + 0x03E0: []rune{0x03E1}, // Case map + 0x03E2: []rune{0x03E3}, // Case map + 0x03E4: []rune{0x03E5}, // Case map + 0x03E6: []rune{0x03E7}, // Case map + 0x03E8: []rune{0x03E9}, // Case map + 0x03EA: []rune{0x03EB}, // Case map + 0x03EC: []rune{0x03ED}, // Case map + 0x03EE: []rune{0x03EF}, // Case map + 0x03F0: []rune{0x03BA}, // Case map + 0x03F1: []rune{0x03C1}, // Case map + 0x03F2: []rune{0x03C3}, // Case map + 0x03F4: []rune{0x03B8}, // Case map + 0x03F5: []rune{0x03B5}, // Case map + 0x0400: []rune{0x0450}, // Case map + 0x0401: []rune{0x0451}, // Case map + 0x0402: []rune{0x0452}, // Case map + 0x0403: []rune{0x0453}, // Case map + 0x0404: []rune{0x0454}, // Case map + 0x0405: []rune{0x0455}, // Case map + 0x0406: []rune{0x0456}, // Case map + 0x0407: []rune{0x0457}, // Case map + 0x0408: []rune{0x0458}, // Case map + 0x0409: []rune{0x0459}, // Case map + 0x040A: []rune{0x045A}, // Case map + 0x040B: []rune{0x045B}, // Case map + 0x040C: []rune{0x045C}, // Case map + 0x040D: []rune{0x045D}, // Case map + 0x040E: []rune{0x045E}, // Case map + 0x040F: []rune{0x045F}, // Case map + 0x0410: []rune{0x0430}, // Case map + 0x0411: []rune{0x0431}, // Case map + 0x0412: []rune{0x0432}, // Case map + 0x0413: []rune{0x0433}, // Case map + 0x0414: []rune{0x0434}, // Case map + 0x0415: []rune{0x0435}, // Case map + 0x0416: []rune{0x0436}, // Case map + 0x0417: []rune{0x0437}, // Case map + 0x0418: []rune{0x0438}, // Case map + 0x0419: []rune{0x0439}, // Case map + 0x041A: []rune{0x043A}, // Case map + 0x041B: []rune{0x043B}, // Case map + 0x041C: []rune{0x043C}, // Case map + 0x041D: []rune{0x043D}, // Case map + 0x041E: []rune{0x043E}, // Case map + 0x041F: []rune{0x043F}, // Case map + 0x0420: []rune{0x0440}, // Case map + 0x0421: []rune{0x0441}, // Case map + 0x0422: []rune{0x0442}, // Case map + 0x0423: []rune{0x0443}, // Case map + 0x0424: []rune{0x0444}, // Case map + 0x0425: []rune{0x0445}, // Case map + 0x0426: []rune{0x0446}, // Case map + 0x0427: []rune{0x0447}, // Case map + 0x0428: []rune{0x0448}, // Case map + 0x0429: []rune{0x0449}, // Case map + 0x042A: []rune{0x044A}, // Case map + 0x042B: []rune{0x044B}, // Case map + 0x042C: []rune{0x044C}, // Case map + 0x042D: []rune{0x044D}, // Case map + 0x042E: []rune{0x044E}, // Case map + 0x042F: []rune{0x044F}, // Case map + 0x0460: []rune{0x0461}, // Case map + 0x0462: []rune{0x0463}, // Case map + 0x0464: []rune{0x0465}, // Case map + 0x0466: []rune{0x0467}, // Case map + 0x0468: []rune{0x0469}, // Case map + 0x046A: []rune{0x046B}, // Case map + 0x046C: []rune{0x046D}, // Case map + 0x046E: []rune{0x046F}, // Case map + 0x0470: []rune{0x0471}, // Case map + 0x0472: []rune{0x0473}, // Case map + 0x0474: []rune{0x0475}, // Case map + 0x0476: []rune{0x0477}, // Case map + 0x0478: []rune{0x0479}, // Case map + 0x047A: []rune{0x047B}, // Case map + 0x047C: []rune{0x047D}, // Case map + 0x047E: []rune{0x047F}, // Case map + 0x0480: []rune{0x0481}, // Case map + 0x048A: []rune{0x048B}, // Case map + 0x048C: []rune{0x048D}, // Case map + 0x048E: []rune{0x048F}, // Case map + 0x0490: []rune{0x0491}, // Case map + 0x0492: []rune{0x0493}, // Case map + 0x0494: []rune{0x0495}, // Case map + 0x0496: []rune{0x0497}, // Case map + 0x0498: []rune{0x0499}, // Case map + 0x049A: []rune{0x049B}, // Case map + 0x049C: []rune{0x049D}, // Case map + 0x049E: []rune{0x049F}, // Case map + 0x04A0: []rune{0x04A1}, // Case map + 0x04A2: []rune{0x04A3}, // Case map + 0x04A4: []rune{0x04A5}, // Case map + 0x04A6: []rune{0x04A7}, // Case map + 0x04A8: []rune{0x04A9}, // Case map + 0x04AA: []rune{0x04AB}, // Case map + 0x04AC: []rune{0x04AD}, // Case map + 0x04AE: []rune{0x04AF}, // Case map + 0x04B0: []rune{0x04B1}, // Case map + 0x04B2: []rune{0x04B3}, // Case map + 0x04B4: []rune{0x04B5}, // Case map + 0x04B6: []rune{0x04B7}, // Case map + 0x04B8: []rune{0x04B9}, // Case map + 0x04BA: []rune{0x04BB}, // Case map + 0x04BC: []rune{0x04BD}, // Case map + 0x04BE: []rune{0x04BF}, // Case map + 0x04C1: []rune{0x04C2}, // Case map + 0x04C3: []rune{0x04C4}, // Case map + 0x04C5: []rune{0x04C6}, // Case map + 0x04C7: []rune{0x04C8}, // Case map + 0x04C9: []rune{0x04CA}, // Case map + 0x04CB: []rune{0x04CC}, // Case map + 0x04CD: []rune{0x04CE}, // Case map + 0x04D0: []rune{0x04D1}, // Case map + 0x04D2: []rune{0x04D3}, // Case map + 0x04D4: []rune{0x04D5}, // Case map + 0x04D6: []rune{0x04D7}, // Case map + 0x04D8: []rune{0x04D9}, // Case map + 0x04DA: []rune{0x04DB}, // Case map + 0x04DC: []rune{0x04DD}, // Case map + 0x04DE: []rune{0x04DF}, // Case map + 0x04E0: []rune{0x04E1}, // Case map + 0x04E2: []rune{0x04E3}, // Case map + 0x04E4: []rune{0x04E5}, // Case map + 0x04E6: []rune{0x04E7}, // Case map + 0x04E8: []rune{0x04E9}, // Case map + 0x04EA: []rune{0x04EB}, // Case map + 0x04EC: []rune{0x04ED}, // Case map + 0x04EE: []rune{0x04EF}, // Case map + 0x04F0: []rune{0x04F1}, // Case map + 0x04F2: []rune{0x04F3}, // Case map + 0x04F4: []rune{0x04F5}, // Case map + 0x04F8: []rune{0x04F9}, // Case map + 0x0500: []rune{0x0501}, // Case map + 0x0502: []rune{0x0503}, // Case map + 0x0504: []rune{0x0505}, // Case map + 0x0506: []rune{0x0507}, // Case map + 0x0508: []rune{0x0509}, // Case map + 0x050A: []rune{0x050B}, // Case map + 0x050C: []rune{0x050D}, // Case map + 0x050E: []rune{0x050F}, // Case map + 0x0531: []rune{0x0561}, // Case map + 0x0532: []rune{0x0562}, // Case map + 0x0533: []rune{0x0563}, // Case map + 0x0534: []rune{0x0564}, // Case map + 0x0535: []rune{0x0565}, // Case map + 0x0536: []rune{0x0566}, // Case map + 0x0537: []rune{0x0567}, // Case map + 0x0538: []rune{0x0568}, // Case map + 0x0539: []rune{0x0569}, // Case map + 0x053A: []rune{0x056A}, // Case map + 0x053B: []rune{0x056B}, // Case map + 0x053C: []rune{0x056C}, // Case map + 0x053D: []rune{0x056D}, // Case map + 0x053E: []rune{0x056E}, // Case map + 0x053F: []rune{0x056F}, // Case map + 0x0540: []rune{0x0570}, // Case map + 0x0541: []rune{0x0571}, // Case map + 0x0542: []rune{0x0572}, // Case map + 0x0543: []rune{0x0573}, // Case map + 0x0544: []rune{0x0574}, // Case map + 0x0545: []rune{0x0575}, // Case map + 0x0546: []rune{0x0576}, // Case map + 0x0547: []rune{0x0577}, // Case map + 0x0548: []rune{0x0578}, // Case map + 0x0549: []rune{0x0579}, // Case map + 0x054A: []rune{0x057A}, // Case map + 0x054B: []rune{0x057B}, // Case map + 0x054C: []rune{0x057C}, // Case map + 0x054D: []rune{0x057D}, // Case map + 0x054E: []rune{0x057E}, // Case map + 0x054F: []rune{0x057F}, // Case map + 0x0550: []rune{0x0580}, // Case map + 0x0551: []rune{0x0581}, // Case map + 0x0552: []rune{0x0582}, // Case map + 0x0553: []rune{0x0583}, // Case map + 0x0554: []rune{0x0584}, // Case map + 0x0555: []rune{0x0585}, // Case map + 0x0556: []rune{0x0586}, // Case map + 0x0587: []rune{0x0565, 0x0582}, // Case map + 0x1E00: []rune{0x1E01}, // Case map + 0x1E02: []rune{0x1E03}, // Case map + 0x1E04: []rune{0x1E05}, // Case map + 0x1E06: []rune{0x1E07}, // Case map + 0x1E08: []rune{0x1E09}, // Case map + 0x1E0A: []rune{0x1E0B}, // Case map + 0x1E0C: []rune{0x1E0D}, // Case map + 0x1E0E: []rune{0x1E0F}, // Case map + 0x1E10: []rune{0x1E11}, // Case map + 0x1E12: []rune{0x1E13}, // Case map + 0x1E14: []rune{0x1E15}, // Case map + 0x1E16: []rune{0x1E17}, // Case map + 0x1E18: []rune{0x1E19}, // Case map + 0x1E1A: []rune{0x1E1B}, // Case map + 0x1E1C: []rune{0x1E1D}, // Case map + 0x1E1E: []rune{0x1E1F}, // Case map + 0x1E20: []rune{0x1E21}, // Case map + 0x1E22: []rune{0x1E23}, // Case map + 0x1E24: []rune{0x1E25}, // Case map + 0x1E26: []rune{0x1E27}, // Case map + 0x1E28: []rune{0x1E29}, // Case map + 0x1E2A: []rune{0x1E2B}, // Case map + 0x1E2C: []rune{0x1E2D}, // Case map + 0x1E2E: []rune{0x1E2F}, // Case map + 0x1E30: []rune{0x1E31}, // Case map + 0x1E32: []rune{0x1E33}, // Case map + 0x1E34: []rune{0x1E35}, // Case map + 0x1E36: []rune{0x1E37}, // Case map + 0x1E38: []rune{0x1E39}, // Case map + 0x1E3A: []rune{0x1E3B}, // Case map + 0x1E3C: []rune{0x1E3D}, // Case map + 0x1E3E: []rune{0x1E3F}, // Case map + 0x1E40: []rune{0x1E41}, // Case map + 0x1E42: []rune{0x1E43}, // Case map + 0x1E44: []rune{0x1E45}, // Case map + 0x1E46: []rune{0x1E47}, // Case map + 0x1E48: []rune{0x1E49}, // Case map + 0x1E4A: []rune{0x1E4B}, // Case map + 0x1E4C: []rune{0x1E4D}, // Case map + 0x1E4E: []rune{0x1E4F}, // Case map + 0x1E50: []rune{0x1E51}, // Case map + 0x1E52: []rune{0x1E53}, // Case map + 0x1E54: []rune{0x1E55}, // Case map + 0x1E56: []rune{0x1E57}, // Case map + 0x1E58: []rune{0x1E59}, // Case map + 0x1E5A: []rune{0x1E5B}, // Case map + 0x1E5C: []rune{0x1E5D}, // Case map + 0x1E5E: []rune{0x1E5F}, // Case map + 0x1E60: []rune{0x1E61}, // Case map + 0x1E62: []rune{0x1E63}, // Case map + 0x1E64: []rune{0x1E65}, // Case map + 0x1E66: []rune{0x1E67}, // Case map + 0x1E68: []rune{0x1E69}, // Case map + 0x1E6A: []rune{0x1E6B}, // Case map + 0x1E6C: []rune{0x1E6D}, // Case map + 0x1E6E: []rune{0x1E6F}, // Case map + 0x1E70: []rune{0x1E71}, // Case map + 0x1E72: []rune{0x1E73}, // Case map + 0x1E74: []rune{0x1E75}, // Case map + 0x1E76: []rune{0x1E77}, // Case map + 0x1E78: []rune{0x1E79}, // Case map + 0x1E7A: []rune{0x1E7B}, // Case map + 0x1E7C: []rune{0x1E7D}, // Case map + 0x1E7E: []rune{0x1E7F}, // Case map + 0x1E80: []rune{0x1E81}, // Case map + 0x1E82: []rune{0x1E83}, // Case map + 0x1E84: []rune{0x1E85}, // Case map + 0x1E86: []rune{0x1E87}, // Case map + 0x1E88: []rune{0x1E89}, // Case map + 0x1E8A: []rune{0x1E8B}, // Case map + 0x1E8C: []rune{0x1E8D}, // Case map + 0x1E8E: []rune{0x1E8F}, // Case map + 0x1E90: []rune{0x1E91}, // Case map + 0x1E92: []rune{0x1E93}, // Case map + 0x1E94: []rune{0x1E95}, // Case map + 0x1E96: []rune{0x0068, 0x0331}, // Case map + 0x1E97: []rune{0x0074, 0x0308}, // Case map + 0x1E98: []rune{0x0077, 0x030A}, // Case map + 0x1E99: []rune{0x0079, 0x030A}, // Case map + 0x1E9A: []rune{0x0061, 0x02BE}, // Case map + 0x1E9B: []rune{0x1E61}, // Case map + 0x1EA0: []rune{0x1EA1}, // Case map + 0x1EA2: []rune{0x1EA3}, // Case map + 0x1EA4: []rune{0x1EA5}, // Case map + 0x1EA6: []rune{0x1EA7}, // Case map + 0x1EA8: []rune{0x1EA9}, // Case map + 0x1EAA: []rune{0x1EAB}, // Case map + 0x1EAC: []rune{0x1EAD}, // Case map + 0x1EAE: []rune{0x1EAF}, // Case map + 0x1EB0: []rune{0x1EB1}, // Case map + 0x1EB2: []rune{0x1EB3}, // Case map + 0x1EB4: []rune{0x1EB5}, // Case map + 0x1EB6: []rune{0x1EB7}, // Case map + 0x1EB8: []rune{0x1EB9}, // Case map + 0x1EBA: []rune{0x1EBB}, // Case map + 0x1EBC: []rune{0x1EBD}, // Case map + 0x1EBE: []rune{0x1EBF}, // Case map + 0x1EC0: []rune{0x1EC1}, // Case map + 0x1EC2: []rune{0x1EC3}, // Case map + 0x1EC4: []rune{0x1EC5}, // Case map + 0x1EC6: []rune{0x1EC7}, // Case map + 0x1EC8: []rune{0x1EC9}, // Case map + 0x1ECA: []rune{0x1ECB}, // Case map + 0x1ECC: []rune{0x1ECD}, // Case map + 0x1ECE: []rune{0x1ECF}, // Case map + 0x1ED0: []rune{0x1ED1}, // Case map + 0x1ED2: []rune{0x1ED3}, // Case map + 0x1ED4: []rune{0x1ED5}, // Case map + 0x1ED6: []rune{0x1ED7}, // Case map + 0x1ED8: []rune{0x1ED9}, // Case map + 0x1EDA: []rune{0x1EDB}, // Case map + 0x1EDC: []rune{0x1EDD}, // Case map + 0x1EDE: []rune{0x1EDF}, // Case map + 0x1EE0: []rune{0x1EE1}, // Case map + 0x1EE2: []rune{0x1EE3}, // Case map + 0x1EE4: []rune{0x1EE5}, // Case map + 0x1EE6: []rune{0x1EE7}, // Case map + 0x1EE8: []rune{0x1EE9}, // Case map + 0x1EEA: []rune{0x1EEB}, // Case map + 0x1EEC: []rune{0x1EED}, // Case map + 0x1EEE: []rune{0x1EEF}, // Case map + 0x1EF0: []rune{0x1EF1}, // Case map + 0x1EF2: []rune{0x1EF3}, // Case map + 0x1EF4: []rune{0x1EF5}, // Case map + 0x1EF6: []rune{0x1EF7}, // Case map + 0x1EF8: []rune{0x1EF9}, // Case map + 0x1F08: []rune{0x1F00}, // Case map + 0x1F09: []rune{0x1F01}, // Case map + 0x1F0A: []rune{0x1F02}, // Case map + 0x1F0B: []rune{0x1F03}, // Case map + 0x1F0C: []rune{0x1F04}, // Case map + 0x1F0D: []rune{0x1F05}, // Case map + 0x1F0E: []rune{0x1F06}, // Case map + 0x1F0F: []rune{0x1F07}, // Case map + 0x1F18: []rune{0x1F10}, // Case map + 0x1F19: []rune{0x1F11}, // Case map + 0x1F1A: []rune{0x1F12}, // Case map + 0x1F1B: []rune{0x1F13}, // Case map + 0x1F1C: []rune{0x1F14}, // Case map + 0x1F1D: []rune{0x1F15}, // Case map + 0x1F28: []rune{0x1F20}, // Case map + 0x1F29: []rune{0x1F21}, // Case map + 0x1F2A: []rune{0x1F22}, // Case map + 0x1F2B: []rune{0x1F23}, // Case map + 0x1F2C: []rune{0x1F24}, // Case map + 0x1F2D: []rune{0x1F25}, // Case map + 0x1F2E: []rune{0x1F26}, // Case map + 0x1F2F: []rune{0x1F27}, // Case map + 0x1F38: []rune{0x1F30}, // Case map + 0x1F39: []rune{0x1F31}, // Case map + 0x1F3A: []rune{0x1F32}, // Case map + 0x1F3B: []rune{0x1F33}, // Case map + 0x1F3C: []rune{0x1F34}, // Case map + 0x1F3D: []rune{0x1F35}, // Case map + 0x1F3E: []rune{0x1F36}, // Case map + 0x1F3F: []rune{0x1F37}, // Case map + 0x1F48: []rune{0x1F40}, // Case map + 0x1F49: []rune{0x1F41}, // Case map + 0x1F4A: []rune{0x1F42}, // Case map + 0x1F4B: []rune{0x1F43}, // Case map + 0x1F4C: []rune{0x1F44}, // Case map + 0x1F4D: []rune{0x1F45}, // Case map + 0x1F50: []rune{0x03C5, 0x0313}, // Case map + 0x1F52: []rune{0x03C5, 0x0313, 0x0300}, // Case map + 0x1F54: []rune{0x03C5, 0x0313, 0x0301}, // Case map + 0x1F56: []rune{0x03C5, 0x0313, 0x0342}, // Case map + 0x1F59: []rune{0x1F51}, // Case map + 0x1F5B: []rune{0x1F53}, // Case map + 0x1F5D: []rune{0x1F55}, // Case map + 0x1F5F: []rune{0x1F57}, // Case map + 0x1F68: []rune{0x1F60}, // Case map + 0x1F69: []rune{0x1F61}, // Case map + 0x1F6A: []rune{0x1F62}, // Case map + 0x1F6B: []rune{0x1F63}, // Case map + 0x1F6C: []rune{0x1F64}, // Case map + 0x1F6D: []rune{0x1F65}, // Case map + 0x1F6E: []rune{0x1F66}, // Case map + 0x1F6F: []rune{0x1F67}, // Case map + 0x1F80: []rune{0x1F00, 0x03B9}, // Case map + 0x1F81: []rune{0x1F01, 0x03B9}, // Case map + 0x1F82: []rune{0x1F02, 0x03B9}, // Case map + 0x1F83: []rune{0x1F03, 0x03B9}, // Case map + 0x1F84: []rune{0x1F04, 0x03B9}, // Case map + 0x1F85: []rune{0x1F05, 0x03B9}, // Case map + 0x1F86: []rune{0x1F06, 0x03B9}, // Case map + 0x1F87: []rune{0x1F07, 0x03B9}, // Case map + 0x1F88: []rune{0x1F00, 0x03B9}, // Case map + 0x1F89: []rune{0x1F01, 0x03B9}, // Case map + 0x1F8A: []rune{0x1F02, 0x03B9}, // Case map + 0x1F8B: []rune{0x1F03, 0x03B9}, // Case map + 0x1F8C: []rune{0x1F04, 0x03B9}, // Case map + 0x1F8D: []rune{0x1F05, 0x03B9}, // Case map + 0x1F8E: []rune{0x1F06, 0x03B9}, // Case map + 0x1F8F: []rune{0x1F07, 0x03B9}, // Case map + 0x1F90: []rune{0x1F20, 0x03B9}, // Case map + 0x1F91: []rune{0x1F21, 0x03B9}, // Case map + 0x1F92: []rune{0x1F22, 0x03B9}, // Case map + 0x1F93: []rune{0x1F23, 0x03B9}, // Case map + 0x1F94: []rune{0x1F24, 0x03B9}, // Case map + 0x1F95: []rune{0x1F25, 0x03B9}, // Case map + 0x1F96: []rune{0x1F26, 0x03B9}, // Case map + 0x1F97: []rune{0x1F27, 0x03B9}, // Case map + 0x1F98: []rune{0x1F20, 0x03B9}, // Case map + 0x1F99: []rune{0x1F21, 0x03B9}, // Case map + 0x1F9A: []rune{0x1F22, 0x03B9}, // Case map + 0x1F9B: []rune{0x1F23, 0x03B9}, // Case map + 0x1F9C: []rune{0x1F24, 0x03B9}, // Case map + 0x1F9D: []rune{0x1F25, 0x03B9}, // Case map + 0x1F9E: []rune{0x1F26, 0x03B9}, // Case map + 0x1F9F: []rune{0x1F27, 0x03B9}, // Case map + 0x1FA0: []rune{0x1F60, 0x03B9}, // Case map + 0x1FA1: []rune{0x1F61, 0x03B9}, // Case map + 0x1FA2: []rune{0x1F62, 0x03B9}, // Case map + 0x1FA3: []rune{0x1F63, 0x03B9}, // Case map + 0x1FA4: []rune{0x1F64, 0x03B9}, // Case map + 0x1FA5: []rune{0x1F65, 0x03B9}, // Case map + 0x1FA6: []rune{0x1F66, 0x03B9}, // Case map + 0x1FA7: []rune{0x1F67, 0x03B9}, // Case map + 0x1FA8: []rune{0x1F60, 0x03B9}, // Case map + 0x1FA9: []rune{0x1F61, 0x03B9}, // Case map + 0x1FAA: []rune{0x1F62, 0x03B9}, // Case map + 0x1FAB: []rune{0x1F63, 0x03B9}, // Case map + 0x1FAC: []rune{0x1F64, 0x03B9}, // Case map + 0x1FAD: []rune{0x1F65, 0x03B9}, // Case map + 0x1FAE: []rune{0x1F66, 0x03B9}, // Case map + 0x1FAF: []rune{0x1F67, 0x03B9}, // Case map + 0x1FB2: []rune{0x1F70, 0x03B9}, // Case map + 0x1FB3: []rune{0x03B1, 0x03B9}, // Case map + 0x1FB4: []rune{0x03AC, 0x03B9}, // Case map + 0x1FB6: []rune{0x03B1, 0x0342}, // Case map + 0x1FB7: []rune{0x03B1, 0x0342, 0x03B9}, // Case map + 0x1FB8: []rune{0x1FB0}, // Case map + 0x1FB9: []rune{0x1FB1}, // Case map + 0x1FBA: []rune{0x1F70}, // Case map + 0x1FBB: []rune{0x1F71}, // Case map + 0x1FBC: []rune{0x03B1, 0x03B9}, // Case map + 0x1FBE: []rune{0x03B9}, // Case map + 0x1FC2: []rune{0x1F74, 0x03B9}, // Case map + 0x1FC3: []rune{0x03B7, 0x03B9}, // Case map + 0x1FC4: []rune{0x03AE, 0x03B9}, // Case map + 0x1FC6: []rune{0x03B7, 0x0342}, // Case map + 0x1FC7: []rune{0x03B7, 0x0342, 0x03B9}, // Case map + 0x1FC8: []rune{0x1F72}, // Case map + 0x1FC9: []rune{0x1F73}, // Case map + 0x1FCA: []rune{0x1F74}, // Case map + 0x1FCB: []rune{0x1F75}, // Case map + 0x1FCC: []rune{0x03B7, 0x03B9}, // Case map + 0x1FD2: []rune{0x03B9, 0x0308, 0x0300}, // Case map + 0x1FD3: []rune{0x03B9, 0x0308, 0x0301}, // Case map + 0x1FD6: []rune{0x03B9, 0x0342}, // Case map + 0x1FD7: []rune{0x03B9, 0x0308, 0x0342}, // Case map + 0x1FD8: []rune{0x1FD0}, // Case map + 0x1FD9: []rune{0x1FD1}, // Case map + 0x1FDA: []rune{0x1F76}, // Case map + 0x1FDB: []rune{0x1F77}, // Case map + 0x1FE2: []rune{0x03C5, 0x0308, 0x0300}, // Case map + 0x1FE3: []rune{0x03C5, 0x0308, 0x0301}, // Case map + 0x1FE4: []rune{0x03C1, 0x0313}, // Case map + 0x1FE6: []rune{0x03C5, 0x0342}, // Case map + 0x1FE7: []rune{0x03C5, 0x0308, 0x0342}, // Case map + 0x1FE8: []rune{0x1FE0}, // Case map + 0x1FE9: []rune{0x1FE1}, // Case map + 0x1FEA: []rune{0x1F7A}, // Case map + 0x1FEB: []rune{0x1F7B}, // Case map + 0x1FEC: []rune{0x1FE5}, // Case map + 0x1FF2: []rune{0x1F7C, 0x03B9}, // Case map + 0x1FF3: []rune{0x03C9, 0x03B9}, // Case map + 0x1FF4: []rune{0x03CE, 0x03B9}, // Case map + 0x1FF6: []rune{0x03C9, 0x0342}, // Case map + 0x1FF7: []rune{0x03C9, 0x0342, 0x03B9}, // Case map + 0x1FF8: []rune{0x1F78}, // Case map + 0x1FF9: []rune{0x1F79}, // Case map + 0x1FFA: []rune{0x1F7C}, // Case map + 0x1FFB: []rune{0x1F7D}, // Case map + 0x1FFC: []rune{0x03C9, 0x03B9}, // Case map + 0x2126: []rune{0x03C9}, // Case map + 0x212A: []rune{0x006B}, // Case map + 0x212B: []rune{0x00E5}, // Case map + 0x2160: []rune{0x2170}, // Case map + 0x2161: []rune{0x2171}, // Case map + 0x2162: []rune{0x2172}, // Case map + 0x2163: []rune{0x2173}, // Case map + 0x2164: []rune{0x2174}, // Case map + 0x2165: []rune{0x2175}, // Case map + 0x2166: []rune{0x2176}, // Case map + 0x2167: []rune{0x2177}, // Case map + 0x2168: []rune{0x2178}, // Case map + 0x2169: []rune{0x2179}, // Case map + 0x216A: []rune{0x217A}, // Case map + 0x216B: []rune{0x217B}, // Case map + 0x216C: []rune{0x217C}, // Case map + 0x216D: []rune{0x217D}, // Case map + 0x216E: []rune{0x217E}, // Case map + 0x216F: []rune{0x217F}, // Case map + 0x24B6: []rune{0x24D0}, // Case map + 0x24B7: []rune{0x24D1}, // Case map + 0x24B8: []rune{0x24D2}, // Case map + 0x24B9: []rune{0x24D3}, // Case map + 0x24BA: []rune{0x24D4}, // Case map + 0x24BB: []rune{0x24D5}, // Case map + 0x24BC: []rune{0x24D6}, // Case map + 0x24BD: []rune{0x24D7}, // Case map + 0x24BE: []rune{0x24D8}, // Case map + 0x24BF: []rune{0x24D9}, // Case map + 0x24C0: []rune{0x24DA}, // Case map + 0x24C1: []rune{0x24DB}, // Case map + 0x24C2: []rune{0x24DC}, // Case map + 0x24C3: []rune{0x24DD}, // Case map + 0x24C4: []rune{0x24DE}, // Case map + 0x24C5: []rune{0x24DF}, // Case map + 0x24C6: []rune{0x24E0}, // Case map + 0x24C7: []rune{0x24E1}, // Case map + 0x24C8: []rune{0x24E2}, // Case map + 0x24C9: []rune{0x24E3}, // Case map + 0x24CA: []rune{0x24E4}, // Case map + 0x24CB: []rune{0x24E5}, // Case map + 0x24CC: []rune{0x24E6}, // Case map + 0x24CD: []rune{0x24E7}, // Case map + 0x24CE: []rune{0x24E8}, // Case map + 0x24CF: []rune{0x24E9}, // Case map + 0xFB00: []rune{0x0066, 0x0066}, // Case map + 0xFB01: []rune{0x0066, 0x0069}, // Case map + 0xFB02: []rune{0x0066, 0x006C}, // Case map + 0xFB03: []rune{0x0066, 0x0066, 0x0069}, // Case map + 0xFB04: []rune{0x0066, 0x0066, 0x006C}, // Case map + 0xFB05: []rune{0x0073, 0x0074}, // Case map + 0xFB06: []rune{0x0073, 0x0074}, // Case map + 0xFB13: []rune{0x0574, 0x0576}, // Case map + 0xFB14: []rune{0x0574, 0x0565}, // Case map + 0xFB15: []rune{0x0574, 0x056B}, // Case map + 0xFB16: []rune{0x057E, 0x0576}, // Case map + 0xFB17: []rune{0x0574, 0x056D}, // Case map + 0xFF21: []rune{0xFF41}, // Case map + 0xFF22: []rune{0xFF42}, // Case map + 0xFF23: []rune{0xFF43}, // Case map + 0xFF24: []rune{0xFF44}, // Case map + 0xFF25: []rune{0xFF45}, // Case map + 0xFF26: []rune{0xFF46}, // Case map + 0xFF27: []rune{0xFF47}, // Case map + 0xFF28: []rune{0xFF48}, // Case map + 0xFF29: []rune{0xFF49}, // Case map + 0xFF2A: []rune{0xFF4A}, // Case map + 0xFF2B: []rune{0xFF4B}, // Case map + 0xFF2C: []rune{0xFF4C}, // Case map + 0xFF2D: []rune{0xFF4D}, // Case map + 0xFF2E: []rune{0xFF4E}, // Case map + 0xFF2F: []rune{0xFF4F}, // Case map + 0xFF30: []rune{0xFF50}, // Case map + 0xFF31: []rune{0xFF51}, // Case map + 0xFF32: []rune{0xFF52}, // Case map + 0xFF33: []rune{0xFF53}, // Case map + 0xFF34: []rune{0xFF54}, // Case map + 0xFF35: []rune{0xFF55}, // Case map + 0xFF36: []rune{0xFF56}, // Case map + 0xFF37: []rune{0xFF57}, // Case map + 0xFF38: []rune{0xFF58}, // Case map + 0xFF39: []rune{0xFF59}, // Case map + 0xFF3A: []rune{0xFF5A}, // Case map + 0x10400: []rune{0x10428}, // Case map + 0x10401: []rune{0x10429}, // Case map + 0x10402: []rune{0x1042A}, // Case map + 0x10403: []rune{0x1042B}, // Case map + 0x10404: []rune{0x1042C}, // Case map + 0x10405: []rune{0x1042D}, // Case map + 0x10406: []rune{0x1042E}, // Case map + 0x10407: []rune{0x1042F}, // Case map + 0x10408: []rune{0x10430}, // Case map + 0x10409: []rune{0x10431}, // Case map + 0x1040A: []rune{0x10432}, // Case map + 0x1040B: []rune{0x10433}, // Case map + 0x1040C: []rune{0x10434}, // Case map + 0x1040D: []rune{0x10435}, // Case map + 0x1040E: []rune{0x10436}, // Case map + 0x1040F: []rune{0x10437}, // Case map + 0x10410: []rune{0x10438}, // Case map + 0x10411: []rune{0x10439}, // Case map + 0x10412: []rune{0x1043A}, // Case map + 0x10413: []rune{0x1043B}, // Case map + 0x10414: []rune{0x1043C}, // Case map + 0x10415: []rune{0x1043D}, // Case map + 0x10416: []rune{0x1043E}, // Case map + 0x10417: []rune{0x1043F}, // Case map + 0x10418: []rune{0x10440}, // Case map + 0x10419: []rune{0x10441}, // Case map + 0x1041A: []rune{0x10442}, // Case map + 0x1041B: []rune{0x10443}, // Case map + 0x1041C: []rune{0x10444}, // Case map + 0x1041D: []rune{0x10445}, // Case map + 0x1041E: []rune{0x10446}, // Case map + 0x1041F: []rune{0x10447}, // Case map + 0x10420: []rune{0x10448}, // Case map + 0x10421: []rune{0x10449}, // Case map + 0x10422: []rune{0x1044A}, // Case map + 0x10423: []rune{0x1044B}, // Case map + 0x10424: []rune{0x1044C}, // Case map + 0x10425: []rune{0x1044D}, // Case map +} + +// TableB3 represents RFC-3454 Table B.3. +var TableB3 Mapping = tableB3 + +var tableC1_1 = Set{ + RuneRange{0x0020, 0x0020}, // SPACE +} + +// TableC1_1 represents RFC-3454 Table C.1.1. +var TableC1_1 Set = tableC1_1 + +var tableC1_2 = Set{ + RuneRange{0x00A0, 0x00A0}, // NO-BREAK SPACE + RuneRange{0x1680, 0x1680}, // OGHAM SPACE MARK + RuneRange{0x2000, 0x2000}, // EN QUAD + RuneRange{0x2001, 0x2001}, // EM QUAD + RuneRange{0x2002, 0x2002}, // EN SPACE + RuneRange{0x2003, 0x2003}, // EM SPACE + RuneRange{0x2004, 0x2004}, // THREE-PER-EM SPACE + RuneRange{0x2005, 0x2005}, // FOUR-PER-EM SPACE + RuneRange{0x2006, 0x2006}, // SIX-PER-EM SPACE + RuneRange{0x2007, 0x2007}, // FIGURE SPACE + RuneRange{0x2008, 0x2008}, // PUNCTUATION SPACE + RuneRange{0x2009, 0x2009}, // THIN SPACE + RuneRange{0x200A, 0x200A}, // HAIR SPACE + RuneRange{0x200B, 0x200B}, // ZERO WIDTH SPACE + RuneRange{0x202F, 0x202F}, // NARROW NO-BREAK SPACE + RuneRange{0x205F, 0x205F}, // MEDIUM MATHEMATICAL SPACE + RuneRange{0x3000, 0x3000}, // IDEOGRAPHIC SPACE +} + +// TableC1_2 represents RFC-3454 Table C.1.2. +var TableC1_2 Set = tableC1_2 + +var tableC2_1 = Set{ + RuneRange{0x0000, 0x001F}, // [CONTROL CHARACTERS] + RuneRange{0x007F, 0x007F}, // DELETE +} + +// TableC2_1 represents RFC-3454 Table C.2.1. +var TableC2_1 Set = tableC2_1 + +var tableC2_2 = Set{ + RuneRange{0x0080, 0x009F}, // [CONTROL CHARACTERS] + RuneRange{0x06DD, 0x06DD}, // ARABIC END OF AYAH + RuneRange{0x070F, 0x070F}, // SYRIAC ABBREVIATION MARK + RuneRange{0x180E, 0x180E}, // MONGOLIAN VOWEL SEPARATOR + RuneRange{0x200C, 0x200C}, // ZERO WIDTH NON-JOINER + RuneRange{0x200D, 0x200D}, // ZERO WIDTH JOINER + RuneRange{0x2028, 0x2028}, // LINE SEPARATOR + RuneRange{0x2029, 0x2029}, // PARAGRAPH SEPARATOR + RuneRange{0x2060, 0x2060}, // WORD JOINER + RuneRange{0x2061, 0x2061}, // FUNCTION APPLICATION + RuneRange{0x2062, 0x2062}, // INVISIBLE TIMES + RuneRange{0x2063, 0x2063}, // INVISIBLE SEPARATOR + RuneRange{0x206A, 0x206F}, // [CONTROL CHARACTERS] + RuneRange{0xFEFF, 0xFEFF}, // ZERO WIDTH NO-BREAK SPACE + RuneRange{0xFFF9, 0xFFFC}, // [CONTROL CHARACTERS] + RuneRange{0x1D173, 0x1D17A}, // [MUSICAL CONTROL CHARACTERS] +} + +// TableC2_2 represents RFC-3454 Table C.2.2. +var TableC2_2 Set = tableC2_2 + +var tableC3 = Set{ + RuneRange{0xE000, 0xF8FF}, // [PRIVATE USE, PLANE 0] + RuneRange{0xF0000, 0xFFFFD}, // [PRIVATE USE, PLANE 15] + RuneRange{0x100000, 0x10FFFD}, // [PRIVATE USE, PLANE 16] +} + +// TableC3 represents RFC-3454 Table C.3. +var TableC3 Set = tableC3 + +var tableC4 = Set{ + RuneRange{0xFDD0, 0xFDEF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xFFFE, 0xFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x1FFFE, 0x1FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x2FFFE, 0x2FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x3FFFE, 0x3FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x4FFFE, 0x4FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x5FFFE, 0x5FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x6FFFE, 0x6FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x7FFFE, 0x7FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x8FFFE, 0x8FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x9FFFE, 0x9FFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xAFFFE, 0xAFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xBFFFE, 0xBFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xCFFFE, 0xCFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xDFFFE, 0xDFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xEFFFE, 0xEFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0xFFFFE, 0xFFFFF}, // [NONCHARACTER CODE POINTS] + RuneRange{0x10FFFE, 0x10FFFF}, // [NONCHARACTER CODE POINTS] +} + +// TableC4 represents RFC-3454 Table C.4. +var TableC4 Set = tableC4 + +var tableC5 = Set{ + RuneRange{0xD800, 0xDFFF}, // [SURROGATE CODES] +} + +// TableC5 represents RFC-3454 Table C.5. +var TableC5 Set = tableC5 + +var tableC6 = Set{ + RuneRange{0xFFF9, 0xFFF9}, // INTERLINEAR ANNOTATION ANCHOR + RuneRange{0xFFFA, 0xFFFA}, // INTERLINEAR ANNOTATION SEPARATOR + RuneRange{0xFFFB, 0xFFFB}, // INTERLINEAR ANNOTATION TERMINATOR + RuneRange{0xFFFC, 0xFFFC}, // OBJECT REPLACEMENT CHARACTER + RuneRange{0xFFFD, 0xFFFD}, // REPLACEMENT CHARACTER +} + +// TableC6 represents RFC-3454 Table C.6. +var TableC6 Set = tableC6 + +var tableC7 = Set{ + RuneRange{0x2FF0, 0x2FFB}, // [IDEOGRAPHIC DESCRIPTION CHARACTERS] +} + +// TableC7 represents RFC-3454 Table C.7. +var TableC7 Set = tableC7 + +var tableC8 = Set{ + RuneRange{0x0340, 0x0340}, // COMBINING GRAVE TONE MARK + RuneRange{0x0341, 0x0341}, // COMBINING ACUTE TONE MARK + RuneRange{0x200E, 0x200E}, // LEFT-TO-RIGHT MARK + RuneRange{0x200F, 0x200F}, // RIGHT-TO-LEFT MARK + RuneRange{0x202A, 0x202A}, // LEFT-TO-RIGHT EMBEDDING + RuneRange{0x202B, 0x202B}, // RIGHT-TO-LEFT EMBEDDING + RuneRange{0x202C, 0x202C}, // POP DIRECTIONAL FORMATTING + RuneRange{0x202D, 0x202D}, // LEFT-TO-RIGHT OVERRIDE + RuneRange{0x202E, 0x202E}, // RIGHT-TO-LEFT OVERRIDE + RuneRange{0x206A, 0x206A}, // INHIBIT SYMMETRIC SWAPPING + RuneRange{0x206B, 0x206B}, // ACTIVATE SYMMETRIC SWAPPING + RuneRange{0x206C, 0x206C}, // INHIBIT ARABIC FORM SHAPING + RuneRange{0x206D, 0x206D}, // ACTIVATE ARABIC FORM SHAPING + RuneRange{0x206E, 0x206E}, // NATIONAL DIGIT SHAPES + RuneRange{0x206F, 0x206F}, // NOMINAL DIGIT SHAPES +} + +// TableC8 represents RFC-3454 Table C.8. +var TableC8 Set = tableC8 + +var tableC9 = Set{ + RuneRange{0xE0001, 0xE0001}, // LANGUAGE TAG + RuneRange{0xE0020, 0xE007F}, // [TAGGING CHARACTERS] +} + +// TableC9 represents RFC-3454 Table C.9. +var TableC9 Set = tableC9 + +var tableD1 = Set{ + RuneRange{0x05BE, 0x05BE}, + RuneRange{0x05C0, 0x05C0}, + RuneRange{0x05C3, 0x05C3}, + RuneRange{0x05D0, 0x05EA}, + RuneRange{0x05F0, 0x05F4}, + RuneRange{0x061B, 0x061B}, + RuneRange{0x061F, 0x061F}, + RuneRange{0x0621, 0x063A}, + RuneRange{0x0640, 0x064A}, + RuneRange{0x066D, 0x066F}, + RuneRange{0x0671, 0x06D5}, + RuneRange{0x06DD, 0x06DD}, + RuneRange{0x06E5, 0x06E6}, + RuneRange{0x06FA, 0x06FE}, + RuneRange{0x0700, 0x070D}, + RuneRange{0x0710, 0x0710}, + RuneRange{0x0712, 0x072C}, + RuneRange{0x0780, 0x07A5}, + RuneRange{0x07B1, 0x07B1}, + RuneRange{0x200F, 0x200F}, + RuneRange{0xFB1D, 0xFB1D}, + RuneRange{0xFB1F, 0xFB28}, + RuneRange{0xFB2A, 0xFB36}, + RuneRange{0xFB38, 0xFB3C}, + RuneRange{0xFB3E, 0xFB3E}, + RuneRange{0xFB40, 0xFB41}, + RuneRange{0xFB43, 0xFB44}, + RuneRange{0xFB46, 0xFBB1}, + RuneRange{0xFBD3, 0xFD3D}, + RuneRange{0xFD50, 0xFD8F}, + RuneRange{0xFD92, 0xFDC7}, + RuneRange{0xFDF0, 0xFDFC}, + RuneRange{0xFE70, 0xFE74}, + RuneRange{0xFE76, 0xFEFC}, +} + +// TableD1 represents RFC-3454 Table D.1. +var TableD1 Set = tableD1 + +var tableD2 = Set{ + RuneRange{0x0041, 0x005A}, + RuneRange{0x0061, 0x007A}, + RuneRange{0x00AA, 0x00AA}, + RuneRange{0x00B5, 0x00B5}, + RuneRange{0x00BA, 0x00BA}, + RuneRange{0x00C0, 0x00D6}, + RuneRange{0x00D8, 0x00F6}, + RuneRange{0x00F8, 0x0220}, + RuneRange{0x0222, 0x0233}, + RuneRange{0x0250, 0x02AD}, + RuneRange{0x02B0, 0x02B8}, + RuneRange{0x02BB, 0x02C1}, + RuneRange{0x02D0, 0x02D1}, + RuneRange{0x02E0, 0x02E4}, + RuneRange{0x02EE, 0x02EE}, + RuneRange{0x037A, 0x037A}, + RuneRange{0x0386, 0x0386}, + RuneRange{0x0388, 0x038A}, + RuneRange{0x038C, 0x038C}, + RuneRange{0x038E, 0x03A1}, + RuneRange{0x03A3, 0x03CE}, + RuneRange{0x03D0, 0x03F5}, + RuneRange{0x0400, 0x0482}, + RuneRange{0x048A, 0x04CE}, + RuneRange{0x04D0, 0x04F5}, + RuneRange{0x04F8, 0x04F9}, + RuneRange{0x0500, 0x050F}, + RuneRange{0x0531, 0x0556}, + RuneRange{0x0559, 0x055F}, + RuneRange{0x0561, 0x0587}, + RuneRange{0x0589, 0x0589}, + RuneRange{0x0903, 0x0903}, + RuneRange{0x0905, 0x0939}, + RuneRange{0x093D, 0x0940}, + RuneRange{0x0949, 0x094C}, + RuneRange{0x0950, 0x0950}, + RuneRange{0x0958, 0x0961}, + RuneRange{0x0964, 0x0970}, + RuneRange{0x0982, 0x0983}, + RuneRange{0x0985, 0x098C}, + RuneRange{0x098F, 0x0990}, + RuneRange{0x0993, 0x09A8}, + RuneRange{0x09AA, 0x09B0}, + RuneRange{0x09B2, 0x09B2}, + RuneRange{0x09B6, 0x09B9}, + RuneRange{0x09BE, 0x09C0}, + RuneRange{0x09C7, 0x09C8}, + RuneRange{0x09CB, 0x09CC}, + RuneRange{0x09D7, 0x09D7}, + RuneRange{0x09DC, 0x09DD}, + RuneRange{0x09DF, 0x09E1}, + RuneRange{0x09E6, 0x09F1}, + RuneRange{0x09F4, 0x09FA}, + RuneRange{0x0A05, 0x0A0A}, + RuneRange{0x0A0F, 0x0A10}, + RuneRange{0x0A13, 0x0A28}, + RuneRange{0x0A2A, 0x0A30}, + RuneRange{0x0A32, 0x0A33}, + RuneRange{0x0A35, 0x0A36}, + RuneRange{0x0A38, 0x0A39}, + RuneRange{0x0A3E, 0x0A40}, + RuneRange{0x0A59, 0x0A5C}, + RuneRange{0x0A5E, 0x0A5E}, + RuneRange{0x0A66, 0x0A6F}, + RuneRange{0x0A72, 0x0A74}, + RuneRange{0x0A83, 0x0A83}, + RuneRange{0x0A85, 0x0A8B}, + RuneRange{0x0A8D, 0x0A8D}, + RuneRange{0x0A8F, 0x0A91}, + RuneRange{0x0A93, 0x0AA8}, + RuneRange{0x0AAA, 0x0AB0}, + RuneRange{0x0AB2, 0x0AB3}, + RuneRange{0x0AB5, 0x0AB9}, + RuneRange{0x0ABD, 0x0AC0}, + RuneRange{0x0AC9, 0x0AC9}, + RuneRange{0x0ACB, 0x0ACC}, + RuneRange{0x0AD0, 0x0AD0}, + RuneRange{0x0AE0, 0x0AE0}, + RuneRange{0x0AE6, 0x0AEF}, + RuneRange{0x0B02, 0x0B03}, + RuneRange{0x0B05, 0x0B0C}, + RuneRange{0x0B0F, 0x0B10}, + RuneRange{0x0B13, 0x0B28}, + RuneRange{0x0B2A, 0x0B30}, + RuneRange{0x0B32, 0x0B33}, + RuneRange{0x0B36, 0x0B39}, + RuneRange{0x0B3D, 0x0B3E}, + RuneRange{0x0B40, 0x0B40}, + RuneRange{0x0B47, 0x0B48}, + RuneRange{0x0B4B, 0x0B4C}, + RuneRange{0x0B57, 0x0B57}, + RuneRange{0x0B5C, 0x0B5D}, + RuneRange{0x0B5F, 0x0B61}, + RuneRange{0x0B66, 0x0B70}, + RuneRange{0x0B83, 0x0B83}, + RuneRange{0x0B85, 0x0B8A}, + RuneRange{0x0B8E, 0x0B90}, + RuneRange{0x0B92, 0x0B95}, + RuneRange{0x0B99, 0x0B9A}, + RuneRange{0x0B9C, 0x0B9C}, + RuneRange{0x0B9E, 0x0B9F}, + RuneRange{0x0BA3, 0x0BA4}, + RuneRange{0x0BA8, 0x0BAA}, + RuneRange{0x0BAE, 0x0BB5}, + RuneRange{0x0BB7, 0x0BB9}, + RuneRange{0x0BBE, 0x0BBF}, + RuneRange{0x0BC1, 0x0BC2}, + RuneRange{0x0BC6, 0x0BC8}, + RuneRange{0x0BCA, 0x0BCC}, + RuneRange{0x0BD7, 0x0BD7}, + RuneRange{0x0BE7, 0x0BF2}, + RuneRange{0x0C01, 0x0C03}, + RuneRange{0x0C05, 0x0C0C}, + RuneRange{0x0C0E, 0x0C10}, + RuneRange{0x0C12, 0x0C28}, + RuneRange{0x0C2A, 0x0C33}, + RuneRange{0x0C35, 0x0C39}, + RuneRange{0x0C41, 0x0C44}, + RuneRange{0x0C60, 0x0C61}, + RuneRange{0x0C66, 0x0C6F}, + RuneRange{0x0C82, 0x0C83}, + RuneRange{0x0C85, 0x0C8C}, + RuneRange{0x0C8E, 0x0C90}, + RuneRange{0x0C92, 0x0CA8}, + RuneRange{0x0CAA, 0x0CB3}, + RuneRange{0x0CB5, 0x0CB9}, + RuneRange{0x0CBE, 0x0CBE}, + RuneRange{0x0CC0, 0x0CC4}, + RuneRange{0x0CC7, 0x0CC8}, + RuneRange{0x0CCA, 0x0CCB}, + RuneRange{0x0CD5, 0x0CD6}, + RuneRange{0x0CDE, 0x0CDE}, + RuneRange{0x0CE0, 0x0CE1}, + RuneRange{0x0CE6, 0x0CEF}, + RuneRange{0x0D02, 0x0D03}, + RuneRange{0x0D05, 0x0D0C}, + RuneRange{0x0D0E, 0x0D10}, + RuneRange{0x0D12, 0x0D28}, + RuneRange{0x0D2A, 0x0D39}, + RuneRange{0x0D3E, 0x0D40}, + RuneRange{0x0D46, 0x0D48}, + RuneRange{0x0D4A, 0x0D4C}, + RuneRange{0x0D57, 0x0D57}, + RuneRange{0x0D60, 0x0D61}, + RuneRange{0x0D66, 0x0D6F}, + RuneRange{0x0D82, 0x0D83}, + RuneRange{0x0D85, 0x0D96}, + RuneRange{0x0D9A, 0x0DB1}, + RuneRange{0x0DB3, 0x0DBB}, + RuneRange{0x0DBD, 0x0DBD}, + RuneRange{0x0DC0, 0x0DC6}, + RuneRange{0x0DCF, 0x0DD1}, + RuneRange{0x0DD8, 0x0DDF}, + RuneRange{0x0DF2, 0x0DF4}, + RuneRange{0x0E01, 0x0E30}, + RuneRange{0x0E32, 0x0E33}, + RuneRange{0x0E40, 0x0E46}, + RuneRange{0x0E4F, 0x0E5B}, + RuneRange{0x0E81, 0x0E82}, + RuneRange{0x0E84, 0x0E84}, + RuneRange{0x0E87, 0x0E88}, + RuneRange{0x0E8A, 0x0E8A}, + RuneRange{0x0E8D, 0x0E8D}, + RuneRange{0x0E94, 0x0E97}, + RuneRange{0x0E99, 0x0E9F}, + RuneRange{0x0EA1, 0x0EA3}, + RuneRange{0x0EA5, 0x0EA5}, + RuneRange{0x0EA7, 0x0EA7}, + RuneRange{0x0EAA, 0x0EAB}, + RuneRange{0x0EAD, 0x0EB0}, + RuneRange{0x0EB2, 0x0EB3}, + RuneRange{0x0EBD, 0x0EBD}, + RuneRange{0x0EC0, 0x0EC4}, + RuneRange{0x0EC6, 0x0EC6}, + RuneRange{0x0ED0, 0x0ED9}, + RuneRange{0x0EDC, 0x0EDD}, + RuneRange{0x0F00, 0x0F17}, + RuneRange{0x0F1A, 0x0F34}, + RuneRange{0x0F36, 0x0F36}, + RuneRange{0x0F38, 0x0F38}, + RuneRange{0x0F3E, 0x0F47}, + RuneRange{0x0F49, 0x0F6A}, + RuneRange{0x0F7F, 0x0F7F}, + RuneRange{0x0F85, 0x0F85}, + RuneRange{0x0F88, 0x0F8B}, + RuneRange{0x0FBE, 0x0FC5}, + RuneRange{0x0FC7, 0x0FCC}, + RuneRange{0x0FCF, 0x0FCF}, + RuneRange{0x1000, 0x1021}, + RuneRange{0x1023, 0x1027}, + RuneRange{0x1029, 0x102A}, + RuneRange{0x102C, 0x102C}, + RuneRange{0x1031, 0x1031}, + RuneRange{0x1038, 0x1038}, + RuneRange{0x1040, 0x1057}, + RuneRange{0x10A0, 0x10C5}, + RuneRange{0x10D0, 0x10F8}, + RuneRange{0x10FB, 0x10FB}, + RuneRange{0x1100, 0x1159}, + RuneRange{0x115F, 0x11A2}, + RuneRange{0x11A8, 0x11F9}, + RuneRange{0x1200, 0x1206}, + RuneRange{0x1208, 0x1246}, + RuneRange{0x1248, 0x1248}, + RuneRange{0x124A, 0x124D}, + RuneRange{0x1250, 0x1256}, + RuneRange{0x1258, 0x1258}, + RuneRange{0x125A, 0x125D}, + RuneRange{0x1260, 0x1286}, + RuneRange{0x1288, 0x1288}, + RuneRange{0x128A, 0x128D}, + RuneRange{0x1290, 0x12AE}, + RuneRange{0x12B0, 0x12B0}, + RuneRange{0x12B2, 0x12B5}, + RuneRange{0x12B8, 0x12BE}, + RuneRange{0x12C0, 0x12C0}, + RuneRange{0x12C2, 0x12C5}, + RuneRange{0x12C8, 0x12CE}, + RuneRange{0x12D0, 0x12D6}, + RuneRange{0x12D8, 0x12EE}, + RuneRange{0x12F0, 0x130E}, + RuneRange{0x1310, 0x1310}, + RuneRange{0x1312, 0x1315}, + RuneRange{0x1318, 0x131E}, + RuneRange{0x1320, 0x1346}, + RuneRange{0x1348, 0x135A}, + RuneRange{0x1361, 0x137C}, + RuneRange{0x13A0, 0x13F4}, + RuneRange{0x1401, 0x1676}, + RuneRange{0x1681, 0x169A}, + RuneRange{0x16A0, 0x16F0}, + RuneRange{0x1700, 0x170C}, + RuneRange{0x170E, 0x1711}, + RuneRange{0x1720, 0x1731}, + RuneRange{0x1735, 0x1736}, + RuneRange{0x1740, 0x1751}, + RuneRange{0x1760, 0x176C}, + RuneRange{0x176E, 0x1770}, + RuneRange{0x1780, 0x17B6}, + RuneRange{0x17BE, 0x17C5}, + RuneRange{0x17C7, 0x17C8}, + RuneRange{0x17D4, 0x17DA}, + RuneRange{0x17DC, 0x17DC}, + RuneRange{0x17E0, 0x17E9}, + RuneRange{0x1810, 0x1819}, + RuneRange{0x1820, 0x1877}, + RuneRange{0x1880, 0x18A8}, + RuneRange{0x1E00, 0x1E9B}, + RuneRange{0x1EA0, 0x1EF9}, + RuneRange{0x1F00, 0x1F15}, + RuneRange{0x1F18, 0x1F1D}, + RuneRange{0x1F20, 0x1F45}, + RuneRange{0x1F48, 0x1F4D}, + RuneRange{0x1F50, 0x1F57}, + RuneRange{0x1F59, 0x1F59}, + RuneRange{0x1F5B, 0x1F5B}, + RuneRange{0x1F5D, 0x1F5D}, + RuneRange{0x1F5F, 0x1F7D}, + RuneRange{0x1F80, 0x1FB4}, + RuneRange{0x1FB6, 0x1FBC}, + RuneRange{0x1FBE, 0x1FBE}, + RuneRange{0x1FC2, 0x1FC4}, + RuneRange{0x1FC6, 0x1FCC}, + RuneRange{0x1FD0, 0x1FD3}, + RuneRange{0x1FD6, 0x1FDB}, + RuneRange{0x1FE0, 0x1FEC}, + RuneRange{0x1FF2, 0x1FF4}, + RuneRange{0x1FF6, 0x1FFC}, + RuneRange{0x200E, 0x200E}, + RuneRange{0x2071, 0x2071}, + RuneRange{0x207F, 0x207F}, + RuneRange{0x2102, 0x2102}, + RuneRange{0x2107, 0x2107}, + RuneRange{0x210A, 0x2113}, + RuneRange{0x2115, 0x2115}, + RuneRange{0x2119, 0x211D}, + RuneRange{0x2124, 0x2124}, + RuneRange{0x2126, 0x2126}, + RuneRange{0x2128, 0x2128}, + RuneRange{0x212A, 0x212D}, + RuneRange{0x212F, 0x2131}, + RuneRange{0x2133, 0x2139}, + RuneRange{0x213D, 0x213F}, + RuneRange{0x2145, 0x2149}, + RuneRange{0x2160, 0x2183}, + RuneRange{0x2336, 0x237A}, + RuneRange{0x2395, 0x2395}, + RuneRange{0x249C, 0x24E9}, + RuneRange{0x3005, 0x3007}, + RuneRange{0x3021, 0x3029}, + RuneRange{0x3031, 0x3035}, + RuneRange{0x3038, 0x303C}, + RuneRange{0x3041, 0x3096}, + RuneRange{0x309D, 0x309F}, + RuneRange{0x30A1, 0x30FA}, + RuneRange{0x30FC, 0x30FF}, + RuneRange{0x3105, 0x312C}, + RuneRange{0x3131, 0x318E}, + RuneRange{0x3190, 0x31B7}, + RuneRange{0x31F0, 0x321C}, + RuneRange{0x3220, 0x3243}, + RuneRange{0x3260, 0x327B}, + RuneRange{0x327F, 0x32B0}, + RuneRange{0x32C0, 0x32CB}, + RuneRange{0x32D0, 0x32FE}, + RuneRange{0x3300, 0x3376}, + RuneRange{0x337B, 0x33DD}, + RuneRange{0x33E0, 0x33FE}, + RuneRange{0x3400, 0x4DB5}, + RuneRange{0x4E00, 0x9FA5}, + RuneRange{0xA000, 0xA48C}, + RuneRange{0xAC00, 0xD7A3}, + RuneRange{0xD800, 0xFA2D}, + RuneRange{0xFA30, 0xFA6A}, + RuneRange{0xFB00, 0xFB06}, + RuneRange{0xFB13, 0xFB17}, + RuneRange{0xFF21, 0xFF3A}, + RuneRange{0xFF41, 0xFF5A}, + RuneRange{0xFF66, 0xFFBE}, + RuneRange{0xFFC2, 0xFFC7}, + RuneRange{0xFFCA, 0xFFCF}, + RuneRange{0xFFD2, 0xFFD7}, + RuneRange{0xFFDA, 0xFFDC}, + RuneRange{0x10300, 0x1031E}, + RuneRange{0x10320, 0x10323}, + RuneRange{0x10330, 0x1034A}, + RuneRange{0x10400, 0x10425}, + RuneRange{0x10428, 0x1044D}, + RuneRange{0x1D000, 0x1D0F5}, + RuneRange{0x1D100, 0x1D126}, + RuneRange{0x1D12A, 0x1D166}, + RuneRange{0x1D16A, 0x1D172}, + RuneRange{0x1D183, 0x1D184}, + RuneRange{0x1D18C, 0x1D1A9}, + RuneRange{0x1D1AE, 0x1D1DD}, + RuneRange{0x1D400, 0x1D454}, + RuneRange{0x1D456, 0x1D49C}, + RuneRange{0x1D49E, 0x1D49F}, + RuneRange{0x1D4A2, 0x1D4A2}, + RuneRange{0x1D4A5, 0x1D4A6}, + RuneRange{0x1D4A9, 0x1D4AC}, + RuneRange{0x1D4AE, 0x1D4B9}, + RuneRange{0x1D4BB, 0x1D4BB}, + RuneRange{0x1D4BD, 0x1D4C0}, + RuneRange{0x1D4C2, 0x1D4C3}, + RuneRange{0x1D4C5, 0x1D505}, + RuneRange{0x1D507, 0x1D50A}, + RuneRange{0x1D50D, 0x1D514}, + RuneRange{0x1D516, 0x1D51C}, + RuneRange{0x1D51E, 0x1D539}, + RuneRange{0x1D53B, 0x1D53E}, + RuneRange{0x1D540, 0x1D544}, + RuneRange{0x1D546, 0x1D546}, + RuneRange{0x1D54A, 0x1D550}, + RuneRange{0x1D552, 0x1D6A3}, + RuneRange{0x1D6A8, 0x1D7C9}, + RuneRange{0x20000, 0x2A6D6}, + RuneRange{0x2F800, 0x2FA1D}, + RuneRange{0xF0000, 0xFFFFD}, + RuneRange{0x100000, 0x10FFFD}, +} + +// TableD2 represents RFC-3454 Table D.2. +var TableD2 Set = tableD2 diff --git a/vendor/github.com/youmark/pkcs8/.gitignore b/vendor/github.com/youmark/pkcs8/.gitignore new file mode 100644 index 0000000..8365624 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/.gitignore @@ -0,0 +1,23 @@ +# Compiled Object files, Static and Dynamic libs (Shared Objects) +*.o +*.a +*.so + +# Folders +_obj +_test + +# Architecture specific extensions/prefixes +*.[568vq] +[568vq].out + +*.cgo1.go +*.cgo2.c +_cgo_defun.c +_cgo_gotypes.go +_cgo_export.* + +_testmain.go + +*.exe +*.test diff --git a/vendor/github.com/youmark/pkcs8/.travis.yml b/vendor/github.com/youmark/pkcs8/.travis.yml new file mode 100644 index 0000000..3608f7d --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/.travis.yml @@ -0,0 +1,14 @@ +arch: + - amd64 + - ppc64le +language: go + +go: + - "1.10.x" + - "1.11.x" + - "1.12.x" + - "1.13.x" + - master + +script: + - go test -v ./... diff --git a/vendor/github.com/youmark/pkcs8/LICENSE b/vendor/github.com/youmark/pkcs8/LICENSE new file mode 100644 index 0000000..c939f44 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/LICENSE @@ -0,0 +1,21 @@ +The MIT License (MIT) + +Copyright (c) 2014 youmark + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. \ No newline at end of file diff --git a/vendor/github.com/youmark/pkcs8/README b/vendor/github.com/youmark/pkcs8/README new file mode 100644 index 0000000..376fcaf --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/README @@ -0,0 +1 @@ +pkcs8 package: implement PKCS#8 private key parsing and conversion as defined in RFC5208 and RFC5958 diff --git a/vendor/github.com/youmark/pkcs8/README.md b/vendor/github.com/youmark/pkcs8/README.md new file mode 100644 index 0000000..ef6c762 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/README.md @@ -0,0 +1,22 @@ +pkcs8 +=== +OpenSSL can generate private keys in both "traditional format" and PKCS#8 format. Newer applications are advised to use more secure PKCS#8 format. Go standard crypto package provides a [function](http://golang.org/pkg/crypto/x509/#ParsePKCS8PrivateKey) to parse private key in PKCS#8 format. There is a limitation to this function. It can only handle unencrypted PKCS#8 private keys. To use this function, the user has to save the private key in file without encryption, which is a bad practice to leave private keys unprotected on file systems. In addition, Go standard package lacks the functions to convert RSA/ECDSA private keys into PKCS#8 format. + +pkcs8 package fills the gap here. It implements functions to process private keys in PKCS#8 format, as defined in [RFC5208](https://tools.ietf.org/html/rfc5208) and [RFC5958](https://tools.ietf.org/html/rfc5958). It can handle both unencrypted PKCS#8 PrivateKeyInfo format and EncryptedPrivateKeyInfo format with PKCS#5 (v2.0) algorithms. + + +[**Godoc**](http://godoc.org/github.com/youmark/pkcs8) + +## Installation +Supports Go 1.10+. Release v1.1 is the last release supporting Go 1.9 + +```text +go get github.com/youmark/pkcs8 +``` +## dependency +This package depends on golang.org/x/crypto/pbkdf2 and golang.org/x/crypto/scrypt packages. Use the following command to retrieve them +```text +go get golang.org/x/crypto/pbkdf2 +go get golang.org/x/crypto/scrypt +``` + diff --git a/vendor/github.com/youmark/pkcs8/cipher.go b/vendor/github.com/youmark/pkcs8/cipher.go new file mode 100644 index 0000000..2946c93 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/cipher.go @@ -0,0 +1,60 @@ +package pkcs8 + +import ( + "bytes" + "crypto/cipher" + "encoding/asn1" +) + +type cipherWithBlock struct { + oid asn1.ObjectIdentifier + ivSize int + keySize int + newBlock func(key []byte) (cipher.Block, error) +} + +func (c cipherWithBlock) IVSize() int { + return c.ivSize +} + +func (c cipherWithBlock) KeySize() int { + return c.keySize +} + +func (c cipherWithBlock) OID() asn1.ObjectIdentifier { + return c.oid +} + +func (c cipherWithBlock) Encrypt(key, iv, plaintext []byte) ([]byte, error) { + block, err := c.newBlock(key) + if err != nil { + return nil, err + } + return cbcEncrypt(block, key, iv, plaintext) +} + +func (c cipherWithBlock) Decrypt(key, iv, ciphertext []byte) ([]byte, error) { + block, err := c.newBlock(key) + if err != nil { + return nil, err + } + return cbcDecrypt(block, key, iv, ciphertext) +} + +func cbcEncrypt(block cipher.Block, key, iv, plaintext []byte) ([]byte, error) { + mode := cipher.NewCBCEncrypter(block, iv) + paddingLen := block.BlockSize() - (len(plaintext) % block.BlockSize()) + ciphertext := make([]byte, len(plaintext)+paddingLen) + copy(ciphertext, plaintext) + copy(ciphertext[len(plaintext):], bytes.Repeat([]byte{byte(paddingLen)}, paddingLen)) + mode.CryptBlocks(ciphertext, ciphertext) + return ciphertext, nil +} + +func cbcDecrypt(block cipher.Block, key, iv, ciphertext []byte) ([]byte, error) { + mode := cipher.NewCBCDecrypter(block, iv) + plaintext := make([]byte, len(ciphertext)) + mode.CryptBlocks(plaintext, ciphertext) + // TODO: remove padding + return plaintext, nil +} diff --git a/vendor/github.com/youmark/pkcs8/cipher_3des.go b/vendor/github.com/youmark/pkcs8/cipher_3des.go new file mode 100644 index 0000000..5629664 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/cipher_3des.go @@ -0,0 +1,24 @@ +package pkcs8 + +import ( + "crypto/des" + "encoding/asn1" +) + +var ( + oidDESEDE3CBC = asn1.ObjectIdentifier{1, 2, 840, 113549, 3, 7} +) + +func init() { + RegisterCipher(oidDESEDE3CBC, func() Cipher { + return TripleDESCBC + }) +} + +// TripleDESCBC is the 168-bit key 3DES cipher in CBC mode. +var TripleDESCBC = cipherWithBlock{ + ivSize: des.BlockSize, + keySize: 24, + newBlock: des.NewTripleDESCipher, + oid: oidDESEDE3CBC, +} diff --git a/vendor/github.com/youmark/pkcs8/cipher_aes.go b/vendor/github.com/youmark/pkcs8/cipher_aes.go new file mode 100644 index 0000000..c0372d1 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/cipher_aes.go @@ -0,0 +1,84 @@ +package pkcs8 + +import ( + "crypto/aes" + "encoding/asn1" +) + +var ( + oidAES128CBC = asn1.ObjectIdentifier{2, 16, 840, 1, 101, 3, 4, 1, 2} + oidAES128GCM = asn1.ObjectIdentifier{2, 16, 840, 1, 101, 3, 4, 1, 6} + oidAES192CBC = asn1.ObjectIdentifier{2, 16, 840, 1, 101, 3, 4, 1, 22} + oidAES192GCM = asn1.ObjectIdentifier{2, 16, 840, 1, 101, 3, 4, 1, 26} + oidAES256CBC = asn1.ObjectIdentifier{2, 16, 840, 1, 101, 3, 4, 1, 42} + oidAES256GCM = asn1.ObjectIdentifier{2, 16, 840, 1, 101, 3, 4, 1, 46} +) + +func init() { + RegisterCipher(oidAES128CBC, func() Cipher { + return AES128CBC + }) + RegisterCipher(oidAES128GCM, func() Cipher { + return AES128GCM + }) + RegisterCipher(oidAES192CBC, func() Cipher { + return AES192CBC + }) + RegisterCipher(oidAES192GCM, func() Cipher { + return AES192GCM + }) + RegisterCipher(oidAES256CBC, func() Cipher { + return AES256CBC + }) + RegisterCipher(oidAES256GCM, func() Cipher { + return AES256GCM + }) +} + +// AES128CBC is the 128-bit key AES cipher in CBC mode. +var AES128CBC = cipherWithBlock{ + ivSize: aes.BlockSize, + keySize: 16, + newBlock: aes.NewCipher, + oid: oidAES128CBC, +} + +// AES128GCM is the 128-bit key AES cipher in GCM mode. +var AES128GCM = cipherWithBlock{ + ivSize: aes.BlockSize, + keySize: 16, + newBlock: aes.NewCipher, + oid: oidAES128GCM, +} + +// AES192CBC is the 192-bit key AES cipher in CBC mode. +var AES192CBC = cipherWithBlock{ + ivSize: aes.BlockSize, + keySize: 24, + newBlock: aes.NewCipher, + oid: oidAES192CBC, +} + +// AES192GCM is the 912-bit key AES cipher in GCM mode. +var AES192GCM = cipherWithBlock{ + ivSize: aes.BlockSize, + keySize: 24, + newBlock: aes.NewCipher, + oid: oidAES192GCM, +} + +// AES256CBC is the 256-bit key AES cipher in CBC mode. +var AES256CBC = cipherWithBlock{ + ivSize: aes.BlockSize, + keySize: 32, + newBlock: aes.NewCipher, + oid: oidAES256CBC, +} + +// AES256GCM is the 256-bit key AES cipher in GCM mode. +var AES256GCM = cipherWithBlock{ + ivSize: aes.BlockSize, + keySize: 32, + newBlock: aes.NewCipher, + oid: oidAES256GCM, +} diff --git a/vendor/github.com/youmark/pkcs8/kdf_pbkdf2.go b/vendor/github.com/youmark/pkcs8/kdf_pbkdf2.go new file mode 100644 index 0000000..79697dd --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/kdf_pbkdf2.go @@ -0,0 +1,91 @@ +package pkcs8 + +import ( + "crypto" + "crypto/sha1" + "crypto/sha256" + "crypto/x509/pkix" + "encoding/asn1" + "errors" + "hash" + + "golang.org/x/crypto/pbkdf2" +) + +var ( + oidPKCS5PBKDF2 = asn1.ObjectIdentifier{1, 2, 840, 113549, 1, 5, 12} + oidHMACWithSHA1 = asn1.ObjectIdentifier{1, 2, 840, 113549, 2, 7} + oidHMACWithSHA256 = asn1.ObjectIdentifier{1, 2, 840, 113549, 2, 9} +) + +func init() { + RegisterKDF(oidPKCS5PBKDF2, func() KDFParameters { + return new(pbkdf2Params) + }) +} + +func newHashFromPRF(ai pkix.AlgorithmIdentifier) (func() hash.Hash, error) { + switch { + case len(ai.Algorithm) == 0 || ai.Algorithm.Equal(oidHMACWithSHA1): + return sha1.New, nil + case ai.Algorithm.Equal(oidHMACWithSHA256): + return sha256.New, nil + default: + return nil, errors.New("pkcs8: unsupported hash function") + } +} + +func newPRFParamFromHash(h crypto.Hash) (pkix.AlgorithmIdentifier, error) { + switch h { + case crypto.SHA1: + return pkix.AlgorithmIdentifier{ + Algorithm: oidHMACWithSHA1, + Parameters: asn1.RawValue{Tag: asn1.TagNull}}, nil + case crypto.SHA256: + return pkix.AlgorithmIdentifier{ + Algorithm: oidHMACWithSHA256, + Parameters: asn1.RawValue{Tag: asn1.TagNull}}, nil + } + return pkix.AlgorithmIdentifier{}, errors.New("pkcs8: unsupported hash function") +} + +type pbkdf2Params struct { + Salt []byte + IterationCount int + PRF pkix.AlgorithmIdentifier `asn1:"optional"` +} + +func (p pbkdf2Params) DeriveKey(password []byte, size int) (key []byte, err error) { + h, err := newHashFromPRF(p.PRF) + if err != nil { + return nil, err + } + return pbkdf2.Key(password, p.Salt, p.IterationCount, size, h), nil +} + +// PBKDF2Opts contains options for the PBKDF2 key derivation function. +type PBKDF2Opts struct { + SaltSize int + IterationCount int + HMACHash crypto.Hash +} + +func (p PBKDF2Opts) DeriveKey(password, salt []byte, size int) ( + key []byte, params KDFParameters, err error) { + + key = pbkdf2.Key(password, salt, p.IterationCount, size, p.HMACHash.New) + prfParam, err := newPRFParamFromHash(p.HMACHash) + if err != nil { + return nil, nil, err + } + params = pbkdf2Params{salt, p.IterationCount, prfParam} + return key, params, nil +} + +func (p PBKDF2Opts) GetSaltSize() int { + return p.SaltSize +} + +func (p PBKDF2Opts) OID() asn1.ObjectIdentifier { + return oidPKCS5PBKDF2 +} diff --git a/vendor/github.com/youmark/pkcs8/kdf_scrypt.go b/vendor/github.com/youmark/pkcs8/kdf_scrypt.go new file mode 100644 index 0000000..36c4f4f --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/kdf_scrypt.go @@ -0,0 +1,62 @@ +package pkcs8 + +import ( + "encoding/asn1" + + "golang.org/x/crypto/scrypt" +) + +var ( + oidScrypt = asn1.ObjectIdentifier{1, 3, 6, 1, 4, 1, 11591, 4, 11} +) + +func init() { + RegisterKDF(oidScrypt, func() KDFParameters { + return new(scryptParams) + }) +} + +type scryptParams struct { + Salt []byte + CostParameter int + BlockSize int + ParallelizationParameter int +} + +func (p scryptParams) DeriveKey(password []byte, size int) (key []byte, err error) { + return scrypt.Key(password, p.Salt, p.CostParameter, p.BlockSize, + p.ParallelizationParameter, size) +} + +// ScryptOpts contains options for the scrypt key derivation function. +type ScryptOpts struct { + SaltSize int + CostParameter int + BlockSize int + ParallelizationParameter int +} + +func (p ScryptOpts) DeriveKey(password, salt []byte, size int) ( + key []byte, params KDFParameters, err error) { + + key, err = scrypt.Key(password, salt, p.CostParameter, p.BlockSize, + p.ParallelizationParameter, size) + if err != nil { + return nil, nil, err + } + params = scryptParams{ + BlockSize: p.BlockSize, + CostParameter: p.CostParameter, + ParallelizationParameter: p.ParallelizationParameter, + Salt: salt, + } + return key, params, nil +} + +func (p ScryptOpts) GetSaltSize() int { + return p.SaltSize +} + +func (p ScryptOpts) OID() asn1.ObjectIdentifier { + return oidScrypt +} diff --git a/vendor/github.com/youmark/pkcs8/pkcs8.go b/vendor/github.com/youmark/pkcs8/pkcs8.go new file mode 100644 index 0000000..f27f627 --- /dev/null +++ b/vendor/github.com/youmark/pkcs8/pkcs8.go @@ -0,0 +1,309 @@ +// Package pkcs8 implements functions to parse and convert private keys in PKCS#8 format, as defined in RFC5208 and RFC5958 +package pkcs8 + +import ( + "crypto" + "crypto/ecdsa" + "crypto/rand" + "crypto/rsa" + "crypto/x509" + "crypto/x509/pkix" + "encoding/asn1" + "errors" + "fmt" +) + +// DefaultOpts are the default options for encrypting a key if none are given. +// The defaults can be changed by the library user. +var DefaultOpts = &Opts{ + Cipher: AES256CBC, + KDFOpts: PBKDF2Opts{ + SaltSize: 8, + IterationCount: 10000, + HMACHash: crypto.SHA256, + }, +} + +// KDFOpts contains options for a key derivation function. +// An implementation of this interface must be specified when encrypting a PKCS#8 key. +type KDFOpts interface { + // DeriveKey derives a key of size bytes from the given password and salt. + // It returns the key and the ASN.1-encodable parameters used. + DeriveKey(password, salt []byte, size int) (key []byte, params KDFParameters, err error) + // GetSaltSize returns the salt size specified. + GetSaltSize() int + // OID returns the OID of the KDF specified. + OID() asn1.ObjectIdentifier +} + +// KDFParameters contains parameters (salt, etc.) for a key deriviation function. +// It must be a ASN.1-decodable structure. +// An implementation of this interface is created when decoding an encrypted PKCS#8 key. +type KDFParameters interface { + // DeriveKey derives a key of size bytes from the given password. + // It uses the salt from the decoded parameters. + DeriveKey(password []byte, size int) (key []byte, err error) +} + +var kdfs = make(map[string]func() KDFParameters) + +// RegisterKDF registers a function that returns a new instance of the given KDF +// parameters. This allows the library to support client-provided KDFs. +func RegisterKDF(oid asn1.ObjectIdentifier, params func() KDFParameters) { + kdfs[oid.String()] = params +} + +// Cipher represents a cipher for encrypting the key material. +type Cipher interface { + // IVSize returns the IV size of the cipher, in bytes. + IVSize() int + // KeySize returns the key size of the cipher, in bytes. + KeySize() int + // Encrypt encrypts the key material. + Encrypt(key, iv, plaintext []byte) ([]byte, error) + // Decrypt decrypts the key material. + Decrypt(key, iv, ciphertext []byte) ([]byte, error) + // OID returns the OID of the cipher specified. + OID() asn1.ObjectIdentifier +} + +var ciphers = make(map[string]func() Cipher) + +// RegisterCipher registers a function that returns a new instance of the given +// cipher. This allows the library to support client-provided ciphers. +func RegisterCipher(oid asn1.ObjectIdentifier, cipher func() Cipher) { + ciphers[oid.String()] = cipher +} + +// Opts contains options for encrypting a PKCS#8 key. +type Opts struct { + Cipher Cipher + KDFOpts KDFOpts +} + +// Unecrypted PKCS8 +var ( + oidPBES2 = asn1.ObjectIdentifier{1, 2, 840, 113549, 1, 5, 13} +) + +type encryptedPrivateKeyInfo struct { + EncryptionAlgorithm pkix.AlgorithmIdentifier + EncryptedData []byte +} + +type pbes2Params struct { + KeyDerivationFunc pkix.AlgorithmIdentifier + EncryptionScheme pkix.AlgorithmIdentifier +} + +type privateKeyInfo struct { + Version int + PrivateKeyAlgorithm pkix.AlgorithmIdentifier + PrivateKey []byte +} + +func parseKeyDerivationFunc(keyDerivationFunc pkix.AlgorithmIdentifier) (KDFParameters, error) { + oid := keyDerivationFunc.Algorithm.String() + newParams, ok := kdfs[oid] + if !ok { + return nil, fmt.Errorf("pkcs8: unsupported KDF (OID: %s)", oid) + } + params := newParams() + _, err := asn1.Unmarshal(keyDerivationFunc.Parameters.FullBytes, params) + if err != nil { + return nil, errors.New("pkcs8: invalid KDF parameters") + } + return params, nil +} + +func parseEncryptionScheme(encryptionScheme pkix.AlgorithmIdentifier) (Cipher, []byte, error) { + oid := encryptionScheme.Algorithm.String() + newCipher, ok := ciphers[oid] + if !ok { + return nil, nil, fmt.Errorf("pkcs8: unsupported cipher (OID: %s)", oid) + } + cipher := newCipher() + var iv []byte + if _, err := asn1.Unmarshal(encryptionScheme.Parameters.FullBytes, &iv); err != nil { + return nil, nil, errors.New("pkcs8: invalid cipher parameters") + } + return cipher, iv, nil +} + +// ParsePrivateKey parses a DER-encoded PKCS#8 private key. +// Password can be nil. +// This is equivalent to ParsePKCS8PrivateKey. +func ParsePrivateKey(der []byte, password []byte) (interface{}, KDFParameters, error) { + // No password provided, assume the private key is unencrypted + if len(password) == 0 { + privateKey, err := x509.ParsePKCS8PrivateKey(der) + return privateKey, nil, err + } + + // Use the password provided to decrypt the private key + var privKey encryptedPrivateKeyInfo + if _, err := asn1.Unmarshal(der, &privKey); err != nil { + return nil, nil, errors.New("pkcs8: only PKCS #5 v2.0 supported") + } + + if !privKey.EncryptionAlgorithm.Algorithm.Equal(oidPBES2) { + return nil, nil, errors.New("pkcs8: only PBES2 supported") + } + + var params pbes2Params + if _, err := asn1.Unmarshal(privKey.EncryptionAlgorithm.Parameters.FullBytes, ¶ms); err != nil { + return nil, nil, errors.New("pkcs8: invalid PBES2 parameters") + } + + cipher, iv, err := parseEncryptionScheme(params.EncryptionScheme) + if err != nil { + return nil, nil, err + } + + kdfParams, err := parseKeyDerivationFunc(params.KeyDerivationFunc) + if err != nil { + return nil, nil, err + } + + keySize := cipher.KeySize() + symkey, err := kdfParams.DeriveKey(password, keySize) + if err != nil { + return nil, nil, err + } + + encryptedKey := privKey.EncryptedData + decryptedKey, err := cipher.Decrypt(symkey, iv, encryptedKey) + if err != nil { + return nil, nil, err + } + + key, err := x509.ParsePKCS8PrivateKey(decryptedKey) + if err != nil { + return nil, nil, errors.New("pkcs8: incorrect password") + } + return key, kdfParams, nil +} + +// MarshalPrivateKey encodes a private key into DER-encoded PKCS#8 with the given options. +// Password can be nil. +func MarshalPrivateKey(priv interface{}, password []byte, opts *Opts) ([]byte, error) { + if len(password) == 0 { + return x509.MarshalPKCS8PrivateKey(priv) + } + + if opts == nil { + opts = DefaultOpts + } + + // Convert private key into PKCS8 format + pkey, err := x509.MarshalPKCS8PrivateKey(priv) + if err != nil { + return nil, err + } + + encAlg := opts.Cipher + salt := make([]byte, opts.KDFOpts.GetSaltSize()) + _, err = rand.Read(salt) + if err != nil { + return nil, err + } + iv := make([]byte, encAlg.IVSize()) + _, err = rand.Read(iv) + if err != nil { + return nil, err + } + key, kdfParams, err := opts.KDFOpts.DeriveKey(password, salt, encAlg.KeySize()) + if err != nil { + return nil, err + } + + encryptedKey, err := encAlg.Encrypt(key, iv, pkey) + if err != nil { + return nil, err + } + + marshalledParams, err := asn1.Marshal(kdfParams) + if err != nil { + return nil, err + } + keyDerivationFunc := pkix.AlgorithmIdentifier{ + Algorithm: opts.KDFOpts.OID(), + Parameters: asn1.RawValue{FullBytes: marshalledParams}, + } + marshalledIV, err := asn1.Marshal(iv) + if err != nil { + return nil, err + } + encryptionScheme := pkix.AlgorithmIdentifier{ + Algorithm: encAlg.OID(), + Parameters: asn1.RawValue{FullBytes: marshalledIV}, + } + + encryptionAlgorithmParams := pbes2Params{ + EncryptionScheme: encryptionScheme, + KeyDerivationFunc: keyDerivationFunc, + } + marshalledEncryptionAlgorithmParams, err := asn1.Marshal(encryptionAlgorithmParams) + if err != nil { + return nil, err + } + encryptionAlgorithm := pkix.AlgorithmIdentifier{ + Algorithm: oidPBES2, + Parameters: asn1.RawValue{FullBytes: marshalledEncryptionAlgorithmParams}, + } + + encryptedPkey := encryptedPrivateKeyInfo{ + EncryptionAlgorithm: encryptionAlgorithm, + EncryptedData: encryptedKey, + } + + return asn1.Marshal(encryptedPkey) +} + +// ParsePKCS8PrivateKey parses encrypted/unencrypted private keys in PKCS#8 format. To parse encrypted private keys, a password of []byte type should be provided to the function as the second parameter. +func ParsePKCS8PrivateKey(der []byte, v ...[]byte) (interface{}, error) { + var password []byte + if len(v) > 0 { + password = v[0] + } + privateKey, _, err := ParsePrivateKey(der, password) + return privateKey, err +} + +// ParsePKCS8PrivateKeyRSA parses encrypted/unencrypted private keys in PKCS#8 format. To parse encrypted private keys, a password of []byte type should be provided to the function as the second parameter. +func ParsePKCS8PrivateKeyRSA(der []byte, v ...[]byte) (*rsa.PrivateKey, error) { + key, err := ParsePKCS8PrivateKey(der, v...) + if err != nil { + return nil, err + } + typedKey, ok := key.(*rsa.PrivateKey) + if !ok { + return nil, errors.New("key block is not of type RSA") + } + return typedKey, nil +} + +// ParsePKCS8PrivateKeyECDSA parses encrypted/unencrypted private keys in PKCS#8 format. To parse encrypted private keys, a password of []byte type should be provided to the function as the second parameter. +func ParsePKCS8PrivateKeyECDSA(der []byte, v ...[]byte) (*ecdsa.PrivateKey, error) { + key, err := ParsePKCS8PrivateKey(der, v...) + if err != nil { + return nil, err + } + typedKey, ok := key.(*ecdsa.PrivateKey) + if !ok { + return nil, errors.New("key block is not of type ECDSA") + } + return typedKey, nil +} + +// ConvertPrivateKeyToPKCS8 converts the private key into PKCS#8 format. +// To encrypt the private key, the password of []byte type should be provided as the second parameter. +// +// The only supported key types are RSA and ECDSA (*rsa.PrivateKey or *ecdsa.PrivateKey for priv) +func ConvertPrivateKeyToPKCS8(priv interface{}, v ...[]byte) ([]byte, error) { + var password []byte + if len(v) > 0 { + password = v[0] + } + return MarshalPrivateKey(priv, password, nil) +} diff --git a/vendor/go.mongodb.org/mongo-driver/LICENSE b/vendor/go.mongodb.org/mongo-driver/LICENSE new file mode 100644 index 0000000..261eeb9 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/LICENSE @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright [yyyy] [name of copyright owner] + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bson.go b/vendor/go.mongodb.org/mongo-driver/bson/bson.go new file mode 100644 index 0000000..95ffc10 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bson.go @@ -0,0 +1,50 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 +// +// Based on gopkg.in/mgo.v2/bson by Gustavo Niemeyer +// See THIRD-PARTY-NOTICES for original license terms. + +package bson // import "go.mongodb.org/mongo-driver/bson" + +import ( + "go.mongodb.org/mongo-driver/bson/primitive" +) + +// Zeroer allows custom struct types to implement a report of zero +// state. All struct types that don't implement Zeroer or where IsZero +// returns false are considered to be not zero. +type Zeroer interface { + IsZero() bool +} + +// D is an ordered representation of a BSON document. This type should be used when the order of the elements matters, +// such as MongoDB command documents. If the order of the elements does not matter, an M should be used instead. +// +// A D should not be constructed with duplicate key names, as that can cause undefined server behavior. +// +// Example usage: +// +// bson.D{{"foo", "bar"}, {"hello", "world"}, {"pi", 3.14159}} +type D = primitive.D + +// E represents a BSON element for a D. It is usually used inside a D. +type E = primitive.E + +// M is an unordered representation of a BSON document. This type should be used when the order of the elements does not +// matter. This type is handled as a regular map[string]interface{} when encoding and decoding. Elements will be +// serialized in an undefined, random order. If the order of the elements matters, a D should be used instead. +// +// Example usage: +// +// bson.M{"foo": "bar", "hello": "world", "pi": 3.14159} +type M = primitive.M + +// An A is an ordered representation of a BSON array. +// +// Example usage: +// +// bson.A{"bar", "world", 3.14159, bson.D{{"qux", 12345}}} +type A = primitive.A diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/array_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/array_codec.go new file mode 100644 index 0000000..4e24f9e --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/array_codec.go @@ -0,0 +1,50 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/x/bsonx/bsoncore" +) + +// ArrayCodec is the Codec used for bsoncore.Array values. +type ArrayCodec struct{} + +var defaultArrayCodec = NewArrayCodec() + +// NewArrayCodec returns an ArrayCodec. +func NewArrayCodec() *ArrayCodec { + return &ArrayCodec{} +} + +// EncodeValue is the ValueEncoder for bsoncore.Array values. +func (ac *ArrayCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tCoreArray { + return ValueEncoderError{Name: "CoreArrayEncodeValue", Types: []reflect.Type{tCoreArray}, Received: val} + } + + arr := val.Interface().(bsoncore.Array) + return bsonrw.Copier{}.CopyArrayFromBytes(vw, arr) +} + +// DecodeValue is the ValueDecoder for bsoncore.Array values. +func (ac *ArrayCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tCoreArray { + return ValueDecoderError{Name: "CoreArrayDecodeValue", Types: []reflect.Type{tCoreArray}, Received: val} + } + + if val.IsNil() { + val.Set(reflect.MakeSlice(val.Type(), 0, 0)) + } + + val.SetLen(0) + arr, err := bsonrw.Copier{}.AppendArrayBytes(val.Interface().(bsoncore.Array), vr) + val.Set(reflect.ValueOf(arr)) + return err +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/bsoncodec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/bsoncodec.go new file mode 100644 index 0000000..96195bc --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/bsoncodec.go @@ -0,0 +1,216 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec // import "go.mongodb.org/mongo-driver/bson/bsoncodec" + +import ( + "fmt" + "reflect" + "strings" + + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +var ( + emptyValue = reflect.Value{} +) + +// Marshaler is an interface implemented by types that can marshal themselves +// into a BSON document represented as bytes. The bytes returned must be a valid +// BSON document if the error is nil. +type Marshaler interface { + MarshalBSON() ([]byte, error) +} + +// ValueMarshaler is an interface implemented by types that can marshal +// themselves into a BSON value as bytes. The type must be the valid type for +// the bytes returned. The bytes and byte type together must be valid if the +// error is nil. +type ValueMarshaler interface { + MarshalBSONValue() (bsontype.Type, []byte, error) +} + +// Unmarshaler is an interface implemented by types that can unmarshal a BSON +// document representation of themselves. The BSON bytes can be assumed to be +// valid. UnmarshalBSON must copy the BSON bytes if it wishes to retain the data +// after returning. +type Unmarshaler interface { + UnmarshalBSON([]byte) error +} + +// ValueUnmarshaler is an interface implemented by types that can unmarshal a +// BSON value representation of themselves. The BSON bytes and type can be +// assumed to be valid. UnmarshalBSONValue must copy the BSON value bytes if it +// wishes to retain the data after returning. +type ValueUnmarshaler interface { + UnmarshalBSONValue(bsontype.Type, []byte) error +} + +// ValueEncoderError is an error returned from a ValueEncoder when the provided value can't be +// encoded by the ValueEncoder. +type ValueEncoderError struct { + Name string + Types []reflect.Type + Kinds []reflect.Kind + Received reflect.Value +} + +func (vee ValueEncoderError) Error() string { + typeKinds := make([]string, 0, len(vee.Types)+len(vee.Kinds)) + for _, t := range vee.Types { + typeKinds = append(typeKinds, t.String()) + } + for _, k := range vee.Kinds { + if k == reflect.Map { + typeKinds = append(typeKinds, "map[string]*") + continue + } + typeKinds = append(typeKinds, k.String()) + } + received := vee.Received.Kind().String() + if vee.Received.IsValid() { + received = vee.Received.Type().String() + } + return fmt.Sprintf("%s can only encode valid %s, but got %s", vee.Name, strings.Join(typeKinds, ", "), received) +} + +// ValueDecoderError is an error returned from a ValueDecoder when the provided value can't be +// decoded by the ValueDecoder. +type ValueDecoderError struct { + Name string + Types []reflect.Type + Kinds []reflect.Kind + Received reflect.Value +} + +func (vde ValueDecoderError) Error() string { + typeKinds := make([]string, 0, len(vde.Types)+len(vde.Kinds)) + for _, t := range vde.Types { + typeKinds = append(typeKinds, t.String()) + } + for _, k := range vde.Kinds { + if k == reflect.Map { + typeKinds = append(typeKinds, "map[string]*") + continue + } + typeKinds = append(typeKinds, k.String()) + } + received := vde.Received.Kind().String() + if vde.Received.IsValid() { + received = vde.Received.Type().String() + } + return fmt.Sprintf("%s can only decode valid and settable %s, but got %s", vde.Name, strings.Join(typeKinds, ", "), received) +} + +// EncodeContext is the contextual information required for a Codec to encode a +// value. +type EncodeContext struct { + *Registry + MinSize bool +} + +// DecodeContext is the contextual information required for a Codec to decode a +// value. +type DecodeContext struct { + *Registry + Truncate bool + // Ancestor is the type of a containing document. This is mainly used to determine what type + // should be used when decoding an embedded document into an empty interface. For example, if + // Ancestor is a bson.M, BSON embedded document values being decoded into an empty interface + // will be decoded into a bson.M. + Ancestor reflect.Type +} + +// ValueCodec is the interface that groups the methods to encode and decode +// values. +type ValueCodec interface { + ValueEncoder + ValueDecoder +} + +// ValueEncoder is the interface implemented by types that can handle the encoding of a value. +type ValueEncoder interface { + EncodeValue(EncodeContext, bsonrw.ValueWriter, reflect.Value) error +} + +// ValueEncoderFunc is an adapter function that allows a function with the correct signature to be +// used as a ValueEncoder. +type ValueEncoderFunc func(EncodeContext, bsonrw.ValueWriter, reflect.Value) error + +// EncodeValue implements the ValueEncoder interface. +func (fn ValueEncoderFunc) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + return fn(ec, vw, val) +} + +// ValueDecoder is the interface implemented by types that can handle the decoding of a value. +type ValueDecoder interface { + DecodeValue(DecodeContext, bsonrw.ValueReader, reflect.Value) error +} + +// ValueDecoderFunc is an adapter function that allows a function with the correct signature to be +// used as a ValueDecoder. +type ValueDecoderFunc func(DecodeContext, bsonrw.ValueReader, reflect.Value) error + +// DecodeValue implements the ValueDecoder interface. +func (fn ValueDecoderFunc) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + return fn(dc, vr, val) +} + +// typeDecoder is the interface implemented by types that can handle the decoding of a value given its type. +type typeDecoder interface { + decodeType(DecodeContext, bsonrw.ValueReader, reflect.Type) (reflect.Value, error) +} + +// typeDecoderFunc is an adapter function that allows a function with the correct signature to be used as a typeDecoder. +type typeDecoderFunc func(DecodeContext, bsonrw.ValueReader, reflect.Type) (reflect.Value, error) + +func (fn typeDecoderFunc) decodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + return fn(dc, vr, t) +} + +// decodeAdapter allows two functions with the correct signatures to be used as both a ValueDecoder and typeDecoder. +type decodeAdapter struct { + ValueDecoderFunc + typeDecoderFunc +} + +var _ ValueDecoder = decodeAdapter{} +var _ typeDecoder = decodeAdapter{} + +// decodeTypeOrValue calls decoder.decodeType is decoder is a typeDecoder. Otherwise, it allocates a new element of type +// t and calls decoder.DecodeValue on it. +func decodeTypeOrValue(decoder ValueDecoder, dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + td, _ := decoder.(typeDecoder) + return decodeTypeOrValueWithInfo(decoder, td, dc, vr, t, true) +} + +func decodeTypeOrValueWithInfo(vd ValueDecoder, td typeDecoder, dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type, convert bool) (reflect.Value, error) { + if td != nil { + val, err := td.decodeType(dc, vr, t) + if err == nil && convert && val.Type() != t { + // This conversion step is necessary for slices and maps. If a user declares variables like: + // + // type myBool bool + // var m map[string]myBool + // + // and tries to decode BSON bytes into the map, the decoding will fail if this conversion is not present + // because we'll try to assign a value of type bool to one of type myBool. + val = val.Convert(t) + } + return val, err + } + + val := reflect.New(t).Elem() + err := vd.DecodeValue(dc, vr, val) + return val, err +} + +// CodecZeroer is the interface implemented by Codecs that can also determine if +// a value of the type that would be encoded is zero. +type CodecZeroer interface { + IsTypeZero(interface{}) bool +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/byte_slice_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/byte_slice_codec.go new file mode 100644 index 0000000..5a916cc --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/byte_slice_codec.go @@ -0,0 +1,111 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "fmt" + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +// ByteSliceCodec is the Codec used for []byte values. +type ByteSliceCodec struct { + EncodeNilAsEmpty bool +} + +var ( + defaultByteSliceCodec = NewByteSliceCodec() + + _ ValueCodec = defaultByteSliceCodec + _ typeDecoder = defaultByteSliceCodec +) + +// NewByteSliceCodec returns a StringCodec with options opts. +func NewByteSliceCodec(opts ...*bsonoptions.ByteSliceCodecOptions) *ByteSliceCodec { + byteSliceOpt := bsonoptions.MergeByteSliceCodecOptions(opts...) + codec := ByteSliceCodec{} + if byteSliceOpt.EncodeNilAsEmpty != nil { + codec.EncodeNilAsEmpty = *byteSliceOpt.EncodeNilAsEmpty + } + return &codec +} + +// EncodeValue is the ValueEncoder for []byte. +func (bsc *ByteSliceCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tByteSlice { + return ValueEncoderError{Name: "ByteSliceEncodeValue", Types: []reflect.Type{tByteSlice}, Received: val} + } + if val.IsNil() && !bsc.EncodeNilAsEmpty { + return vw.WriteNull() + } + return vw.WriteBinary(val.Interface().([]byte)) +} + +func (bsc *ByteSliceCodec) decodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tByteSlice { + return emptyValue, ValueDecoderError{ + Name: "ByteSliceDecodeValue", + Types: []reflect.Type{tByteSlice}, + Received: reflect.Zero(t), + } + } + + var data []byte + var err error + switch vrType := vr.Type(); vrType { + case bsontype.String: + str, err := vr.ReadString() + if err != nil { + return emptyValue, err + } + data = []byte(str) + case bsontype.Symbol: + sym, err := vr.ReadSymbol() + if err != nil { + return emptyValue, err + } + data = []byte(sym) + case bsontype.Binary: + var subtype byte + data, subtype, err = vr.ReadBinary() + if err != nil { + return emptyValue, err + } + if subtype != bsontype.BinaryGeneric && subtype != bsontype.BinaryBinaryOld { + return emptyValue, decodeBinaryError{subtype: subtype, typeName: "[]byte"} + } + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a []byte", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(data), nil +} + +// DecodeValue is the ValueDecoder for []byte. +func (bsc *ByteSliceCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tByteSlice { + return ValueDecoderError{Name: "ByteSliceDecodeValue", Types: []reflect.Type{tByteSlice}, Received: val} + } + + elem, err := bsc.decodeType(dc, vr, tByteSlice) + if err != nil { + return err + } + + val.Set(elem) + return nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/cond_addr_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/cond_addr_codec.go new file mode 100644 index 0000000..cb8180f --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/cond_addr_codec.go @@ -0,0 +1,63 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonrw" +) + +// condAddrEncoder is the encoder used when a pointer to the encoding value has an encoder. +type condAddrEncoder struct { + canAddrEnc ValueEncoder + elseEnc ValueEncoder +} + +var _ ValueEncoder = (*condAddrEncoder)(nil) + +// newCondAddrEncoder returns an condAddrEncoder. +func newCondAddrEncoder(canAddrEnc, elseEnc ValueEncoder) *condAddrEncoder { + encoder := condAddrEncoder{canAddrEnc: canAddrEnc, elseEnc: elseEnc} + return &encoder +} + +// EncodeValue is the ValueEncoderFunc for a value that may be addressable. +func (cae *condAddrEncoder) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if val.CanAddr() { + return cae.canAddrEnc.EncodeValue(ec, vw, val) + } + if cae.elseEnc != nil { + return cae.elseEnc.EncodeValue(ec, vw, val) + } + return ErrNoEncoder{Type: val.Type()} +} + +// condAddrDecoder is the decoder used when a pointer to the value has a decoder. +type condAddrDecoder struct { + canAddrDec ValueDecoder + elseDec ValueDecoder +} + +var _ ValueDecoder = (*condAddrDecoder)(nil) + +// newCondAddrDecoder returns an CondAddrDecoder. +func newCondAddrDecoder(canAddrDec, elseDec ValueDecoder) *condAddrDecoder { + decoder := condAddrDecoder{canAddrDec: canAddrDec, elseDec: elseDec} + return &decoder +} + +// DecodeValue is the ValueDecoderFunc for a value that may be addressable. +func (cad *condAddrDecoder) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if val.CanAddr() { + return cad.canAddrDec.DecodeValue(dc, vr, val) + } + if cad.elseDec != nil { + return cad.elseDec.DecodeValue(dc, vr, val) + } + return ErrNoDecoder{Type: val.Type()} +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/default_value_decoders.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/default_value_decoders.go new file mode 100644 index 0000000..20f4797 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/default_value_decoders.go @@ -0,0 +1,1729 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "encoding/json" + "errors" + "fmt" + "math" + "net/url" + "reflect" + "strconv" + "time" + + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" + "go.mongodb.org/mongo-driver/x/bsonx/bsoncore" +) + +var ( + defaultValueDecoders DefaultValueDecoders + errCannotTruncate = errors.New("float64 can only be truncated to an integer type when truncation is enabled") +) + +type decodeBinaryError struct { + subtype byte + typeName string +} + +func (d decodeBinaryError) Error() string { + return fmt.Sprintf("only binary values with subtype 0x00 or 0x02 can be decoded into %s, but got subtype %v", d.typeName, d.subtype) +} + +func newDefaultStructCodec() *StructCodec { + codec, err := NewStructCodec(DefaultStructTagParser) + if err != nil { + // This function is called from the codec registration path, so errors can't be propagated. If there's an error + // constructing the StructCodec, we panic to avoid losing it. + panic(fmt.Errorf("error creating default StructCodec: %v", err)) + } + return codec +} + +// DefaultValueDecoders is a namespace type for the default ValueDecoders used +// when creating a registry. +type DefaultValueDecoders struct{} + +// RegisterDefaultDecoders will register the decoder methods attached to DefaultValueDecoders with +// the provided RegistryBuilder. +// +// There is no support for decoding map[string]interface{} because there is no decoder for +// interface{}, so users must either register this decoder themselves or use the +// EmptyInterfaceDecoder available in the bson package. +func (dvd DefaultValueDecoders) RegisterDefaultDecoders(rb *RegistryBuilder) { + if rb == nil { + panic(errors.New("argument to RegisterDefaultDecoders must not be nil")) + } + + intDecoder := decodeAdapter{dvd.IntDecodeValue, dvd.intDecodeType} + floatDecoder := decodeAdapter{dvd.FloatDecodeValue, dvd.floatDecodeType} + + rb. + RegisterTypeDecoder(tD, ValueDecoderFunc(dvd.DDecodeValue)). + RegisterTypeDecoder(tBinary, decodeAdapter{dvd.BinaryDecodeValue, dvd.binaryDecodeType}). + RegisterTypeDecoder(tUndefined, decodeAdapter{dvd.UndefinedDecodeValue, dvd.undefinedDecodeType}). + RegisterTypeDecoder(tDateTime, decodeAdapter{dvd.DateTimeDecodeValue, dvd.dateTimeDecodeType}). + RegisterTypeDecoder(tNull, decodeAdapter{dvd.NullDecodeValue, dvd.nullDecodeType}). + RegisterTypeDecoder(tRegex, decodeAdapter{dvd.RegexDecodeValue, dvd.regexDecodeType}). + RegisterTypeDecoder(tDBPointer, decodeAdapter{dvd.DBPointerDecodeValue, dvd.dBPointerDecodeType}). + RegisterTypeDecoder(tTimestamp, decodeAdapter{dvd.TimestampDecodeValue, dvd.timestampDecodeType}). + RegisterTypeDecoder(tMinKey, decodeAdapter{dvd.MinKeyDecodeValue, dvd.minKeyDecodeType}). + RegisterTypeDecoder(tMaxKey, decodeAdapter{dvd.MaxKeyDecodeValue, dvd.maxKeyDecodeType}). + RegisterTypeDecoder(tJavaScript, decodeAdapter{dvd.JavaScriptDecodeValue, dvd.javaScriptDecodeType}). + RegisterTypeDecoder(tSymbol, decodeAdapter{dvd.SymbolDecodeValue, dvd.symbolDecodeType}). + RegisterTypeDecoder(tByteSlice, defaultByteSliceCodec). + RegisterTypeDecoder(tTime, defaultTimeCodec). + RegisterTypeDecoder(tEmpty, defaultEmptyInterfaceCodec). + RegisterTypeDecoder(tCoreArray, defaultArrayCodec). + RegisterTypeDecoder(tOID, decodeAdapter{dvd.ObjectIDDecodeValue, dvd.objectIDDecodeType}). + RegisterTypeDecoder(tDecimal, decodeAdapter{dvd.Decimal128DecodeValue, dvd.decimal128DecodeType}). + RegisterTypeDecoder(tJSONNumber, decodeAdapter{dvd.JSONNumberDecodeValue, dvd.jsonNumberDecodeType}). + RegisterTypeDecoder(tURL, decodeAdapter{dvd.URLDecodeValue, dvd.urlDecodeType}). + RegisterTypeDecoder(tCoreDocument, ValueDecoderFunc(dvd.CoreDocumentDecodeValue)). + RegisterTypeDecoder(tCodeWithScope, decodeAdapter{dvd.CodeWithScopeDecodeValue, dvd.codeWithScopeDecodeType}). + RegisterDefaultDecoder(reflect.Bool, decodeAdapter{dvd.BooleanDecodeValue, dvd.booleanDecodeType}). + RegisterDefaultDecoder(reflect.Int, intDecoder). + RegisterDefaultDecoder(reflect.Int8, intDecoder). + RegisterDefaultDecoder(reflect.Int16, intDecoder). + RegisterDefaultDecoder(reflect.Int32, intDecoder). + RegisterDefaultDecoder(reflect.Int64, intDecoder). + RegisterDefaultDecoder(reflect.Uint, defaultUIntCodec). + RegisterDefaultDecoder(reflect.Uint8, defaultUIntCodec). + RegisterDefaultDecoder(reflect.Uint16, defaultUIntCodec). + RegisterDefaultDecoder(reflect.Uint32, defaultUIntCodec). + RegisterDefaultDecoder(reflect.Uint64, defaultUIntCodec). + RegisterDefaultDecoder(reflect.Float32, floatDecoder). + RegisterDefaultDecoder(reflect.Float64, floatDecoder). + RegisterDefaultDecoder(reflect.Array, ValueDecoderFunc(dvd.ArrayDecodeValue)). + RegisterDefaultDecoder(reflect.Map, defaultMapCodec). + RegisterDefaultDecoder(reflect.Slice, defaultSliceCodec). + RegisterDefaultDecoder(reflect.String, defaultStringCodec). + RegisterDefaultDecoder(reflect.Struct, newDefaultStructCodec()). + RegisterDefaultDecoder(reflect.Ptr, NewPointerCodec()). + RegisterTypeMapEntry(bsontype.Double, tFloat64). + RegisterTypeMapEntry(bsontype.String, tString). + RegisterTypeMapEntry(bsontype.Array, tA). + RegisterTypeMapEntry(bsontype.Binary, tBinary). + RegisterTypeMapEntry(bsontype.Undefined, tUndefined). + RegisterTypeMapEntry(bsontype.ObjectID, tOID). + RegisterTypeMapEntry(bsontype.Boolean, tBool). + RegisterTypeMapEntry(bsontype.DateTime, tDateTime). + RegisterTypeMapEntry(bsontype.Regex, tRegex). + RegisterTypeMapEntry(bsontype.DBPointer, tDBPointer). + RegisterTypeMapEntry(bsontype.JavaScript, tJavaScript). + RegisterTypeMapEntry(bsontype.Symbol, tSymbol). + RegisterTypeMapEntry(bsontype.CodeWithScope, tCodeWithScope). + RegisterTypeMapEntry(bsontype.Int32, tInt32). + RegisterTypeMapEntry(bsontype.Int64, tInt64). + RegisterTypeMapEntry(bsontype.Timestamp, tTimestamp). + RegisterTypeMapEntry(bsontype.Decimal128, tDecimal). + RegisterTypeMapEntry(bsontype.MinKey, tMinKey). + RegisterTypeMapEntry(bsontype.MaxKey, tMaxKey). + RegisterTypeMapEntry(bsontype.Type(0), tD). + RegisterTypeMapEntry(bsontype.EmbeddedDocument, tD). + RegisterHookDecoder(tValueUnmarshaler, ValueDecoderFunc(dvd.ValueUnmarshalerDecodeValue)). + RegisterHookDecoder(tUnmarshaler, ValueDecoderFunc(dvd.UnmarshalerDecodeValue)) +} + +// DDecodeValue is the ValueDecoderFunc for primitive.D instances. +func (dvd DefaultValueDecoders) DDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.IsValid() || !val.CanSet() || val.Type() != tD { + return ValueDecoderError{Name: "DDecodeValue", Kinds: []reflect.Kind{reflect.Slice}, Received: val} + } + + switch vrType := vr.Type(); vrType { + case bsontype.Type(0), bsontype.EmbeddedDocument: + dc.Ancestor = tD + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + default: + return fmt.Errorf("cannot decode %v into a primitive.D", vrType) + } + + dr, err := vr.ReadDocument() + if err != nil { + return err + } + + decoder, err := dc.LookupDecoder(tEmpty) + if err != nil { + return err + } + tEmptyTypeDecoder, _ := decoder.(typeDecoder) + + // Use the elements in the provided value if it's non nil. Otherwise, allocate a new D instance. + var elems primitive.D + if !val.IsNil() { + val.SetLen(0) + elems = val.Interface().(primitive.D) + } else { + elems = make(primitive.D, 0) + } + + for { + key, elemVr, err := dr.ReadElement() + if err == bsonrw.ErrEOD { + break + } else if err != nil { + return err + } + + // Pass false for convert because we don't need to call reflect.Value.Convert for tEmpty. + elem, err := decodeTypeOrValueWithInfo(decoder, tEmptyTypeDecoder, dc, elemVr, tEmpty, false) + if err != nil { + return err + } + + elems = append(elems, primitive.E{Key: key, Value: elem.Interface()}) + } + + val.Set(reflect.ValueOf(elems)) + return nil +} + +func (dvd DefaultValueDecoders) booleanDecodeType(dctx DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t.Kind() != reflect.Bool { + return emptyValue, ValueDecoderError{ + Name: "BooleanDecodeValue", + Kinds: []reflect.Kind{reflect.Bool}, + Received: reflect.Zero(t), + } + } + + var b bool + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Int32: + i32, err := vr.ReadInt32() + if err != nil { + return emptyValue, err + } + b = (i32 != 0) + case bsontype.Int64: + i64, err := vr.ReadInt64() + if err != nil { + return emptyValue, err + } + b = (i64 != 0) + case bsontype.Double: + f64, err := vr.ReadDouble() + if err != nil { + return emptyValue, err + } + b = (f64 != 0) + case bsontype.Boolean: + b, err = vr.ReadBoolean() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a boolean", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(b), nil +} + +// BooleanDecodeValue is the ValueDecoderFunc for bool types. +func (dvd DefaultValueDecoders) BooleanDecodeValue(dctx DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.IsValid() || !val.CanSet() || val.Kind() != reflect.Bool { + return ValueDecoderError{Name: "BooleanDecodeValue", Kinds: []reflect.Kind{reflect.Bool}, Received: val} + } + + elem, err := dvd.booleanDecodeType(dctx, vr, val.Type()) + if err != nil { + return err + } + + val.SetBool(elem.Bool()) + return nil +} + +func (DefaultValueDecoders) intDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + var i64 int64 + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Int32: + i32, err := vr.ReadInt32() + if err != nil { + return emptyValue, err + } + i64 = int64(i32) + case bsontype.Int64: + i64, err = vr.ReadInt64() + if err != nil { + return emptyValue, err + } + case bsontype.Double: + f64, err := vr.ReadDouble() + if err != nil { + return emptyValue, err + } + if !dc.Truncate && math.Floor(f64) != f64 { + return emptyValue, errCannotTruncate + } + if f64 > float64(math.MaxInt64) { + return emptyValue, fmt.Errorf("%g overflows int64", f64) + } + i64 = int64(f64) + case bsontype.Boolean: + b, err := vr.ReadBoolean() + if err != nil { + return emptyValue, err + } + if b { + i64 = 1 + } + case bsontype.Null: + if err = vr.ReadNull(); err != nil { + return emptyValue, err + } + case bsontype.Undefined: + if err = vr.ReadUndefined(); err != nil { + return emptyValue, err + } + default: + return emptyValue, fmt.Errorf("cannot decode %v into an integer type", vrType) + } + + switch t.Kind() { + case reflect.Int8: + if i64 < math.MinInt8 || i64 > math.MaxInt8 { + return emptyValue, fmt.Errorf("%d overflows int8", i64) + } + + return reflect.ValueOf(int8(i64)), nil + case reflect.Int16: + if i64 < math.MinInt16 || i64 > math.MaxInt16 { + return emptyValue, fmt.Errorf("%d overflows int16", i64) + } + + return reflect.ValueOf(int16(i64)), nil + case reflect.Int32: + if i64 < math.MinInt32 || i64 > math.MaxInt32 { + return emptyValue, fmt.Errorf("%d overflows int32", i64) + } + + return reflect.ValueOf(int32(i64)), nil + case reflect.Int64: + return reflect.ValueOf(i64), nil + case reflect.Int: + if int64(int(i64)) != i64 { // Can we fit this inside of an int + return emptyValue, fmt.Errorf("%d overflows int", i64) + } + + return reflect.ValueOf(int(i64)), nil + default: + return emptyValue, ValueDecoderError{ + Name: "IntDecodeValue", + Kinds: []reflect.Kind{reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64, reflect.Int}, + Received: reflect.Zero(t), + } + } +} + +// IntDecodeValue is the ValueDecoderFunc for int types. +func (dvd DefaultValueDecoders) IntDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() { + return ValueDecoderError{ + Name: "IntDecodeValue", + Kinds: []reflect.Kind{reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64, reflect.Int}, + Received: val, + } + } + + elem, err := dvd.intDecodeType(dc, vr, val.Type()) + if err != nil { + return err + } + + val.SetInt(elem.Int()) + return nil +} + +// UintDecodeValue is the ValueDecoderFunc for uint types. +// +// Deprecated: UintDecodeValue is not registered by default. Use UintCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) UintDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + var i64 int64 + var err error + switch vr.Type() { + case bsontype.Int32: + i32, err := vr.ReadInt32() + if err != nil { + return err + } + i64 = int64(i32) + case bsontype.Int64: + i64, err = vr.ReadInt64() + if err != nil { + return err + } + case bsontype.Double: + f64, err := vr.ReadDouble() + if err != nil { + return err + } + if !dc.Truncate && math.Floor(f64) != f64 { + return errors.New("UintDecodeValue can only truncate float64 to an integer type when truncation is enabled") + } + if f64 > float64(math.MaxInt64) { + return fmt.Errorf("%g overflows int64", f64) + } + i64 = int64(f64) + case bsontype.Boolean: + b, err := vr.ReadBoolean() + if err != nil { + return err + } + if b { + i64 = 1 + } + default: + return fmt.Errorf("cannot decode %v into an integer type", vr.Type()) + } + + if !val.CanSet() { + return ValueDecoderError{ + Name: "UintDecodeValue", + Kinds: []reflect.Kind{reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uint}, + Received: val, + } + } + + switch val.Kind() { + case reflect.Uint8: + if i64 < 0 || i64 > math.MaxUint8 { + return fmt.Errorf("%d overflows uint8", i64) + } + case reflect.Uint16: + if i64 < 0 || i64 > math.MaxUint16 { + return fmt.Errorf("%d overflows uint16", i64) + } + case reflect.Uint32: + if i64 < 0 || i64 > math.MaxUint32 { + return fmt.Errorf("%d overflows uint32", i64) + } + case reflect.Uint64: + if i64 < 0 { + return fmt.Errorf("%d overflows uint64", i64) + } + case reflect.Uint: + if i64 < 0 || int64(uint(i64)) != i64 { // Can we fit this inside of an uint + return fmt.Errorf("%d overflows uint", i64) + } + default: + return ValueDecoderError{ + Name: "UintDecodeValue", + Kinds: []reflect.Kind{reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uint}, + Received: val, + } + } + + val.SetUint(uint64(i64)) + return nil +} + +func (dvd DefaultValueDecoders) floatDecodeType(ec DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + var f float64 + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Int32: + i32, err := vr.ReadInt32() + if err != nil { + return emptyValue, err + } + f = float64(i32) + case bsontype.Int64: + i64, err := vr.ReadInt64() + if err != nil { + return emptyValue, err + } + f = float64(i64) + case bsontype.Double: + f, err = vr.ReadDouble() + if err != nil { + return emptyValue, err + } + case bsontype.Boolean: + b, err := vr.ReadBoolean() + if err != nil { + return emptyValue, err + } + if b { + f = 1 + } + case bsontype.Null: + if err = vr.ReadNull(); err != nil { + return emptyValue, err + } + case bsontype.Undefined: + if err = vr.ReadUndefined(); err != nil { + return emptyValue, err + } + default: + return emptyValue, fmt.Errorf("cannot decode %v into a float32 or float64 type", vrType) + } + + switch t.Kind() { + case reflect.Float32: + if !ec.Truncate && float64(float32(f)) != f { + return emptyValue, errCannotTruncate + } + + return reflect.ValueOf(float32(f)), nil + case reflect.Float64: + return reflect.ValueOf(f), nil + default: + return emptyValue, ValueDecoderError{ + Name: "FloatDecodeValue", + Kinds: []reflect.Kind{reflect.Float32, reflect.Float64}, + Received: reflect.Zero(t), + } + } +} + +// FloatDecodeValue is the ValueDecoderFunc for float types. +func (dvd DefaultValueDecoders) FloatDecodeValue(ec DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() { + return ValueDecoderError{ + Name: "FloatDecodeValue", + Kinds: []reflect.Kind{reflect.Float32, reflect.Float64}, + Received: val, + } + } + + elem, err := dvd.floatDecodeType(ec, vr, val.Type()) + if err != nil { + return err + } + + val.SetFloat(elem.Float()) + return nil +} + +// StringDecodeValue is the ValueDecoderFunc for string types. +// +// Deprecated: StringDecodeValue is not registered by default. Use StringCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) StringDecodeValue(dctx DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + var str string + var err error + switch vr.Type() { + // TODO(GODRIVER-577): Handle JavaScript and Symbol BSON types when allowed. + case bsontype.String: + str, err = vr.ReadString() + if err != nil { + return err + } + default: + return fmt.Errorf("cannot decode %v into a string type", vr.Type()) + } + if !val.CanSet() || val.Kind() != reflect.String { + return ValueDecoderError{Name: "StringDecodeValue", Kinds: []reflect.Kind{reflect.String}, Received: val} + } + + val.SetString(str) + return nil +} + +func (DefaultValueDecoders) javaScriptDecodeType(dctx DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tJavaScript { + return emptyValue, ValueDecoderError{ + Name: "JavaScriptDecodeValue", + Types: []reflect.Type{tJavaScript}, + Received: reflect.Zero(t), + } + } + + var js string + var err error + switch vrType := vr.Type(); vrType { + case bsontype.JavaScript: + js, err = vr.ReadJavascript() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a primitive.JavaScript", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.JavaScript(js)), nil +} + +// JavaScriptDecodeValue is the ValueDecoderFunc for the primitive.JavaScript type. +func (dvd DefaultValueDecoders) JavaScriptDecodeValue(dctx DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tJavaScript { + return ValueDecoderError{Name: "JavaScriptDecodeValue", Types: []reflect.Type{tJavaScript}, Received: val} + } + + elem, err := dvd.javaScriptDecodeType(dctx, vr, tJavaScript) + if err != nil { + return err + } + + val.SetString(elem.String()) + return nil +} + +func (DefaultValueDecoders) symbolDecodeType(dctx DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tSymbol { + return emptyValue, ValueDecoderError{ + Name: "SymbolDecodeValue", + Types: []reflect.Type{tSymbol}, + Received: reflect.Zero(t), + } + } + + var symbol string + var err error + switch vrType := vr.Type(); vrType { + case bsontype.String: + symbol, err = vr.ReadString() + case bsontype.Symbol: + symbol, err = vr.ReadSymbol() + case bsontype.Binary: + data, subtype, err := vr.ReadBinary() + if err != nil { + return emptyValue, err + } + + if subtype != bsontype.BinaryGeneric && subtype != bsontype.BinaryBinaryOld { + return emptyValue, decodeBinaryError{subtype: subtype, typeName: "primitive.Symbol"} + } + symbol = string(data) + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a primitive.Symbol", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.Symbol(symbol)), nil +} + +// SymbolDecodeValue is the ValueDecoderFunc for the primitive.Symbol type. +func (dvd DefaultValueDecoders) SymbolDecodeValue(dctx DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tSymbol { + return ValueDecoderError{Name: "SymbolDecodeValue", Types: []reflect.Type{tSymbol}, Received: val} + } + + elem, err := dvd.symbolDecodeType(dctx, vr, tSymbol) + if err != nil { + return err + } + + val.SetString(elem.String()) + return nil +} + +func (DefaultValueDecoders) binaryDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tBinary { + return emptyValue, ValueDecoderError{ + Name: "BinaryDecodeValue", + Types: []reflect.Type{tBinary}, + Received: reflect.Zero(t), + } + } + + var data []byte + var subtype byte + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Binary: + data, subtype, err = vr.ReadBinary() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a Binary", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.Binary{Subtype: subtype, Data: data}), nil +} + +// BinaryDecodeValue is the ValueDecoderFunc for Binary. +func (dvd DefaultValueDecoders) BinaryDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tBinary { + return ValueDecoderError{Name: "BinaryDecodeValue", Types: []reflect.Type{tBinary}, Received: val} + } + + elem, err := dvd.binaryDecodeType(dc, vr, tBinary) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) undefinedDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tUndefined { + return emptyValue, ValueDecoderError{ + Name: "UndefinedDecodeValue", + Types: []reflect.Type{tUndefined}, + Received: reflect.Zero(t), + } + } + + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Undefined: + err = vr.ReadUndefined() + case bsontype.Null: + err = vr.ReadNull() + default: + return emptyValue, fmt.Errorf("cannot decode %v into an Undefined", vr.Type()) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.Undefined{}), nil +} + +// UndefinedDecodeValue is the ValueDecoderFunc for Undefined. +func (dvd DefaultValueDecoders) UndefinedDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tUndefined { + return ValueDecoderError{Name: "UndefinedDecodeValue", Types: []reflect.Type{tUndefined}, Received: val} + } + + elem, err := dvd.undefinedDecodeType(dc, vr, tUndefined) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +// Accept both 12-byte string and pretty-printed 24-byte hex string formats. +func (dvd DefaultValueDecoders) objectIDDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tOID { + return emptyValue, ValueDecoderError{ + Name: "ObjectIDDecodeValue", + Types: []reflect.Type{tOID}, + Received: reflect.Zero(t), + } + } + + var oid primitive.ObjectID + var err error + switch vrType := vr.Type(); vrType { + case bsontype.ObjectID: + oid, err = vr.ReadObjectID() + if err != nil { + return emptyValue, err + } + case bsontype.String: + str, err := vr.ReadString() + if err != nil { + return emptyValue, err + } + if oid, err = primitive.ObjectIDFromHex(str); err == nil { + break + } + if len(str) != 12 { + return emptyValue, fmt.Errorf("an ObjectID string must be exactly 12 bytes long (got %v)", len(str)) + } + byteArr := []byte(str) + copy(oid[:], byteArr) + case bsontype.Null: + if err = vr.ReadNull(); err != nil { + return emptyValue, err + } + case bsontype.Undefined: + if err = vr.ReadUndefined(); err != nil { + return emptyValue, err + } + default: + return emptyValue, fmt.Errorf("cannot decode %v into an ObjectID", vrType) + } + + return reflect.ValueOf(oid), nil +} + +// ObjectIDDecodeValue is the ValueDecoderFunc for primitive.ObjectID. +func (dvd DefaultValueDecoders) ObjectIDDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tOID { + return ValueDecoderError{Name: "ObjectIDDecodeValue", Types: []reflect.Type{tOID}, Received: val} + } + + elem, err := dvd.objectIDDecodeType(dc, vr, tOID) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) dateTimeDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tDateTime { + return emptyValue, ValueDecoderError{ + Name: "DateTimeDecodeValue", + Types: []reflect.Type{tDateTime}, + Received: reflect.Zero(t), + } + } + + var dt int64 + var err error + switch vrType := vr.Type(); vrType { + case bsontype.DateTime: + dt, err = vr.ReadDateTime() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a DateTime", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.DateTime(dt)), nil +} + +// DateTimeDecodeValue is the ValueDecoderFunc for DateTime. +func (dvd DefaultValueDecoders) DateTimeDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tDateTime { + return ValueDecoderError{Name: "DateTimeDecodeValue", Types: []reflect.Type{tDateTime}, Received: val} + } + + elem, err := dvd.dateTimeDecodeType(dc, vr, tDateTime) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) nullDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tNull { + return emptyValue, ValueDecoderError{ + Name: "NullDecodeValue", + Types: []reflect.Type{tNull}, + Received: reflect.Zero(t), + } + } + + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Undefined: + err = vr.ReadUndefined() + case bsontype.Null: + err = vr.ReadNull() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a Null", vr.Type()) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.Null{}), nil +} + +// NullDecodeValue is the ValueDecoderFunc for Null. +func (dvd DefaultValueDecoders) NullDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tNull { + return ValueDecoderError{Name: "NullDecodeValue", Types: []reflect.Type{tNull}, Received: val} + } + + elem, err := dvd.nullDecodeType(dc, vr, tNull) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) regexDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tRegex { + return emptyValue, ValueDecoderError{ + Name: "RegexDecodeValue", + Types: []reflect.Type{tRegex}, + Received: reflect.Zero(t), + } + } + + var pattern, options string + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Regex: + pattern, options, err = vr.ReadRegex() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a Regex", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.Regex{Pattern: pattern, Options: options}), nil +} + +// RegexDecodeValue is the ValueDecoderFunc for Regex. +func (dvd DefaultValueDecoders) RegexDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tRegex { + return ValueDecoderError{Name: "RegexDecodeValue", Types: []reflect.Type{tRegex}, Received: val} + } + + elem, err := dvd.regexDecodeType(dc, vr, tRegex) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) dBPointerDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tDBPointer { + return emptyValue, ValueDecoderError{ + Name: "DBPointerDecodeValue", + Types: []reflect.Type{tDBPointer}, + Received: reflect.Zero(t), + } + } + + var ns string + var pointer primitive.ObjectID + var err error + switch vrType := vr.Type(); vrType { + case bsontype.DBPointer: + ns, pointer, err = vr.ReadDBPointer() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a DBPointer", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.DBPointer{DB: ns, Pointer: pointer}), nil +} + +// DBPointerDecodeValue is the ValueDecoderFunc for DBPointer. +func (dvd DefaultValueDecoders) DBPointerDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tDBPointer { + return ValueDecoderError{Name: "DBPointerDecodeValue", Types: []reflect.Type{tDBPointer}, Received: val} + } + + elem, err := dvd.dBPointerDecodeType(dc, vr, tDBPointer) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) timestampDecodeType(dc DecodeContext, vr bsonrw.ValueReader, reflectType reflect.Type) (reflect.Value, error) { + if reflectType != tTimestamp { + return emptyValue, ValueDecoderError{ + Name: "TimestampDecodeValue", + Types: []reflect.Type{tTimestamp}, + Received: reflect.Zero(reflectType), + } + } + + var t, incr uint32 + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Timestamp: + t, incr, err = vr.ReadTimestamp() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a Timestamp", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.Timestamp{T: t, I: incr}), nil +} + +// TimestampDecodeValue is the ValueDecoderFunc for Timestamp. +func (dvd DefaultValueDecoders) TimestampDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tTimestamp { + return ValueDecoderError{Name: "TimestampDecodeValue", Types: []reflect.Type{tTimestamp}, Received: val} + } + + elem, err := dvd.timestampDecodeType(dc, vr, tTimestamp) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) minKeyDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tMinKey { + return emptyValue, ValueDecoderError{ + Name: "MinKeyDecodeValue", + Types: []reflect.Type{tMinKey}, + Received: reflect.Zero(t), + } + } + + var err error + switch vrType := vr.Type(); vrType { + case bsontype.MinKey: + err = vr.ReadMinKey() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a MinKey", vr.Type()) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.MinKey{}), nil +} + +// MinKeyDecodeValue is the ValueDecoderFunc for MinKey. +func (dvd DefaultValueDecoders) MinKeyDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tMinKey { + return ValueDecoderError{Name: "MinKeyDecodeValue", Types: []reflect.Type{tMinKey}, Received: val} + } + + elem, err := dvd.minKeyDecodeType(dc, vr, tMinKey) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (DefaultValueDecoders) maxKeyDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tMaxKey { + return emptyValue, ValueDecoderError{ + Name: "MaxKeyDecodeValue", + Types: []reflect.Type{tMaxKey}, + Received: reflect.Zero(t), + } + } + + var err error + switch vrType := vr.Type(); vrType { + case bsontype.MaxKey: + err = vr.ReadMaxKey() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a MaxKey", vr.Type()) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(primitive.MaxKey{}), nil +} + +// MaxKeyDecodeValue is the ValueDecoderFunc for MaxKey. +func (dvd DefaultValueDecoders) MaxKeyDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tMaxKey { + return ValueDecoderError{Name: "MaxKeyDecodeValue", Types: []reflect.Type{tMaxKey}, Received: val} + } + + elem, err := dvd.maxKeyDecodeType(dc, vr, tMaxKey) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (dvd DefaultValueDecoders) decimal128DecodeType(dctx DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tDecimal { + return emptyValue, ValueDecoderError{ + Name: "Decimal128DecodeValue", + Types: []reflect.Type{tDecimal}, + Received: reflect.Zero(t), + } + } + + var d128 primitive.Decimal128 + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Decimal128: + d128, err = vr.ReadDecimal128() + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a primitive.Decimal128", vr.Type()) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(d128), nil +} + +// Decimal128DecodeValue is the ValueDecoderFunc for primitive.Decimal128. +func (dvd DefaultValueDecoders) Decimal128DecodeValue(dctx DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tDecimal { + return ValueDecoderError{Name: "Decimal128DecodeValue", Types: []reflect.Type{tDecimal}, Received: val} + } + + elem, err := dvd.decimal128DecodeType(dctx, vr, tDecimal) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (dvd DefaultValueDecoders) jsonNumberDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tJSONNumber { + return emptyValue, ValueDecoderError{ + Name: "JSONNumberDecodeValue", + Types: []reflect.Type{tJSONNumber}, + Received: reflect.Zero(t), + } + } + + var jsonNum json.Number + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Double: + f64, err := vr.ReadDouble() + if err != nil { + return emptyValue, err + } + jsonNum = json.Number(strconv.FormatFloat(f64, 'f', -1, 64)) + case bsontype.Int32: + i32, err := vr.ReadInt32() + if err != nil { + return emptyValue, err + } + jsonNum = json.Number(strconv.FormatInt(int64(i32), 10)) + case bsontype.Int64: + i64, err := vr.ReadInt64() + if err != nil { + return emptyValue, err + } + jsonNum = json.Number(strconv.FormatInt(i64, 10)) + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a json.Number", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(jsonNum), nil +} + +// JSONNumberDecodeValue is the ValueDecoderFunc for json.Number. +func (dvd DefaultValueDecoders) JSONNumberDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tJSONNumber { + return ValueDecoderError{Name: "JSONNumberDecodeValue", Types: []reflect.Type{tJSONNumber}, Received: val} + } + + elem, err := dvd.jsonNumberDecodeType(dc, vr, tJSONNumber) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (dvd DefaultValueDecoders) urlDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tURL { + return emptyValue, ValueDecoderError{ + Name: "URLDecodeValue", + Types: []reflect.Type{tURL}, + Received: reflect.Zero(t), + } + } + + urlPtr := &url.URL{} + var err error + switch vrType := vr.Type(); vrType { + case bsontype.String: + var str string // Declare str here to avoid shadowing err during the ReadString call. + str, err = vr.ReadString() + if err != nil { + return emptyValue, err + } + + urlPtr, err = url.Parse(str) + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a *url.URL", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(urlPtr).Elem(), nil +} + +// URLDecodeValue is the ValueDecoderFunc for url.URL. +func (dvd DefaultValueDecoders) URLDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tURL { + return ValueDecoderError{Name: "URLDecodeValue", Types: []reflect.Type{tURL}, Received: val} + } + + elem, err := dvd.urlDecodeType(dc, vr, tURL) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +// TimeDecodeValue is the ValueDecoderFunc for time.Time. +// +// Deprecated: TimeDecodeValue is not registered by default. Use TimeCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) TimeDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if vr.Type() != bsontype.DateTime { + return fmt.Errorf("cannot decode %v into a time.Time", vr.Type()) + } + + dt, err := vr.ReadDateTime() + if err != nil { + return err + } + + if !val.CanSet() || val.Type() != tTime { + return ValueDecoderError{Name: "TimeDecodeValue", Types: []reflect.Type{tTime}, Received: val} + } + + val.Set(reflect.ValueOf(time.Unix(dt/1000, dt%1000*1000000).UTC())) + return nil +} + +// ByteSliceDecodeValue is the ValueDecoderFunc for []byte. +// +// Deprecated: ByteSliceDecodeValue is not registered by default. Use ByteSliceCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) ByteSliceDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if vr.Type() != bsontype.Binary && vr.Type() != bsontype.Null { + return fmt.Errorf("cannot decode %v into a []byte", vr.Type()) + } + + if !val.CanSet() || val.Type() != tByteSlice { + return ValueDecoderError{Name: "ByteSliceDecodeValue", Types: []reflect.Type{tByteSlice}, Received: val} + } + + if vr.Type() == bsontype.Null { + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + } + + data, subtype, err := vr.ReadBinary() + if err != nil { + return err + } + if subtype != 0x00 { + return fmt.Errorf("ByteSliceDecodeValue can only be used to decode subtype 0x00 for %s, got %v", bsontype.Binary, subtype) + } + + val.Set(reflect.ValueOf(data)) + return nil +} + +// MapDecodeValue is the ValueDecoderFunc for map[string]* types. +// +// Deprecated: MapDecodeValue is not registered by default. Use MapCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) MapDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Kind() != reflect.Map || val.Type().Key().Kind() != reflect.String { + return ValueDecoderError{Name: "MapDecodeValue", Kinds: []reflect.Kind{reflect.Map}, Received: val} + } + + switch vr.Type() { + case bsontype.Type(0), bsontype.EmbeddedDocument: + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + default: + return fmt.Errorf("cannot decode %v into a %s", vr.Type(), val.Type()) + } + + dr, err := vr.ReadDocument() + if err != nil { + return err + } + + if val.IsNil() { + val.Set(reflect.MakeMap(val.Type())) + } + + eType := val.Type().Elem() + decoder, err := dc.LookupDecoder(eType) + if err != nil { + return err + } + + if eType == tEmpty { + dc.Ancestor = val.Type() + } + + keyType := val.Type().Key() + for { + key, vr, err := dr.ReadElement() + if err == bsonrw.ErrEOD { + break + } + if err != nil { + return err + } + + elem := reflect.New(eType).Elem() + + err = decoder.DecodeValue(dc, vr, elem) + if err != nil { + return err + } + + val.SetMapIndex(reflect.ValueOf(key).Convert(keyType), elem) + } + return nil +} + +// ArrayDecodeValue is the ValueDecoderFunc for array types. +func (dvd DefaultValueDecoders) ArrayDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Array { + return ValueDecoderError{Name: "ArrayDecodeValue", Kinds: []reflect.Kind{reflect.Array}, Received: val} + } + + switch vrType := vr.Type(); vrType { + case bsontype.Array: + case bsontype.Type(0), bsontype.EmbeddedDocument: + if val.Type().Elem() != tE { + return fmt.Errorf("cannot decode document into %s", val.Type()) + } + case bsontype.Binary: + if val.Type().Elem() != tByte { + return fmt.Errorf("ArrayDecodeValue can only be used to decode binary into a byte array, got %v", vrType) + } + data, subtype, err := vr.ReadBinary() + if err != nil { + return err + } + if subtype != bsontype.BinaryGeneric && subtype != bsontype.BinaryBinaryOld { + return fmt.Errorf("ArrayDecodeValue can only be used to decode subtype 0x00 or 0x02 for %s, got %v", bsontype.Binary, subtype) + } + + if len(data) > val.Len() { + return fmt.Errorf("more elements returned in array than can fit inside %s", val.Type()) + } + + for idx, elem := range data { + val.Index(idx).Set(reflect.ValueOf(elem)) + } + return nil + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + case bsontype.Undefined: + val.Set(reflect.Zero(val.Type())) + return vr.ReadUndefined() + default: + return fmt.Errorf("cannot decode %v into an array", vrType) + } + + var elemsFunc func(DecodeContext, bsonrw.ValueReader, reflect.Value) ([]reflect.Value, error) + switch val.Type().Elem() { + case tE: + elemsFunc = dvd.decodeD + default: + elemsFunc = dvd.decodeDefault + } + + elems, err := elemsFunc(dc, vr, val) + if err != nil { + return err + } + + if len(elems) > val.Len() { + return fmt.Errorf("more elements returned in array than can fit inside %s, got %v elements", val.Type(), len(elems)) + } + + for idx, elem := range elems { + val.Index(idx).Set(elem) + } + + return nil +} + +// SliceDecodeValue is the ValueDecoderFunc for slice types. +// +// Deprecated: SliceDecodeValue is not registered by default. Use SliceCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) SliceDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Kind() != reflect.Slice { + return ValueDecoderError{Name: "SliceDecodeValue", Kinds: []reflect.Kind{reflect.Slice}, Received: val} + } + + switch vr.Type() { + case bsontype.Array: + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + case bsontype.Type(0), bsontype.EmbeddedDocument: + if val.Type().Elem() != tE { + return fmt.Errorf("cannot decode document into %s", val.Type()) + } + default: + return fmt.Errorf("cannot decode %v into a slice", vr.Type()) + } + + var elemsFunc func(DecodeContext, bsonrw.ValueReader, reflect.Value) ([]reflect.Value, error) + switch val.Type().Elem() { + case tE: + dc.Ancestor = val.Type() + elemsFunc = dvd.decodeD + default: + elemsFunc = dvd.decodeDefault + } + + elems, err := elemsFunc(dc, vr, val) + if err != nil { + return err + } + + if val.IsNil() { + val.Set(reflect.MakeSlice(val.Type(), 0, len(elems))) + } + + val.SetLen(0) + val.Set(reflect.Append(val, elems...)) + + return nil +} + +// ValueUnmarshalerDecodeValue is the ValueDecoderFunc for ValueUnmarshaler implementations. +func (dvd DefaultValueDecoders) ValueUnmarshalerDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.IsValid() || (!val.Type().Implements(tValueUnmarshaler) && !reflect.PtrTo(val.Type()).Implements(tValueUnmarshaler)) { + return ValueDecoderError{Name: "ValueUnmarshalerDecodeValue", Types: []reflect.Type{tValueUnmarshaler}, Received: val} + } + + if val.Kind() == reflect.Ptr && val.IsNil() { + if !val.CanSet() { + return ValueDecoderError{Name: "ValueUnmarshalerDecodeValue", Types: []reflect.Type{tValueUnmarshaler}, Received: val} + } + val.Set(reflect.New(val.Type().Elem())) + } + + if !val.Type().Implements(tValueUnmarshaler) { + if !val.CanAddr() { + return ValueDecoderError{Name: "ValueUnmarshalerDecodeValue", Types: []reflect.Type{tValueUnmarshaler}, Received: val} + } + val = val.Addr() // If they type doesn't implement the interface, a pointer to it must. + } + + t, src, err := bsonrw.Copier{}.CopyValueToBytes(vr) + if err != nil { + return err + } + + fn := val.Convert(tValueUnmarshaler).MethodByName("UnmarshalBSONValue") + errVal := fn.Call([]reflect.Value{reflect.ValueOf(t), reflect.ValueOf(src)})[0] + if !errVal.IsNil() { + return errVal.Interface().(error) + } + return nil +} + +// UnmarshalerDecodeValue is the ValueDecoderFunc for Unmarshaler implementations. +func (dvd DefaultValueDecoders) UnmarshalerDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.IsValid() || (!val.Type().Implements(tUnmarshaler) && !reflect.PtrTo(val.Type()).Implements(tUnmarshaler)) { + return ValueDecoderError{Name: "UnmarshalerDecodeValue", Types: []reflect.Type{tUnmarshaler}, Received: val} + } + + if val.Kind() == reflect.Ptr && val.IsNil() { + if !val.CanSet() { + return ValueDecoderError{Name: "UnmarshalerDecodeValue", Types: []reflect.Type{tUnmarshaler}, Received: val} + } + val.Set(reflect.New(val.Type().Elem())) + } + + if !val.Type().Implements(tUnmarshaler) { + if !val.CanAddr() { + return ValueDecoderError{Name: "UnmarshalerDecodeValue", Types: []reflect.Type{tUnmarshaler}, Received: val} + } + val = val.Addr() // If they type doesn't implement the interface, a pointer to it must. + } + + _, src, err := bsonrw.Copier{}.CopyValueToBytes(vr) + if err != nil { + return err + } + + // If the target Go value is a pointer and the BSON field value is empty, set the value to the + // zero value of the pointer (nil) and don't call UnmarshalBSON. UnmarshalBSON has no way to + // change the pointer value from within the function (only the value at the pointer address), + // so it can't set the pointer to "nil" itself. Since the most common Go value for an empty BSON + // field value is "nil", we set "nil" here and don't call UnmarshalBSON. This behavior matches + // the behavior of the Go "encoding/json" unmarshaler when the target Go value is a pointer and + // the JSON field value is "null". + if val.Kind() == reflect.Ptr && len(src) == 0 { + val.Set(reflect.Zero(val.Type())) + return nil + } + + fn := val.Convert(tUnmarshaler).MethodByName("UnmarshalBSON") + errVal := fn.Call([]reflect.Value{reflect.ValueOf(src)})[0] + if !errVal.IsNil() { + return errVal.Interface().(error) + } + return nil +} + +// EmptyInterfaceDecodeValue is the ValueDecoderFunc for interface{}. +// +// Deprecated: EmptyInterfaceDecodeValue is not registered by default. Use EmptyInterfaceCodec.DecodeValue instead. +func (dvd DefaultValueDecoders) EmptyInterfaceDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tEmpty { + return ValueDecoderError{Name: "EmptyInterfaceDecodeValue", Types: []reflect.Type{tEmpty}, Received: val} + } + + rtype, err := dc.LookupTypeMapEntry(vr.Type()) + if err != nil { + switch vr.Type() { + case bsontype.EmbeddedDocument: + if dc.Ancestor != nil { + rtype = dc.Ancestor + break + } + rtype = tD + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + default: + return err + } + } + + decoder, err := dc.LookupDecoder(rtype) + if err != nil { + return err + } + + elem := reflect.New(rtype).Elem() + err = decoder.DecodeValue(dc, vr, elem) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +// CoreDocumentDecodeValue is the ValueDecoderFunc for bsoncore.Document. +func (DefaultValueDecoders) CoreDocumentDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tCoreDocument { + return ValueDecoderError{Name: "CoreDocumentDecodeValue", Types: []reflect.Type{tCoreDocument}, Received: val} + } + + if val.IsNil() { + val.Set(reflect.MakeSlice(val.Type(), 0, 0)) + } + + val.SetLen(0) + + cdoc, err := bsonrw.Copier{}.AppendDocumentBytes(val.Interface().(bsoncore.Document), vr) + val.Set(reflect.ValueOf(cdoc)) + return err +} + +func (dvd DefaultValueDecoders) decodeDefault(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) ([]reflect.Value, error) { + elems := make([]reflect.Value, 0) + + ar, err := vr.ReadArray() + if err != nil { + return nil, err + } + + eType := val.Type().Elem() + + decoder, err := dc.LookupDecoder(eType) + if err != nil { + return nil, err + } + eTypeDecoder, _ := decoder.(typeDecoder) + + idx := 0 + for { + vr, err := ar.ReadValue() + if err == bsonrw.ErrEOA { + break + } + if err != nil { + return nil, err + } + + elem, err := decodeTypeOrValueWithInfo(decoder, eTypeDecoder, dc, vr, eType, true) + if err != nil { + return nil, newDecodeError(strconv.Itoa(idx), err) + } + elems = append(elems, elem) + idx++ + } + + return elems, nil +} + +func (dvd DefaultValueDecoders) readCodeWithScope(dc DecodeContext, vr bsonrw.ValueReader) (primitive.CodeWithScope, error) { + var cws primitive.CodeWithScope + + code, dr, err := vr.ReadCodeWithScope() + if err != nil { + return cws, err + } + + scope := reflect.New(tD).Elem() + elems, err := dvd.decodeElemsFromDocumentReader(dc, dr) + if err != nil { + return cws, err + } + + scope.Set(reflect.MakeSlice(tD, 0, len(elems))) + scope.Set(reflect.Append(scope, elems...)) + + cws = primitive.CodeWithScope{ + Code: primitive.JavaScript(code), + Scope: scope.Interface().(primitive.D), + } + return cws, nil +} + +func (dvd DefaultValueDecoders) codeWithScopeDecodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tCodeWithScope { + return emptyValue, ValueDecoderError{ + Name: "CodeWithScopeDecodeValue", + Types: []reflect.Type{tCodeWithScope}, + Received: reflect.Zero(t), + } + } + + var cws primitive.CodeWithScope + var err error + switch vrType := vr.Type(); vrType { + case bsontype.CodeWithScope: + cws, err = dvd.readCodeWithScope(dc, vr) + case bsontype.Null: + err = vr.ReadNull() + case bsontype.Undefined: + err = vr.ReadUndefined() + default: + return emptyValue, fmt.Errorf("cannot decode %v into a primitive.CodeWithScope", vrType) + } + if err != nil { + return emptyValue, err + } + + return reflect.ValueOf(cws), nil +} + +// CodeWithScopeDecodeValue is the ValueDecoderFunc for CodeWithScope. +func (dvd DefaultValueDecoders) CodeWithScopeDecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tCodeWithScope { + return ValueDecoderError{Name: "CodeWithScopeDecodeValue", Types: []reflect.Type{tCodeWithScope}, Received: val} + } + + elem, err := dvd.codeWithScopeDecodeType(dc, vr, tCodeWithScope) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +func (dvd DefaultValueDecoders) decodeD(dc DecodeContext, vr bsonrw.ValueReader, _ reflect.Value) ([]reflect.Value, error) { + switch vr.Type() { + case bsontype.Type(0), bsontype.EmbeddedDocument: + default: + return nil, fmt.Errorf("cannot decode %v into a D", vr.Type()) + } + + dr, err := vr.ReadDocument() + if err != nil { + return nil, err + } + + return dvd.decodeElemsFromDocumentReader(dc, dr) +} + +func (DefaultValueDecoders) decodeElemsFromDocumentReader(dc DecodeContext, dr bsonrw.DocumentReader) ([]reflect.Value, error) { + decoder, err := dc.LookupDecoder(tEmpty) + if err != nil { + return nil, err + } + + elems := make([]reflect.Value, 0) + for { + key, vr, err := dr.ReadElement() + if err == bsonrw.ErrEOD { + break + } + if err != nil { + return nil, err + } + + val := reflect.New(tEmpty).Elem() + err = decoder.DecodeValue(dc, vr, val) + if err != nil { + return nil, newDecodeError(key, err) + } + + elems = append(elems, reflect.ValueOf(primitive.E{Key: key, Value: val.Interface()})) + } + + return elems, nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/default_value_encoders.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/default_value_encoders.go new file mode 100644 index 0000000..6bdb43c --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/default_value_encoders.go @@ -0,0 +1,766 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "encoding/json" + "errors" + "fmt" + "math" + "net/url" + "reflect" + "sync" + "time" + + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" + "go.mongodb.org/mongo-driver/x/bsonx/bsoncore" +) + +var defaultValueEncoders DefaultValueEncoders + +var bvwPool = bsonrw.NewBSONValueWriterPool() + +var errInvalidValue = errors.New("cannot encode invalid element") + +var sliceWriterPool = sync.Pool{ + New: func() interface{} { + sw := make(bsonrw.SliceWriter, 0) + return &sw + }, +} + +func encodeElement(ec EncodeContext, dw bsonrw.DocumentWriter, e primitive.E) error { + vw, err := dw.WriteDocumentElement(e.Key) + if err != nil { + return err + } + + if e.Value == nil { + return vw.WriteNull() + } + encoder, err := ec.LookupEncoder(reflect.TypeOf(e.Value)) + if err != nil { + return err + } + + err = encoder.EncodeValue(ec, vw, reflect.ValueOf(e.Value)) + if err != nil { + return err + } + return nil +} + +// DefaultValueEncoders is a namespace type for the default ValueEncoders used +// when creating a registry. +type DefaultValueEncoders struct{} + +// RegisterDefaultEncoders will register the encoder methods attached to DefaultValueEncoders with +// the provided RegistryBuilder. +func (dve DefaultValueEncoders) RegisterDefaultEncoders(rb *RegistryBuilder) { + if rb == nil { + panic(errors.New("argument to RegisterDefaultEncoders must not be nil")) + } + rb. + RegisterTypeEncoder(tByteSlice, defaultByteSliceCodec). + RegisterTypeEncoder(tTime, defaultTimeCodec). + RegisterTypeEncoder(tEmpty, defaultEmptyInterfaceCodec). + RegisterTypeEncoder(tCoreArray, defaultArrayCodec). + RegisterTypeEncoder(tOID, ValueEncoderFunc(dve.ObjectIDEncodeValue)). + RegisterTypeEncoder(tDecimal, ValueEncoderFunc(dve.Decimal128EncodeValue)). + RegisterTypeEncoder(tJSONNumber, ValueEncoderFunc(dve.JSONNumberEncodeValue)). + RegisterTypeEncoder(tURL, ValueEncoderFunc(dve.URLEncodeValue)). + RegisterTypeEncoder(tJavaScript, ValueEncoderFunc(dve.JavaScriptEncodeValue)). + RegisterTypeEncoder(tSymbol, ValueEncoderFunc(dve.SymbolEncodeValue)). + RegisterTypeEncoder(tBinary, ValueEncoderFunc(dve.BinaryEncodeValue)). + RegisterTypeEncoder(tUndefined, ValueEncoderFunc(dve.UndefinedEncodeValue)). + RegisterTypeEncoder(tDateTime, ValueEncoderFunc(dve.DateTimeEncodeValue)). + RegisterTypeEncoder(tNull, ValueEncoderFunc(dve.NullEncodeValue)). + RegisterTypeEncoder(tRegex, ValueEncoderFunc(dve.RegexEncodeValue)). + RegisterTypeEncoder(tDBPointer, ValueEncoderFunc(dve.DBPointerEncodeValue)). + RegisterTypeEncoder(tTimestamp, ValueEncoderFunc(dve.TimestampEncodeValue)). + RegisterTypeEncoder(tMinKey, ValueEncoderFunc(dve.MinKeyEncodeValue)). + RegisterTypeEncoder(tMaxKey, ValueEncoderFunc(dve.MaxKeyEncodeValue)). + RegisterTypeEncoder(tCoreDocument, ValueEncoderFunc(dve.CoreDocumentEncodeValue)). + RegisterTypeEncoder(tCodeWithScope, ValueEncoderFunc(dve.CodeWithScopeEncodeValue)). + RegisterDefaultEncoder(reflect.Bool, ValueEncoderFunc(dve.BooleanEncodeValue)). + RegisterDefaultEncoder(reflect.Int, ValueEncoderFunc(dve.IntEncodeValue)). + RegisterDefaultEncoder(reflect.Int8, ValueEncoderFunc(dve.IntEncodeValue)). + RegisterDefaultEncoder(reflect.Int16, ValueEncoderFunc(dve.IntEncodeValue)). + RegisterDefaultEncoder(reflect.Int32, ValueEncoderFunc(dve.IntEncodeValue)). + RegisterDefaultEncoder(reflect.Int64, ValueEncoderFunc(dve.IntEncodeValue)). + RegisterDefaultEncoder(reflect.Uint, defaultUIntCodec). + RegisterDefaultEncoder(reflect.Uint8, defaultUIntCodec). + RegisterDefaultEncoder(reflect.Uint16, defaultUIntCodec). + RegisterDefaultEncoder(reflect.Uint32, defaultUIntCodec). + RegisterDefaultEncoder(reflect.Uint64, defaultUIntCodec). + RegisterDefaultEncoder(reflect.Float32, ValueEncoderFunc(dve.FloatEncodeValue)). + RegisterDefaultEncoder(reflect.Float64, ValueEncoderFunc(dve.FloatEncodeValue)). + RegisterDefaultEncoder(reflect.Array, ValueEncoderFunc(dve.ArrayEncodeValue)). + RegisterDefaultEncoder(reflect.Map, defaultMapCodec). + RegisterDefaultEncoder(reflect.Slice, defaultSliceCodec). + RegisterDefaultEncoder(reflect.String, defaultStringCodec). + RegisterDefaultEncoder(reflect.Struct, newDefaultStructCodec()). + RegisterDefaultEncoder(reflect.Ptr, NewPointerCodec()). + RegisterHookEncoder(tValueMarshaler, ValueEncoderFunc(dve.ValueMarshalerEncodeValue)). + RegisterHookEncoder(tMarshaler, ValueEncoderFunc(dve.MarshalerEncodeValue)). + RegisterHookEncoder(tProxy, ValueEncoderFunc(dve.ProxyEncodeValue)) +} + +// BooleanEncodeValue is the ValueEncoderFunc for bool types. +func (dve DefaultValueEncoders) BooleanEncodeValue(ectx EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Bool { + return ValueEncoderError{Name: "BooleanEncodeValue", Kinds: []reflect.Kind{reflect.Bool}, Received: val} + } + return vw.WriteBoolean(val.Bool()) +} + +func fitsIn32Bits(i int64) bool { + return math.MinInt32 <= i && i <= math.MaxInt32 +} + +// IntEncodeValue is the ValueEncoderFunc for int types. +func (dve DefaultValueEncoders) IntEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + switch val.Kind() { + case reflect.Int8, reflect.Int16, reflect.Int32: + return vw.WriteInt32(int32(val.Int())) + case reflect.Int: + i64 := val.Int() + if fitsIn32Bits(i64) { + return vw.WriteInt32(int32(i64)) + } + return vw.WriteInt64(i64) + case reflect.Int64: + i64 := val.Int() + if ec.MinSize && fitsIn32Bits(i64) { + return vw.WriteInt32(int32(i64)) + } + return vw.WriteInt64(i64) + } + + return ValueEncoderError{ + Name: "IntEncodeValue", + Kinds: []reflect.Kind{reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64, reflect.Int}, + Received: val, + } +} + +// UintEncodeValue is the ValueEncoderFunc for uint types. +// +// Deprecated: UintEncodeValue is not registered by default. Use UintCodec.EncodeValue instead. +func (dve DefaultValueEncoders) UintEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + switch val.Kind() { + case reflect.Uint8, reflect.Uint16: + return vw.WriteInt32(int32(val.Uint())) + case reflect.Uint, reflect.Uint32, reflect.Uint64: + u64 := val.Uint() + if ec.MinSize && u64 <= math.MaxInt32 { + return vw.WriteInt32(int32(u64)) + } + if u64 > math.MaxInt64 { + return fmt.Errorf("%d overflows int64", u64) + } + return vw.WriteInt64(int64(u64)) + } + + return ValueEncoderError{ + Name: "UintEncodeValue", + Kinds: []reflect.Kind{reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uint}, + Received: val, + } +} + +// FloatEncodeValue is the ValueEncoderFunc for float types. +func (dve DefaultValueEncoders) FloatEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + switch val.Kind() { + case reflect.Float32, reflect.Float64: + return vw.WriteDouble(val.Float()) + } + + return ValueEncoderError{Name: "FloatEncodeValue", Kinds: []reflect.Kind{reflect.Float32, reflect.Float64}, Received: val} +} + +// StringEncodeValue is the ValueEncoderFunc for string types. +// +// Deprecated: StringEncodeValue is not registered by default. Use StringCodec.EncodeValue instead. +func (dve DefaultValueEncoders) StringEncodeValue(ectx EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if val.Kind() != reflect.String { + return ValueEncoderError{ + Name: "StringEncodeValue", + Kinds: []reflect.Kind{reflect.String}, + Received: val, + } + } + + return vw.WriteString(val.String()) +} + +// ObjectIDEncodeValue is the ValueEncoderFunc for primitive.ObjectID. +func (dve DefaultValueEncoders) ObjectIDEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tOID { + return ValueEncoderError{Name: "ObjectIDEncodeValue", Types: []reflect.Type{tOID}, Received: val} + } + return vw.WriteObjectID(val.Interface().(primitive.ObjectID)) +} + +// Decimal128EncodeValue is the ValueEncoderFunc for primitive.Decimal128. +func (dve DefaultValueEncoders) Decimal128EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tDecimal { + return ValueEncoderError{Name: "Decimal128EncodeValue", Types: []reflect.Type{tDecimal}, Received: val} + } + return vw.WriteDecimal128(val.Interface().(primitive.Decimal128)) +} + +// JSONNumberEncodeValue is the ValueEncoderFunc for json.Number. +func (dve DefaultValueEncoders) JSONNumberEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tJSONNumber { + return ValueEncoderError{Name: "JSONNumberEncodeValue", Types: []reflect.Type{tJSONNumber}, Received: val} + } + jsnum := val.Interface().(json.Number) + + // Attempt int first, then float64 + if i64, err := jsnum.Int64(); err == nil { + return dve.IntEncodeValue(ec, vw, reflect.ValueOf(i64)) + } + + f64, err := jsnum.Float64() + if err != nil { + return err + } + + return dve.FloatEncodeValue(ec, vw, reflect.ValueOf(f64)) +} + +// URLEncodeValue is the ValueEncoderFunc for url.URL. +func (dve DefaultValueEncoders) URLEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tURL { + return ValueEncoderError{Name: "URLEncodeValue", Types: []reflect.Type{tURL}, Received: val} + } + u := val.Interface().(url.URL) + return vw.WriteString(u.String()) +} + +// TimeEncodeValue is the ValueEncoderFunc for time.TIme. +// +// Deprecated: TimeEncodeValue is not registered by default. Use TimeCodec.EncodeValue instead. +func (dve DefaultValueEncoders) TimeEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tTime { + return ValueEncoderError{Name: "TimeEncodeValue", Types: []reflect.Type{tTime}, Received: val} + } + tt := val.Interface().(time.Time) + dt := primitive.NewDateTimeFromTime(tt) + return vw.WriteDateTime(int64(dt)) +} + +// ByteSliceEncodeValue is the ValueEncoderFunc for []byte. +// +// Deprecated: ByteSliceEncodeValue is not registered by default. Use ByteSliceCodec.EncodeValue instead. +func (dve DefaultValueEncoders) ByteSliceEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tByteSlice { + return ValueEncoderError{Name: "ByteSliceEncodeValue", Types: []reflect.Type{tByteSlice}, Received: val} + } + if val.IsNil() { + return vw.WriteNull() + } + return vw.WriteBinary(val.Interface().([]byte)) +} + +// MapEncodeValue is the ValueEncoderFunc for map[string]* types. +// +// Deprecated: MapEncodeValue is not registered by default. Use MapCodec.EncodeValue instead. +func (dve DefaultValueEncoders) MapEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Map || val.Type().Key().Kind() != reflect.String { + return ValueEncoderError{Name: "MapEncodeValue", Kinds: []reflect.Kind{reflect.Map}, Received: val} + } + + if val.IsNil() { + // If we have a nill map but we can't WriteNull, that means we're probably trying to encode + // to a TopLevel document. We can't currently tell if this is what actually happened, but if + // there's a deeper underlying problem, the error will also be returned from WriteDocument, + // so just continue. The operations on a map reflection value are valid, so we can call + // MapKeys within mapEncodeValue without a problem. + err := vw.WriteNull() + if err == nil { + return nil + } + } + + dw, err := vw.WriteDocument() + if err != nil { + return err + } + + return dve.mapEncodeValue(ec, dw, val, nil) +} + +// mapEncodeValue handles encoding of the values of a map. The collisionFn returns +// true if the provided key exists, this is mainly used for inline maps in the +// struct codec. +func (dve DefaultValueEncoders) mapEncodeValue(ec EncodeContext, dw bsonrw.DocumentWriter, val reflect.Value, collisionFn func(string) bool) error { + + elemType := val.Type().Elem() + encoder, err := ec.LookupEncoder(elemType) + if err != nil && elemType.Kind() != reflect.Interface { + return err + } + + keys := val.MapKeys() + for _, key := range keys { + if collisionFn != nil && collisionFn(key.String()) { + return fmt.Errorf("Key %s of inlined map conflicts with a struct field name", key) + } + + currEncoder, currVal, lookupErr := dve.lookupElementEncoder(ec, encoder, val.MapIndex(key)) + if lookupErr != nil && lookupErr != errInvalidValue { + return lookupErr + } + + vw, err := dw.WriteDocumentElement(key.String()) + if err != nil { + return err + } + + if lookupErr == errInvalidValue { + err = vw.WriteNull() + if err != nil { + return err + } + continue + } + + err = currEncoder.EncodeValue(ec, vw, currVal) + if err != nil { + return err + } + } + + return dw.WriteDocumentEnd() +} + +// ArrayEncodeValue is the ValueEncoderFunc for array types. +func (dve DefaultValueEncoders) ArrayEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Array { + return ValueEncoderError{Name: "ArrayEncodeValue", Kinds: []reflect.Kind{reflect.Array}, Received: val} + } + + // If we have a []primitive.E we want to treat it as a document instead of as an array. + if val.Type().Elem() == tE { + dw, err := vw.WriteDocument() + if err != nil { + return err + } + + for idx := 0; idx < val.Len(); idx++ { + e := val.Index(idx).Interface().(primitive.E) + err = encodeElement(ec, dw, e) + if err != nil { + return err + } + } + + return dw.WriteDocumentEnd() + } + + // If we have a []byte we want to treat it as a binary instead of as an array. + if val.Type().Elem() == tByte { + var byteSlice []byte + for idx := 0; idx < val.Len(); idx++ { + byteSlice = append(byteSlice, val.Index(idx).Interface().(byte)) + } + return vw.WriteBinary(byteSlice) + } + + aw, err := vw.WriteArray() + if err != nil { + return err + } + + elemType := val.Type().Elem() + encoder, err := ec.LookupEncoder(elemType) + if err != nil && elemType.Kind() != reflect.Interface { + return err + } + + for idx := 0; idx < val.Len(); idx++ { + currEncoder, currVal, lookupErr := dve.lookupElementEncoder(ec, encoder, val.Index(idx)) + if lookupErr != nil && lookupErr != errInvalidValue { + return lookupErr + } + + vw, err := aw.WriteArrayElement() + if err != nil { + return err + } + + if lookupErr == errInvalidValue { + err = vw.WriteNull() + if err != nil { + return err + } + continue + } + + err = currEncoder.EncodeValue(ec, vw, currVal) + if err != nil { + return err + } + } + return aw.WriteArrayEnd() +} + +// SliceEncodeValue is the ValueEncoderFunc for slice types. +// +// Deprecated: SliceEncodeValue is not registered by default. Use SliceCodec.EncodeValue instead. +func (dve DefaultValueEncoders) SliceEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Slice { + return ValueEncoderError{Name: "SliceEncodeValue", Kinds: []reflect.Kind{reflect.Slice}, Received: val} + } + + if val.IsNil() { + return vw.WriteNull() + } + + // If we have a []primitive.E we want to treat it as a document instead of as an array. + if val.Type().ConvertibleTo(tD) { + d := val.Convert(tD).Interface().(primitive.D) + + dw, err := vw.WriteDocument() + if err != nil { + return err + } + + for _, e := range d { + err = encodeElement(ec, dw, e) + if err != nil { + return err + } + } + + return dw.WriteDocumentEnd() + } + + aw, err := vw.WriteArray() + if err != nil { + return err + } + + elemType := val.Type().Elem() + encoder, err := ec.LookupEncoder(elemType) + if err != nil && elemType.Kind() != reflect.Interface { + return err + } + + for idx := 0; idx < val.Len(); idx++ { + currEncoder, currVal, lookupErr := dve.lookupElementEncoder(ec, encoder, val.Index(idx)) + if lookupErr != nil && lookupErr != errInvalidValue { + return lookupErr + } + + vw, err := aw.WriteArrayElement() + if err != nil { + return err + } + + if lookupErr == errInvalidValue { + err = vw.WriteNull() + if err != nil { + return err + } + continue + } + + err = currEncoder.EncodeValue(ec, vw, currVal) + if err != nil { + return err + } + } + return aw.WriteArrayEnd() +} + +func (dve DefaultValueEncoders) lookupElementEncoder(ec EncodeContext, origEncoder ValueEncoder, currVal reflect.Value) (ValueEncoder, reflect.Value, error) { + if origEncoder != nil || (currVal.Kind() != reflect.Interface) { + return origEncoder, currVal, nil + } + currVal = currVal.Elem() + if !currVal.IsValid() { + return nil, currVal, errInvalidValue + } + currEncoder, err := ec.LookupEncoder(currVal.Type()) + + return currEncoder, currVal, err +} + +// EmptyInterfaceEncodeValue is the ValueEncoderFunc for interface{}. +// +// Deprecated: EmptyInterfaceEncodeValue is not registered by default. Use EmptyInterfaceCodec.EncodeValue instead. +func (dve DefaultValueEncoders) EmptyInterfaceEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tEmpty { + return ValueEncoderError{Name: "EmptyInterfaceEncodeValue", Types: []reflect.Type{tEmpty}, Received: val} + } + + if val.IsNil() { + return vw.WriteNull() + } + encoder, err := ec.LookupEncoder(val.Elem().Type()) + if err != nil { + return err + } + + return encoder.EncodeValue(ec, vw, val.Elem()) +} + +// ValueMarshalerEncodeValue is the ValueEncoderFunc for ValueMarshaler implementations. +func (dve DefaultValueEncoders) ValueMarshalerEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + // Either val or a pointer to val must implement ValueMarshaler + switch { + case !val.IsValid(): + return ValueEncoderError{Name: "ValueMarshalerEncodeValue", Types: []reflect.Type{tValueMarshaler}, Received: val} + case val.Type().Implements(tValueMarshaler): + // If ValueMarshaler is implemented on a concrete type, make sure that val isn't a nil pointer + if isImplementationNil(val, tValueMarshaler) { + return vw.WriteNull() + } + case reflect.PtrTo(val.Type()).Implements(tValueMarshaler) && val.CanAddr(): + val = val.Addr() + default: + return ValueEncoderError{Name: "ValueMarshalerEncodeValue", Types: []reflect.Type{tValueMarshaler}, Received: val} + } + + fn := val.Convert(tValueMarshaler).MethodByName("MarshalBSONValue") + returns := fn.Call(nil) + if !returns[2].IsNil() { + return returns[2].Interface().(error) + } + t, data := returns[0].Interface().(bsontype.Type), returns[1].Interface().([]byte) + return bsonrw.Copier{}.CopyValueFromBytes(vw, t, data) +} + +// MarshalerEncodeValue is the ValueEncoderFunc for Marshaler implementations. +func (dve DefaultValueEncoders) MarshalerEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + // Either val or a pointer to val must implement Marshaler + switch { + case !val.IsValid(): + return ValueEncoderError{Name: "MarshalerEncodeValue", Types: []reflect.Type{tMarshaler}, Received: val} + case val.Type().Implements(tMarshaler): + // If Marshaler is implemented on a concrete type, make sure that val isn't a nil pointer + if isImplementationNil(val, tMarshaler) { + return vw.WriteNull() + } + case reflect.PtrTo(val.Type()).Implements(tMarshaler) && val.CanAddr(): + val = val.Addr() + default: + return ValueEncoderError{Name: "MarshalerEncodeValue", Types: []reflect.Type{tMarshaler}, Received: val} + } + + fn := val.Convert(tMarshaler).MethodByName("MarshalBSON") + returns := fn.Call(nil) + if !returns[1].IsNil() { + return returns[1].Interface().(error) + } + data := returns[0].Interface().([]byte) + return bsonrw.Copier{}.CopyValueFromBytes(vw, bsontype.EmbeddedDocument, data) +} + +// ProxyEncodeValue is the ValueEncoderFunc for Proxy implementations. +func (dve DefaultValueEncoders) ProxyEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + // Either val or a pointer to val must implement Proxy + switch { + case !val.IsValid(): + return ValueEncoderError{Name: "ProxyEncodeValue", Types: []reflect.Type{tProxy}, Received: val} + case val.Type().Implements(tProxy): + // If Proxy is implemented on a concrete type, make sure that val isn't a nil pointer + if isImplementationNil(val, tProxy) { + return vw.WriteNull() + } + case reflect.PtrTo(val.Type()).Implements(tProxy) && val.CanAddr(): + val = val.Addr() + default: + return ValueEncoderError{Name: "ProxyEncodeValue", Types: []reflect.Type{tProxy}, Received: val} + } + + fn := val.Convert(tProxy).MethodByName("ProxyBSON") + returns := fn.Call(nil) + if !returns[1].IsNil() { + return returns[1].Interface().(error) + } + data := returns[0] + var encoder ValueEncoder + var err error + if data.Elem().IsValid() { + encoder, err = ec.LookupEncoder(data.Elem().Type()) + } else { + encoder, err = ec.LookupEncoder(nil) + } + if err != nil { + return err + } + return encoder.EncodeValue(ec, vw, data.Elem()) +} + +// JavaScriptEncodeValue is the ValueEncoderFunc for the primitive.JavaScript type. +func (DefaultValueEncoders) JavaScriptEncodeValue(ectx EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tJavaScript { + return ValueEncoderError{Name: "JavaScriptEncodeValue", Types: []reflect.Type{tJavaScript}, Received: val} + } + + return vw.WriteJavascript(val.String()) +} + +// SymbolEncodeValue is the ValueEncoderFunc for the primitive.Symbol type. +func (DefaultValueEncoders) SymbolEncodeValue(ectx EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tSymbol { + return ValueEncoderError{Name: "SymbolEncodeValue", Types: []reflect.Type{tSymbol}, Received: val} + } + + return vw.WriteSymbol(val.String()) +} + +// BinaryEncodeValue is the ValueEncoderFunc for Binary. +func (DefaultValueEncoders) BinaryEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tBinary { + return ValueEncoderError{Name: "BinaryEncodeValue", Types: []reflect.Type{tBinary}, Received: val} + } + b := val.Interface().(primitive.Binary) + + return vw.WriteBinaryWithSubtype(b.Data, b.Subtype) +} + +// UndefinedEncodeValue is the ValueEncoderFunc for Undefined. +func (DefaultValueEncoders) UndefinedEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tUndefined { + return ValueEncoderError{Name: "UndefinedEncodeValue", Types: []reflect.Type{tUndefined}, Received: val} + } + + return vw.WriteUndefined() +} + +// DateTimeEncodeValue is the ValueEncoderFunc for DateTime. +func (DefaultValueEncoders) DateTimeEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tDateTime { + return ValueEncoderError{Name: "DateTimeEncodeValue", Types: []reflect.Type{tDateTime}, Received: val} + } + + return vw.WriteDateTime(val.Int()) +} + +// NullEncodeValue is the ValueEncoderFunc for Null. +func (DefaultValueEncoders) NullEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tNull { + return ValueEncoderError{Name: "NullEncodeValue", Types: []reflect.Type{tNull}, Received: val} + } + + return vw.WriteNull() +} + +// RegexEncodeValue is the ValueEncoderFunc for Regex. +func (DefaultValueEncoders) RegexEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tRegex { + return ValueEncoderError{Name: "RegexEncodeValue", Types: []reflect.Type{tRegex}, Received: val} + } + + regex := val.Interface().(primitive.Regex) + + return vw.WriteRegex(regex.Pattern, regex.Options) +} + +// DBPointerEncodeValue is the ValueEncoderFunc for DBPointer. +func (DefaultValueEncoders) DBPointerEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tDBPointer { + return ValueEncoderError{Name: "DBPointerEncodeValue", Types: []reflect.Type{tDBPointer}, Received: val} + } + + dbp := val.Interface().(primitive.DBPointer) + + return vw.WriteDBPointer(dbp.DB, dbp.Pointer) +} + +// TimestampEncodeValue is the ValueEncoderFunc for Timestamp. +func (DefaultValueEncoders) TimestampEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tTimestamp { + return ValueEncoderError{Name: "TimestampEncodeValue", Types: []reflect.Type{tTimestamp}, Received: val} + } + + ts := val.Interface().(primitive.Timestamp) + + return vw.WriteTimestamp(ts.T, ts.I) +} + +// MinKeyEncodeValue is the ValueEncoderFunc for MinKey. +func (DefaultValueEncoders) MinKeyEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tMinKey { + return ValueEncoderError{Name: "MinKeyEncodeValue", Types: []reflect.Type{tMinKey}, Received: val} + } + + return vw.WriteMinKey() +} + +// MaxKeyEncodeValue is the ValueEncoderFunc for MaxKey. +func (DefaultValueEncoders) MaxKeyEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tMaxKey { + return ValueEncoderError{Name: "MaxKeyEncodeValue", Types: []reflect.Type{tMaxKey}, Received: val} + } + + return vw.WriteMaxKey() +} + +// CoreDocumentEncodeValue is the ValueEncoderFunc for bsoncore.Document. +func (DefaultValueEncoders) CoreDocumentEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tCoreDocument { + return ValueEncoderError{Name: "CoreDocumentEncodeValue", Types: []reflect.Type{tCoreDocument}, Received: val} + } + + cdoc := val.Interface().(bsoncore.Document) + + return bsonrw.Copier{}.CopyDocumentFromBytes(vw, cdoc) +} + +// CodeWithScopeEncodeValue is the ValueEncoderFunc for CodeWithScope. +func (dve DefaultValueEncoders) CodeWithScopeEncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tCodeWithScope { + return ValueEncoderError{Name: "CodeWithScopeEncodeValue", Types: []reflect.Type{tCodeWithScope}, Received: val} + } + + cws := val.Interface().(primitive.CodeWithScope) + + dw, err := vw.WriteCodeWithScope(string(cws.Code)) + if err != nil { + return err + } + + sw := sliceWriterPool.Get().(*bsonrw.SliceWriter) + defer sliceWriterPool.Put(sw) + *sw = (*sw)[:0] + + scopeVW := bvwPool.Get(sw) + defer bvwPool.Put(scopeVW) + + encoder, err := ec.LookupEncoder(reflect.TypeOf(cws.Scope)) + if err != nil { + return err + } + + err = encoder.EncodeValue(ec, scopeVW, reflect.ValueOf(cws.Scope)) + if err != nil { + return err + } + + err = bsonrw.Copier{}.CopyBytesToDocumentWriter(dw, *sw) + if err != nil { + return err + } + return dw.WriteDocumentEnd() +} + +// isImplementationNil returns if val is a nil pointer and inter is implemented on a concrete type +func isImplementationNil(val reflect.Value, inter reflect.Type) bool { + vt := val.Type() + for vt.Kind() == reflect.Ptr { + vt = vt.Elem() + } + return vt.Implements(inter) && val.Kind() == reflect.Ptr && val.IsNil() +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/doc.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/doc.go new file mode 100644 index 0000000..c1e20f9 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/doc.go @@ -0,0 +1,84 @@ +// Package bsoncodec provides a system for encoding values to BSON representations and decoding +// values from BSON representations. This package considers both binary BSON and ExtendedJSON as +// BSON representations. The types in this package enable a flexible system for handling this +// encoding and decoding. +// +// The codec system is composed of two parts: +// +// 1) ValueEncoders and ValueDecoders that handle encoding and decoding Go values to and from BSON +// representations. +// +// 2) A Registry that holds these ValueEncoders and ValueDecoders and provides methods for +// retrieving them. +// +// ValueEncoders and ValueDecoders +// +// The ValueEncoder interface is implemented by types that can encode a provided Go type to BSON. +// The value to encode is provided as a reflect.Value and a bsonrw.ValueWriter is used within the +// EncodeValue method to actually create the BSON representation. For convenience, ValueEncoderFunc +// is provided to allow use of a function with the correct signature as a ValueEncoder. An +// EncodeContext instance is provided to allow implementations to lookup further ValueEncoders and +// to provide configuration information. +// +// The ValueDecoder interface is the inverse of the ValueEncoder. Implementations should ensure that +// the value they receive is settable. Similar to ValueEncoderFunc, ValueDecoderFunc is provided to +// allow the use of a function with the correct signature as a ValueDecoder. A DecodeContext +// instance is provided and serves similar functionality to the EncodeContext. +// +// Registry and RegistryBuilder +// +// A Registry is an immutable store for ValueEncoders, ValueDecoders, and a type map. See the Registry type +// documentation for examples of registering various custom encoders and decoders. A Registry can be constructed using a +// RegistryBuilder, which handles three main types of codecs: +// +// 1. Type encoders/decoders - These can be registered using the RegisterTypeEncoder and RegisterTypeDecoder methods. +// The registered codec will be invoked when encoding/decoding a value whose type matches the registered type exactly. +// If the registered type is an interface, the codec will be invoked when encoding or decoding values whose type is the +// interface, but not for values with concrete types that implement the interface. +// +// 2. Hook encoders/decoders - These can be registered using the RegisterHookEncoder and RegisterHookDecoder methods. +// These methods only accept interface types and the registered codecs will be invoked when encoding or decoding values +// whose types implement the interface. An example of a hook defined by the driver is bson.Marshaler. The driver will +// call the MarshalBSON method for any value whose type implements bson.Marshaler, regardless of the value's concrete +// type. +// +// 3. Type map entries - This can be used to associate a BSON type with a Go type. These type associations are used when +// decoding into a bson.D/bson.M or a struct field of type interface{}. For example, by default, BSON int32 and int64 +// values decode as Go int32 and int64 instances, respectively, when decoding into a bson.D. The following code would +// change the behavior so these values decode as Go int instances instead: +// +// intType := reflect.TypeOf(int(0)) +// registryBuilder.RegisterTypeMapEntry(bsontype.Int32, intType).RegisterTypeMapEntry(bsontype.Int64, intType) +// +// 4. Kind encoder/decoders - These can be registered using the RegisterDefaultEncoder and RegisterDefaultDecoder +// methods. The registered codec will be invoked when encoding or decoding values whose reflect.Kind matches the +// registered reflect.Kind as long as the value's type doesn't match a registered type or hook encoder/decoder first. +// These methods should be used to change the behavior for all values for a specific kind. +// +// Registry Lookup Procedure +// +// When looking up an encoder in a Registry, the precedence rules are as follows: +// +// 1. A type encoder registered for the exact type of the value. +// +// 2. A hook encoder registered for an interface that is implemented by the value or by a pointer to the value. If the +// value matches multiple hooks (e.g. the type implements bsoncodec.Marshaler and bsoncodec.ValueMarshaler), the first +// one registered will be selected. Note that registries constructed using bson.NewRegistryBuilder have driver-defined +// hooks registered for the bsoncodec.Marshaler, bsoncodec.ValueMarshaler, and bsoncodec.Proxy interfaces, so those +// will take precedence over any new hooks. +// +// 3. A kind encoder registered for the value's kind. +// +// If all of these lookups fail to find an encoder, an error of type ErrNoEncoder is returned. The same precedence +// rules apply for decoders, with the exception that an error of type ErrNoDecoder will be returned if no decoder is +// found. +// +// DefaultValueEncoders and DefaultValueDecoders +// +// The DefaultValueEncoders and DefaultValueDecoders types provide a full set of ValueEncoders and +// ValueDecoders for handling a wide range of Go types, including all of the types within the +// primitive package. To make registering these codecs easier, a helper method on each type is +// provided. For the DefaultValueEncoders type the method is called RegisterDefaultEncoders and for +// the DefaultValueDecoders type the method is called RegisterDefaultDecoders, this method also +// handles registering type map entries for each BSON type. +package bsoncodec diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/empty_interface_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/empty_interface_codec.go new file mode 100644 index 0000000..a15636d --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/empty_interface_codec.go @@ -0,0 +1,140 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" +) + +// EmptyInterfaceCodec is the Codec used for interface{} values. +type EmptyInterfaceCodec struct { + DecodeBinaryAsSlice bool +} + +var ( + defaultEmptyInterfaceCodec = NewEmptyInterfaceCodec() + + _ ValueCodec = defaultEmptyInterfaceCodec + _ typeDecoder = defaultEmptyInterfaceCodec +) + +// NewEmptyInterfaceCodec returns a EmptyInterfaceCodec with options opts. +func NewEmptyInterfaceCodec(opts ...*bsonoptions.EmptyInterfaceCodecOptions) *EmptyInterfaceCodec { + interfaceOpt := bsonoptions.MergeEmptyInterfaceCodecOptions(opts...) + + codec := EmptyInterfaceCodec{} + if interfaceOpt.DecodeBinaryAsSlice != nil { + codec.DecodeBinaryAsSlice = *interfaceOpt.DecodeBinaryAsSlice + } + return &codec +} + +// EncodeValue is the ValueEncoderFunc for interface{}. +func (eic EmptyInterfaceCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tEmpty { + return ValueEncoderError{Name: "EmptyInterfaceEncodeValue", Types: []reflect.Type{tEmpty}, Received: val} + } + + if val.IsNil() { + return vw.WriteNull() + } + encoder, err := ec.LookupEncoder(val.Elem().Type()) + if err != nil { + return err + } + + return encoder.EncodeValue(ec, vw, val.Elem()) +} + +func (eic EmptyInterfaceCodec) getEmptyInterfaceDecodeType(dc DecodeContext, valueType bsontype.Type) (reflect.Type, error) { + isDocument := valueType == bsontype.Type(0) || valueType == bsontype.EmbeddedDocument + if isDocument && dc.Ancestor != nil { + // Using ancestor information rather than looking up the type map entry forces consistent decoding. + // If we're decoding into a bson.D, subdocuments should also be decoded as bson.D, even if a type map entry + // has been registered. + return dc.Ancestor, nil + } + + rtype, err := dc.LookupTypeMapEntry(valueType) + if err == nil { + return rtype, nil + } + + if isDocument { + // For documents, fallback to looking up a type map entry for bsontype.Type(0) or bsontype.EmbeddedDocument, + // depending on the original valueType. + var lookupType bsontype.Type + switch valueType { + case bsontype.Type(0): + lookupType = bsontype.EmbeddedDocument + case bsontype.EmbeddedDocument: + lookupType = bsontype.Type(0) + } + + rtype, err = dc.LookupTypeMapEntry(lookupType) + if err == nil { + return rtype, nil + } + } + + return nil, err +} + +func (eic EmptyInterfaceCodec) decodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tEmpty { + return emptyValue, ValueDecoderError{Name: "EmptyInterfaceDecodeValue", Types: []reflect.Type{tEmpty}, Received: reflect.Zero(t)} + } + + rtype, err := eic.getEmptyInterfaceDecodeType(dc, vr.Type()) + if err != nil { + switch vr.Type() { + case bsontype.Null: + return reflect.Zero(t), vr.ReadNull() + default: + return emptyValue, err + } + } + + decoder, err := dc.LookupDecoder(rtype) + if err != nil { + return emptyValue, err + } + + elem, err := decodeTypeOrValue(decoder, dc, vr, rtype) + if err != nil { + return emptyValue, err + } + + if eic.DecodeBinaryAsSlice && rtype == tBinary { + binElem := elem.Interface().(primitive.Binary) + if binElem.Subtype == bsontype.BinaryGeneric || binElem.Subtype == bsontype.BinaryBinaryOld { + elem = reflect.ValueOf(binElem.Data) + } + } + + return elem, nil +} + +// DecodeValue is the ValueDecoderFunc for interface{}. +func (eic EmptyInterfaceCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tEmpty { + return ValueDecoderError{Name: "EmptyInterfaceDecodeValue", Types: []reflect.Type{tEmpty}, Received: val} + } + + elem, err := eic.decodeType(dc, vr, val.Type()) + if err != nil { + return err + } + + val.Set(elem) + return nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/map_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/map_codec.go new file mode 100644 index 0000000..1f7acbc --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/map_codec.go @@ -0,0 +1,288 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "fmt" + "reflect" + "strconv" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +var defaultMapCodec = NewMapCodec() + +// MapCodec is the Codec used for map values. +type MapCodec struct { + DecodeZerosMap bool + EncodeNilAsEmpty bool + EncodeKeysWithStringer bool +} + +var _ ValueCodec = &MapCodec{} + +// KeyMarshaler is the interface implemented by an object that can marshal itself into a string key. +// This applies to types used as map keys and is similar to encoding.TextMarshaler. +type KeyMarshaler interface { + MarshalKey() (key string, err error) +} + +// KeyUnmarshaler is the interface implemented by an object that can unmarshal a string representation +// of itself. This applies to types used as map keys and is similar to encoding.TextUnmarshaler. +// +// UnmarshalKey must be able to decode the form generated by MarshalKey. +// UnmarshalKey must copy the text if it wishes to retain the text +// after returning. +type KeyUnmarshaler interface { + UnmarshalKey(key string) error +} + +// NewMapCodec returns a MapCodec with options opts. +func NewMapCodec(opts ...*bsonoptions.MapCodecOptions) *MapCodec { + mapOpt := bsonoptions.MergeMapCodecOptions(opts...) + + codec := MapCodec{} + if mapOpt.DecodeZerosMap != nil { + codec.DecodeZerosMap = *mapOpt.DecodeZerosMap + } + if mapOpt.EncodeNilAsEmpty != nil { + codec.EncodeNilAsEmpty = *mapOpt.EncodeNilAsEmpty + } + if mapOpt.EncodeKeysWithStringer != nil { + codec.EncodeKeysWithStringer = *mapOpt.EncodeKeysWithStringer + } + return &codec +} + +// EncodeValue is the ValueEncoder for map[*]* types. +func (mc *MapCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Map { + return ValueEncoderError{Name: "MapEncodeValue", Kinds: []reflect.Kind{reflect.Map}, Received: val} + } + + if val.IsNil() && !mc.EncodeNilAsEmpty { + // If we have a nil map but we can't WriteNull, that means we're probably trying to encode + // to a TopLevel document. We can't currently tell if this is what actually happened, but if + // there's a deeper underlying problem, the error will also be returned from WriteDocument, + // so just continue. The operations on a map reflection value are valid, so we can call + // MapKeys within mapEncodeValue without a problem. + err := vw.WriteNull() + if err == nil { + return nil + } + } + + dw, err := vw.WriteDocument() + if err != nil { + return err + } + + return mc.mapEncodeValue(ec, dw, val, nil) +} + +// mapEncodeValue handles encoding of the values of a map. The collisionFn returns +// true if the provided key exists, this is mainly used for inline maps in the +// struct codec. +func (mc *MapCodec) mapEncodeValue(ec EncodeContext, dw bsonrw.DocumentWriter, val reflect.Value, collisionFn func(string) bool) error { + + elemType := val.Type().Elem() + encoder, err := ec.LookupEncoder(elemType) + if err != nil && elemType.Kind() != reflect.Interface { + return err + } + + keys := val.MapKeys() + for _, key := range keys { + keyStr, err := mc.encodeKey(key) + if err != nil { + return err + } + + if collisionFn != nil && collisionFn(keyStr) { + return fmt.Errorf("Key %s of inlined map conflicts with a struct field name", key) + } + + currEncoder, currVal, lookupErr := defaultValueEncoders.lookupElementEncoder(ec, encoder, val.MapIndex(key)) + if lookupErr != nil && lookupErr != errInvalidValue { + return lookupErr + } + + vw, err := dw.WriteDocumentElement(keyStr) + if err != nil { + return err + } + + if lookupErr == errInvalidValue { + err = vw.WriteNull() + if err != nil { + return err + } + continue + } + + err = currEncoder.EncodeValue(ec, vw, currVal) + if err != nil { + return err + } + } + + return dw.WriteDocumentEnd() +} + +// DecodeValue is the ValueDecoder for map[string/decimal]* types. +func (mc *MapCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if val.Kind() != reflect.Map || (!val.CanSet() && val.IsNil()) { + return ValueDecoderError{Name: "MapDecodeValue", Kinds: []reflect.Kind{reflect.Map}, Received: val} + } + + switch vrType := vr.Type(); vrType { + case bsontype.Type(0), bsontype.EmbeddedDocument: + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + case bsontype.Undefined: + val.Set(reflect.Zero(val.Type())) + return vr.ReadUndefined() + default: + return fmt.Errorf("cannot decode %v into a %s", vrType, val.Type()) + } + + dr, err := vr.ReadDocument() + if err != nil { + return err + } + + if val.IsNil() { + val.Set(reflect.MakeMap(val.Type())) + } + + if val.Len() > 0 && mc.DecodeZerosMap { + clearMap(val) + } + + eType := val.Type().Elem() + decoder, err := dc.LookupDecoder(eType) + if err != nil { + return err + } + eTypeDecoder, _ := decoder.(typeDecoder) + + if eType == tEmpty { + dc.Ancestor = val.Type() + } + + keyType := val.Type().Key() + + for { + key, vr, err := dr.ReadElement() + if err == bsonrw.ErrEOD { + break + } + if err != nil { + return err + } + + k, err := mc.decodeKey(key, keyType) + if err != nil { + return err + } + + elem, err := decodeTypeOrValueWithInfo(decoder, eTypeDecoder, dc, vr, eType, true) + if err != nil { + return newDecodeError(key, err) + } + + val.SetMapIndex(k, elem) + } + return nil +} + +func clearMap(m reflect.Value) { + var none reflect.Value + for _, k := range m.MapKeys() { + m.SetMapIndex(k, none) + } +} + +func (mc *MapCodec) encodeKey(val reflect.Value) (string, error) { + if mc.EncodeKeysWithStringer { + return fmt.Sprint(val), nil + } + + // keys of any string type are used directly + if val.Kind() == reflect.String { + return val.String(), nil + } + // KeyMarshalers are marshaled + if km, ok := val.Interface().(KeyMarshaler); ok { + if val.Kind() == reflect.Ptr && val.IsNil() { + return "", nil + } + buf, err := km.MarshalKey() + if err == nil { + return buf, nil + } + return "", err + } + + switch val.Kind() { + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64: + return strconv.FormatInt(val.Int(), 10), nil + case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr: + return strconv.FormatUint(val.Uint(), 10), nil + } + return "", fmt.Errorf("unsupported key type: %v", val.Type()) +} + +var keyUnmarshalerType = reflect.TypeOf((*KeyUnmarshaler)(nil)).Elem() + +func (mc *MapCodec) decodeKey(key string, keyType reflect.Type) (reflect.Value, error) { + keyVal := reflect.ValueOf(key) + var err error + switch { + // First, if EncodeKeysWithStringer is not enabled, try to decode withKeyUnmarshaler + case !mc.EncodeKeysWithStringer && reflect.PtrTo(keyType).Implements(keyUnmarshalerType): + keyVal = reflect.New(keyType) + v := keyVal.Interface().(KeyUnmarshaler) + err = v.UnmarshalKey(key) + keyVal = keyVal.Elem() + // Otherwise, go to type specific behavior + default: + switch keyType.Kind() { + case reflect.String: + keyVal = reflect.ValueOf(key).Convert(keyType) + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64: + n, parseErr := strconv.ParseInt(key, 10, 64) + if parseErr != nil || reflect.Zero(keyType).OverflowInt(n) { + err = fmt.Errorf("failed to unmarshal number key %v", key) + } + keyVal = reflect.ValueOf(n).Convert(keyType) + case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr: + n, parseErr := strconv.ParseUint(key, 10, 64) + if parseErr != nil || reflect.Zero(keyType).OverflowUint(n) { + err = fmt.Errorf("failed to unmarshal number key %v", key) + break + } + keyVal = reflect.ValueOf(n).Convert(keyType) + case reflect.Float32, reflect.Float64: + if mc.EncodeKeysWithStringer { + parsed, err := strconv.ParseFloat(key, 64) + if err != nil { + return keyVal, fmt.Errorf("Map key is defined to be a decimal type (%v) but got error %v", keyType.Kind(), err) + } + keyVal = reflect.ValueOf(parsed) + break + } + fallthrough + default: + return keyVal, fmt.Errorf("unsupported key type: %v", keyType) + } + } + return keyVal, err +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/mode.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/mode.go new file mode 100644 index 0000000..fbd9f0a --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/mode.go @@ -0,0 +1,65 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import "fmt" + +type mode int + +const ( + _ mode = iota + mTopLevel + mDocument + mArray + mValue + mElement + mCodeWithScope + mSpacer +) + +func (m mode) String() string { + var str string + + switch m { + case mTopLevel: + str = "TopLevel" + case mDocument: + str = "DocumentMode" + case mArray: + str = "ArrayMode" + case mValue: + str = "ValueMode" + case mElement: + str = "ElementMode" + case mCodeWithScope: + str = "CodeWithScopeMode" + case mSpacer: + str = "CodeWithScopeSpacerFrame" + default: + str = "UnknownMode" + } + + return str +} + +// TransitionError is an error returned when an invalid progressing a +// ValueReader or ValueWriter state machine occurs. +type TransitionError struct { + parent mode + current mode + destination mode +} + +func (te TransitionError) Error() string { + if te.destination == mode(0) { + return fmt.Sprintf("invalid state transition: cannot read/write value while in %s", te.current) + } + if te.parent == mode(0) { + return fmt.Sprintf("invalid state transition: %s -> %s", te.current, te.destination) + } + return fmt.Sprintf("invalid state transition: %s -> %s; parent %s", te.current, te.destination, te.parent) +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/pointer_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/pointer_codec.go new file mode 100644 index 0000000..616a3e7 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/pointer_codec.go @@ -0,0 +1,109 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "reflect" + "sync" + + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +var _ ValueEncoder = &PointerCodec{} +var _ ValueDecoder = &PointerCodec{} + +// PointerCodec is the Codec used for pointers. +type PointerCodec struct { + ecache map[reflect.Type]ValueEncoder + dcache map[reflect.Type]ValueDecoder + l sync.RWMutex +} + +// NewPointerCodec returns a PointerCodec that has been initialized. +func NewPointerCodec() *PointerCodec { + return &PointerCodec{ + ecache: make(map[reflect.Type]ValueEncoder), + dcache: make(map[reflect.Type]ValueDecoder), + } +} + +// EncodeValue handles encoding a pointer by either encoding it to BSON Null if the pointer is nil +// or looking up an encoder for the type of value the pointer points to. +func (pc *PointerCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if val.Kind() != reflect.Ptr { + if !val.IsValid() { + return vw.WriteNull() + } + return ValueEncoderError{Name: "PointerCodec.EncodeValue", Kinds: []reflect.Kind{reflect.Ptr}, Received: val} + } + + if val.IsNil() { + return vw.WriteNull() + } + + pc.l.RLock() + enc, ok := pc.ecache[val.Type()] + pc.l.RUnlock() + if ok { + if enc == nil { + return ErrNoEncoder{Type: val.Type()} + } + return enc.EncodeValue(ec, vw, val.Elem()) + } + + enc, err := ec.LookupEncoder(val.Type().Elem()) + pc.l.Lock() + pc.ecache[val.Type()] = enc + pc.l.Unlock() + if err != nil { + return err + } + + return enc.EncodeValue(ec, vw, val.Elem()) +} + +// DecodeValue handles decoding a pointer by looking up a decoder for the type it points to and +// using that to decode. If the BSON value is Null, this method will set the pointer to nil. +func (pc *PointerCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Kind() != reflect.Ptr { + return ValueDecoderError{Name: "PointerCodec.DecodeValue", Kinds: []reflect.Kind{reflect.Ptr}, Received: val} + } + + if vr.Type() == bsontype.Null { + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + } + if vr.Type() == bsontype.Undefined { + val.Set(reflect.Zero(val.Type())) + return vr.ReadUndefined() + } + + if val.IsNil() { + val.Set(reflect.New(val.Type().Elem())) + } + + pc.l.RLock() + dec, ok := pc.dcache[val.Type()] + pc.l.RUnlock() + if ok { + if dec == nil { + return ErrNoDecoder{Type: val.Type()} + } + return dec.DecodeValue(dc, vr, val.Elem()) + } + + dec, err := dc.LookupDecoder(val.Type().Elem()) + pc.l.Lock() + pc.dcache[val.Type()] = dec + pc.l.Unlock() + if err != nil { + return err + } + + return dec.DecodeValue(dc, vr, val.Elem()) +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/proxy.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/proxy.go new file mode 100644 index 0000000..4cf2b01 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/proxy.go @@ -0,0 +1,14 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +// Proxy is an interface implemented by types that cannot themselves be directly encoded. Types +// that implement this interface with have ProxyBSON called during the encoding process and that +// value will be encoded in place for the implementer. +type Proxy interface { + ProxyBSON() (interface{}, error) +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/registry.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/registry.go new file mode 100644 index 0000000..f6f3800 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/registry.go @@ -0,0 +1,468 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "errors" + "fmt" + "reflect" + "sync" + + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +// ErrNilType is returned when nil is passed to either LookupEncoder or LookupDecoder. +var ErrNilType = errors.New("cannot perform a decoder lookup on ") + +// ErrNotPointer is returned when a non-pointer type is provided to LookupDecoder. +var ErrNotPointer = errors.New("non-pointer provided to LookupDecoder") + +// ErrNoEncoder is returned when there wasn't an encoder available for a type. +type ErrNoEncoder struct { + Type reflect.Type +} + +func (ene ErrNoEncoder) Error() string { + if ene.Type == nil { + return "no encoder found for " + } + return "no encoder found for " + ene.Type.String() +} + +// ErrNoDecoder is returned when there wasn't a decoder available for a type. +type ErrNoDecoder struct { + Type reflect.Type +} + +func (end ErrNoDecoder) Error() string { + return "no decoder found for " + end.Type.String() +} + +// ErrNoTypeMapEntry is returned when there wasn't a type available for the provided BSON type. +type ErrNoTypeMapEntry struct { + Type bsontype.Type +} + +func (entme ErrNoTypeMapEntry) Error() string { + return "no type map entry found for " + entme.Type.String() +} + +// ErrNotInterface is returned when the provided type is not an interface. +var ErrNotInterface = errors.New("The provided type is not an interface") + +// A RegistryBuilder is used to build a Registry. This type is not goroutine +// safe. +type RegistryBuilder struct { + typeEncoders map[reflect.Type]ValueEncoder + interfaceEncoders []interfaceValueEncoder + kindEncoders map[reflect.Kind]ValueEncoder + + typeDecoders map[reflect.Type]ValueDecoder + interfaceDecoders []interfaceValueDecoder + kindDecoders map[reflect.Kind]ValueDecoder + + typeMap map[bsontype.Type]reflect.Type +} + +// A Registry is used to store and retrieve codecs for types and interfaces. This type is the main +// typed passed around and Encoders and Decoders are constructed from it. +type Registry struct { + typeEncoders map[reflect.Type]ValueEncoder + typeDecoders map[reflect.Type]ValueDecoder + + interfaceEncoders []interfaceValueEncoder + interfaceDecoders []interfaceValueDecoder + + kindEncoders map[reflect.Kind]ValueEncoder + kindDecoders map[reflect.Kind]ValueDecoder + + typeMap map[bsontype.Type]reflect.Type + + mu sync.RWMutex +} + +// NewRegistryBuilder creates a new empty RegistryBuilder. +func NewRegistryBuilder() *RegistryBuilder { + return &RegistryBuilder{ + typeEncoders: make(map[reflect.Type]ValueEncoder), + typeDecoders: make(map[reflect.Type]ValueDecoder), + + interfaceEncoders: make([]interfaceValueEncoder, 0), + interfaceDecoders: make([]interfaceValueDecoder, 0), + + kindEncoders: make(map[reflect.Kind]ValueEncoder), + kindDecoders: make(map[reflect.Kind]ValueDecoder), + + typeMap: make(map[bsontype.Type]reflect.Type), + } +} + +func buildDefaultRegistry() *Registry { + rb := NewRegistryBuilder() + defaultValueEncoders.RegisterDefaultEncoders(rb) + defaultValueDecoders.RegisterDefaultDecoders(rb) + return rb.Build() +} + +// RegisterCodec will register the provided ValueCodec for the provided type. +func (rb *RegistryBuilder) RegisterCodec(t reflect.Type, codec ValueCodec) *RegistryBuilder { + rb.RegisterTypeEncoder(t, codec) + rb.RegisterTypeDecoder(t, codec) + return rb +} + +// RegisterTypeEncoder will register the provided ValueEncoder for the provided type. +// +// The type will be used directly, so an encoder can be registered for a type and a different encoder can be registered +// for a pointer to that type. +// +// If the given type is an interface, the encoder will be called when marshalling a type that is that interface. It +// will not be called when marshalling a non-interface type that implements the interface. +func (rb *RegistryBuilder) RegisterTypeEncoder(t reflect.Type, enc ValueEncoder) *RegistryBuilder { + rb.typeEncoders[t] = enc + return rb +} + +// RegisterHookEncoder will register an encoder for the provided interface type t. This encoder will be called when +// marshalling a type if the type implements t or a pointer to the type implements t. If the provided type is not +// an interface (i.e. t.Kind() != reflect.Interface), this method will panic. +func (rb *RegistryBuilder) RegisterHookEncoder(t reflect.Type, enc ValueEncoder) *RegistryBuilder { + if t.Kind() != reflect.Interface { + panicStr := fmt.Sprintf("RegisterHookEncoder expects a type with kind reflect.Interface, "+ + "got type %s with kind %s", t, t.Kind()) + panic(panicStr) + } + + for idx, encoder := range rb.interfaceEncoders { + if encoder.i == t { + rb.interfaceEncoders[idx].ve = enc + return rb + } + } + + rb.interfaceEncoders = append(rb.interfaceEncoders, interfaceValueEncoder{i: t, ve: enc}) + return rb +} + +// RegisterTypeDecoder will register the provided ValueDecoder for the provided type. +// +// The type will be used directly, so a decoder can be registered for a type and a different decoder can be registered +// for a pointer to that type. +// +// If the given type is an interface, the decoder will be called when unmarshalling into a type that is that interface. +// It will not be called when unmarshalling into a non-interface type that implements the interface. +func (rb *RegistryBuilder) RegisterTypeDecoder(t reflect.Type, dec ValueDecoder) *RegistryBuilder { + rb.typeDecoders[t] = dec + return rb +} + +// RegisterHookDecoder will register an decoder for the provided interface type t. This decoder will be called when +// unmarshalling into a type if the type implements t or a pointer to the type implements t. If the provided type is not +// an interface (i.e. t.Kind() != reflect.Interface), this method will panic. +func (rb *RegistryBuilder) RegisterHookDecoder(t reflect.Type, dec ValueDecoder) *RegistryBuilder { + if t.Kind() != reflect.Interface { + panicStr := fmt.Sprintf("RegisterHookDecoder expects a type with kind reflect.Interface, "+ + "got type %s with kind %s", t, t.Kind()) + panic(panicStr) + } + + for idx, decoder := range rb.interfaceDecoders { + if decoder.i == t { + rb.interfaceDecoders[idx].vd = dec + return rb + } + } + + rb.interfaceDecoders = append(rb.interfaceDecoders, interfaceValueDecoder{i: t, vd: dec}) + return rb +} + +// RegisterEncoder registers the provided type and encoder pair. +// +// Deprecated: Use RegisterTypeEncoder or RegisterHookEncoder instead. +func (rb *RegistryBuilder) RegisterEncoder(t reflect.Type, enc ValueEncoder) *RegistryBuilder { + if t == tEmpty { + rb.typeEncoders[t] = enc + return rb + } + switch t.Kind() { + case reflect.Interface: + for idx, ir := range rb.interfaceEncoders { + if ir.i == t { + rb.interfaceEncoders[idx].ve = enc + return rb + } + } + + rb.interfaceEncoders = append(rb.interfaceEncoders, interfaceValueEncoder{i: t, ve: enc}) + default: + rb.typeEncoders[t] = enc + } + return rb +} + +// RegisterDecoder registers the provided type and decoder pair. +// +// Deprecated: Use RegisterTypeDecoder or RegisterHookDecoder instead. +func (rb *RegistryBuilder) RegisterDecoder(t reflect.Type, dec ValueDecoder) *RegistryBuilder { + if t == nil { + rb.typeDecoders[nil] = dec + return rb + } + if t == tEmpty { + rb.typeDecoders[t] = dec + return rb + } + switch t.Kind() { + case reflect.Interface: + for idx, ir := range rb.interfaceDecoders { + if ir.i == t { + rb.interfaceDecoders[idx].vd = dec + return rb + } + } + + rb.interfaceDecoders = append(rb.interfaceDecoders, interfaceValueDecoder{i: t, vd: dec}) + default: + rb.typeDecoders[t] = dec + } + return rb +} + +// RegisterDefaultEncoder will registr the provided ValueEncoder to the provided +// kind. +func (rb *RegistryBuilder) RegisterDefaultEncoder(kind reflect.Kind, enc ValueEncoder) *RegistryBuilder { + rb.kindEncoders[kind] = enc + return rb +} + +// RegisterDefaultDecoder will register the provided ValueDecoder to the +// provided kind. +func (rb *RegistryBuilder) RegisterDefaultDecoder(kind reflect.Kind, dec ValueDecoder) *RegistryBuilder { + rb.kindDecoders[kind] = dec + return rb +} + +// RegisterTypeMapEntry will register the provided type to the BSON type. The primary usage for this +// mapping is decoding situations where an empty interface is used and a default type needs to be +// created and decoded into. +// +// By default, BSON documents will decode into interface{} values as bson.D. To change the default type for BSON +// documents, a type map entry for bsontype.EmbeddedDocument should be registered. For example, to force BSON documents +// to decode to bson.Raw, use the following code: +// rb.RegisterTypeMapEntry(bsontype.EmbeddedDocument, reflect.TypeOf(bson.Raw{})) +func (rb *RegistryBuilder) RegisterTypeMapEntry(bt bsontype.Type, rt reflect.Type) *RegistryBuilder { + rb.typeMap[bt] = rt + return rb +} + +// Build creates a Registry from the current state of this RegistryBuilder. +func (rb *RegistryBuilder) Build() *Registry { + registry := new(Registry) + + registry.typeEncoders = make(map[reflect.Type]ValueEncoder) + for t, enc := range rb.typeEncoders { + registry.typeEncoders[t] = enc + } + + registry.typeDecoders = make(map[reflect.Type]ValueDecoder) + for t, dec := range rb.typeDecoders { + registry.typeDecoders[t] = dec + } + + registry.interfaceEncoders = make([]interfaceValueEncoder, len(rb.interfaceEncoders)) + copy(registry.interfaceEncoders, rb.interfaceEncoders) + + registry.interfaceDecoders = make([]interfaceValueDecoder, len(rb.interfaceDecoders)) + copy(registry.interfaceDecoders, rb.interfaceDecoders) + + registry.kindEncoders = make(map[reflect.Kind]ValueEncoder) + for kind, enc := range rb.kindEncoders { + registry.kindEncoders[kind] = enc + } + + registry.kindDecoders = make(map[reflect.Kind]ValueDecoder) + for kind, dec := range rb.kindDecoders { + registry.kindDecoders[kind] = dec + } + + registry.typeMap = make(map[bsontype.Type]reflect.Type) + for bt, rt := range rb.typeMap { + registry.typeMap[bt] = rt + } + + return registry +} + +// LookupEncoder inspects the registry for an encoder for the given type. The lookup precedence works as follows: +// +// 1. An encoder registered for the exact type. If the given type represents an interface, an encoder registered using +// RegisterTypeEncoder for the interface will be selected. +// +// 2. An encoder registered using RegisterHookEncoder for an interface implemented by the type or by a pointer to the +// type. +// +// 3. An encoder registered for the reflect.Kind of the value. +// +// If no encoder is found, an error of type ErrNoEncoder is returned. +func (r *Registry) LookupEncoder(t reflect.Type) (ValueEncoder, error) { + encodererr := ErrNoEncoder{Type: t} + r.mu.RLock() + enc, found := r.lookupTypeEncoder(t) + r.mu.RUnlock() + if found { + if enc == nil { + return nil, ErrNoEncoder{Type: t} + } + return enc, nil + } + + enc, found = r.lookupInterfaceEncoder(t, true) + if found { + r.mu.Lock() + r.typeEncoders[t] = enc + r.mu.Unlock() + return enc, nil + } + + if t == nil { + r.mu.Lock() + r.typeEncoders[t] = nil + r.mu.Unlock() + return nil, encodererr + } + + enc, found = r.kindEncoders[t.Kind()] + if !found { + r.mu.Lock() + r.typeEncoders[t] = nil + r.mu.Unlock() + return nil, encodererr + } + + r.mu.Lock() + r.typeEncoders[t] = enc + r.mu.Unlock() + return enc, nil +} + +func (r *Registry) lookupTypeEncoder(t reflect.Type) (ValueEncoder, bool) { + enc, found := r.typeEncoders[t] + return enc, found +} + +func (r *Registry) lookupInterfaceEncoder(t reflect.Type, allowAddr bool) (ValueEncoder, bool) { + if t == nil { + return nil, false + } + for _, ienc := range r.interfaceEncoders { + if t.Implements(ienc.i) { + return ienc.ve, true + } + if allowAddr && t.Kind() != reflect.Ptr && reflect.PtrTo(t).Implements(ienc.i) { + // if *t implements an interface, this will catch if t implements an interface further ahead + // in interfaceEncoders + defaultEnc, found := r.lookupInterfaceEncoder(t, false) + if !found { + defaultEnc = r.kindEncoders[t.Kind()] + } + return newCondAddrEncoder(ienc.ve, defaultEnc), true + } + } + return nil, false +} + +// LookupDecoder inspects the registry for an decoder for the given type. The lookup precedence works as follows: +// +// 1. A decoder registered for the exact type. If the given type represents an interface, a decoder registered using +// RegisterTypeDecoder for the interface will be selected. +// +// 2. A decoder registered using RegisterHookDecoder for an interface implemented by the type or by a pointer to the +// type. +// +// 3. A decoder registered for the reflect.Kind of the value. +// +// If no decoder is found, an error of type ErrNoDecoder is returned. +func (r *Registry) LookupDecoder(t reflect.Type) (ValueDecoder, error) { + if t == nil { + return nil, ErrNilType + } + decodererr := ErrNoDecoder{Type: t} + r.mu.RLock() + dec, found := r.lookupTypeDecoder(t) + r.mu.RUnlock() + if found { + if dec == nil { + return nil, ErrNoDecoder{Type: t} + } + return dec, nil + } + + dec, found = r.lookupInterfaceDecoder(t, true) + if found { + r.mu.Lock() + r.typeDecoders[t] = dec + r.mu.Unlock() + return dec, nil + } + + dec, found = r.kindDecoders[t.Kind()] + if !found { + r.mu.Lock() + r.typeDecoders[t] = nil + r.mu.Unlock() + return nil, decodererr + } + + r.mu.Lock() + r.typeDecoders[t] = dec + r.mu.Unlock() + return dec, nil +} + +func (r *Registry) lookupTypeDecoder(t reflect.Type) (ValueDecoder, bool) { + dec, found := r.typeDecoders[t] + return dec, found +} + +func (r *Registry) lookupInterfaceDecoder(t reflect.Type, allowAddr bool) (ValueDecoder, bool) { + for _, idec := range r.interfaceDecoders { + if t.Implements(idec.i) { + return idec.vd, true + } + if allowAddr && t.Kind() != reflect.Ptr && reflect.PtrTo(t).Implements(idec.i) { + // if *t implements an interface, this will catch if t implements an interface further ahead + // in interfaceDecoders + defaultDec, found := r.lookupInterfaceDecoder(t, false) + if !found { + defaultDec = r.kindDecoders[t.Kind()] + } + return newCondAddrDecoder(idec.vd, defaultDec), true + } + } + return nil, false +} + +// LookupTypeMapEntry inspects the registry's type map for a Go type for the corresponding BSON +// type. If no type is found, ErrNoTypeMapEntry is returned. +func (r *Registry) LookupTypeMapEntry(bt bsontype.Type) (reflect.Type, error) { + t, ok := r.typeMap[bt] + if !ok || t == nil { + return nil, ErrNoTypeMapEntry{Type: bt} + } + return t, nil +} + +type interfaceValueEncoder struct { + i reflect.Type + ve ValueEncoder +} + +type interfaceValueDecoder struct { + i reflect.Type + vd ValueDecoder +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/slice_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/slice_codec.go new file mode 100644 index 0000000..3c1b6b8 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/slice_codec.go @@ -0,0 +1,199 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "fmt" + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" +) + +var defaultSliceCodec = NewSliceCodec() + +// SliceCodec is the Codec used for slice values. +type SliceCodec struct { + EncodeNilAsEmpty bool +} + +var _ ValueCodec = &MapCodec{} + +// NewSliceCodec returns a MapCodec with options opts. +func NewSliceCodec(opts ...*bsonoptions.SliceCodecOptions) *SliceCodec { + sliceOpt := bsonoptions.MergeSliceCodecOptions(opts...) + + codec := SliceCodec{} + if sliceOpt.EncodeNilAsEmpty != nil { + codec.EncodeNilAsEmpty = *sliceOpt.EncodeNilAsEmpty + } + return &codec +} + +// EncodeValue is the ValueEncoder for slice types. +func (sc SliceCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Slice { + return ValueEncoderError{Name: "SliceEncodeValue", Kinds: []reflect.Kind{reflect.Slice}, Received: val} + } + + if val.IsNil() && !sc.EncodeNilAsEmpty { + return vw.WriteNull() + } + + // If we have a []byte we want to treat it as a binary instead of as an array. + if val.Type().Elem() == tByte { + var byteSlice []byte + for idx := 0; idx < val.Len(); idx++ { + byteSlice = append(byteSlice, val.Index(idx).Interface().(byte)) + } + return vw.WriteBinary(byteSlice) + } + + // If we have a []primitive.E we want to treat it as a document instead of as an array. + if val.Type().ConvertibleTo(tD) { + d := val.Convert(tD).Interface().(primitive.D) + + dw, err := vw.WriteDocument() + if err != nil { + return err + } + + for _, e := range d { + err = encodeElement(ec, dw, e) + if err != nil { + return err + } + } + + return dw.WriteDocumentEnd() + } + + aw, err := vw.WriteArray() + if err != nil { + return err + } + + elemType := val.Type().Elem() + encoder, err := ec.LookupEncoder(elemType) + if err != nil && elemType.Kind() != reflect.Interface { + return err + } + + for idx := 0; idx < val.Len(); idx++ { + currEncoder, currVal, lookupErr := defaultValueEncoders.lookupElementEncoder(ec, encoder, val.Index(idx)) + if lookupErr != nil && lookupErr != errInvalidValue { + return lookupErr + } + + vw, err := aw.WriteArrayElement() + if err != nil { + return err + } + + if lookupErr == errInvalidValue { + err = vw.WriteNull() + if err != nil { + return err + } + continue + } + + err = currEncoder.EncodeValue(ec, vw, currVal) + if err != nil { + return err + } + } + return aw.WriteArrayEnd() +} + +// DecodeValue is the ValueDecoder for slice types. +func (sc *SliceCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Kind() != reflect.Slice { + return ValueDecoderError{Name: "SliceDecodeValue", Kinds: []reflect.Kind{reflect.Slice}, Received: val} + } + + switch vrType := vr.Type(); vrType { + case bsontype.Array: + case bsontype.Null: + val.Set(reflect.Zero(val.Type())) + return vr.ReadNull() + case bsontype.Undefined: + val.Set(reflect.Zero(val.Type())) + return vr.ReadUndefined() + case bsontype.Type(0), bsontype.EmbeddedDocument: + if val.Type().Elem() != tE { + return fmt.Errorf("cannot decode document into %s", val.Type()) + } + case bsontype.Binary: + if val.Type().Elem() != tByte { + return fmt.Errorf("SliceDecodeValue can only decode a binary into a byte array, got %v", vrType) + } + data, subtype, err := vr.ReadBinary() + if err != nil { + return err + } + if subtype != bsontype.BinaryGeneric && subtype != bsontype.BinaryBinaryOld { + return fmt.Errorf("SliceDecodeValue can only be used to decode subtype 0x00 or 0x02 for %s, got %v", bsontype.Binary, subtype) + } + + if val.IsNil() { + val.Set(reflect.MakeSlice(val.Type(), 0, len(data))) + } + + val.SetLen(0) + for _, elem := range data { + val.Set(reflect.Append(val, reflect.ValueOf(elem))) + } + return nil + case bsontype.String: + if sliceType := val.Type().Elem(); sliceType != tByte { + return fmt.Errorf("SliceDecodeValue can only decode a string into a byte array, got %v", sliceType) + } + str, err := vr.ReadString() + if err != nil { + return err + } + byteStr := []byte(str) + + if val.IsNil() { + val.Set(reflect.MakeSlice(val.Type(), 0, len(byteStr))) + } + + val.SetLen(0) + for _, elem := range byteStr { + val.Set(reflect.Append(val, reflect.ValueOf(elem))) + } + return nil + default: + return fmt.Errorf("cannot decode %v into a slice", vrType) + } + + var elemsFunc func(DecodeContext, bsonrw.ValueReader, reflect.Value) ([]reflect.Value, error) + switch val.Type().Elem() { + case tE: + dc.Ancestor = val.Type() + elemsFunc = defaultValueDecoders.decodeD + default: + elemsFunc = defaultValueDecoders.decodeDefault + } + + elems, err := elemsFunc(dc, vr, val) + if err != nil { + return err + } + + if val.IsNil() { + val.Set(reflect.MakeSlice(val.Type(), 0, len(elems))) + } + + val.SetLen(0) + val.Set(reflect.Append(val, elems...)) + + return nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/string_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/string_codec.go new file mode 100644 index 0000000..5332b7c --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/string_codec.go @@ -0,0 +1,119 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "fmt" + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +// StringCodec is the Codec used for struct values. +type StringCodec struct { + DecodeObjectIDAsHex bool +} + +var ( + defaultStringCodec = NewStringCodec() + + _ ValueCodec = defaultStringCodec + _ typeDecoder = defaultStringCodec +) + +// NewStringCodec returns a StringCodec with options opts. +func NewStringCodec(opts ...*bsonoptions.StringCodecOptions) *StringCodec { + stringOpt := bsonoptions.MergeStringCodecOptions(opts...) + return &StringCodec{*stringOpt.DecodeObjectIDAsHex} +} + +// EncodeValue is the ValueEncoder for string types. +func (sc *StringCodec) EncodeValue(ectx EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if val.Kind() != reflect.String { + return ValueEncoderError{ + Name: "StringEncodeValue", + Kinds: []reflect.Kind{reflect.String}, + Received: val, + } + } + + return vw.WriteString(val.String()) +} + +func (sc *StringCodec) decodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t.Kind() != reflect.String { + return emptyValue, ValueDecoderError{ + Name: "StringDecodeValue", + Kinds: []reflect.Kind{reflect.String}, + Received: reflect.Zero(t), + } + } + + var str string + var err error + switch vr.Type() { + case bsontype.String: + str, err = vr.ReadString() + if err != nil { + return emptyValue, err + } + case bsontype.ObjectID: + oid, err := vr.ReadObjectID() + if err != nil { + return emptyValue, err + } + if sc.DecodeObjectIDAsHex { + str = oid.Hex() + } else { + byteArray := [12]byte(oid) + str = string(byteArray[:]) + } + case bsontype.Symbol: + str, err = vr.ReadSymbol() + if err != nil { + return emptyValue, err + } + case bsontype.Binary: + data, subtype, err := vr.ReadBinary() + if err != nil { + return emptyValue, err + } + if subtype != bsontype.BinaryGeneric && subtype != bsontype.BinaryBinaryOld { + return emptyValue, decodeBinaryError{subtype: subtype, typeName: "string"} + } + str = string(data) + case bsontype.Null: + if err = vr.ReadNull(); err != nil { + return emptyValue, err + } + case bsontype.Undefined: + if err = vr.ReadUndefined(); err != nil { + return emptyValue, err + } + default: + return emptyValue, fmt.Errorf("cannot decode %v into a string type", vr.Type()) + } + + return reflect.ValueOf(str), nil +} + +// DecodeValue is the ValueDecoder for string types. +func (sc *StringCodec) DecodeValue(dctx DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Kind() != reflect.String { + return ValueDecoderError{Name: "StringDecodeValue", Kinds: []reflect.Kind{reflect.String}, Received: val} + } + + elem, err := sc.decodeType(dctx, vr, val.Type()) + if err != nil { + return err + } + + val.SetString(elem.String()) + return nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/struct_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/struct_codec.go new file mode 100644 index 0000000..be3f208 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/struct_codec.go @@ -0,0 +1,664 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "errors" + "fmt" + "reflect" + "sort" + "strings" + "sync" + "time" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +// DecodeError represents an error that occurs when unmarshalling BSON bytes into a native Go type. +type DecodeError struct { + keys []string + wrapped error +} + +// Unwrap returns the underlying error +func (de *DecodeError) Unwrap() error { + return de.wrapped +} + +// Error implements the error interface. +func (de *DecodeError) Error() string { + // The keys are stored in reverse order because the de.keys slice is builtup while propagating the error up the + // stack of BSON keys, so we call de.Keys(), which reverses them. + keyPath := strings.Join(de.Keys(), ".") + return fmt.Sprintf("error decoding key %s: %v", keyPath, de.wrapped) +} + +// Keys returns the BSON key path that caused an error as a slice of strings. The keys in the slice are in top-down +// order. For example, if the document being unmarshalled was {a: {b: {c: 1}}} and the value for c was supposed to be +// a string, the keys slice will be ["a", "b", "c"]. +func (de *DecodeError) Keys() []string { + reversedKeys := make([]string, 0, len(de.keys)) + for idx := len(de.keys) - 1; idx >= 0; idx-- { + reversedKeys = append(reversedKeys, de.keys[idx]) + } + + return reversedKeys +} + +// Zeroer allows custom struct types to implement a report of zero +// state. All struct types that don't implement Zeroer or where IsZero +// returns false are considered to be not zero. +type Zeroer interface { + IsZero() bool +} + +// StructCodec is the Codec used for struct values. +type StructCodec struct { + cache map[reflect.Type]*structDescription + l sync.RWMutex + parser StructTagParser + DecodeZeroStruct bool + DecodeDeepZeroInline bool + EncodeOmitDefaultStruct bool + AllowUnexportedFields bool + OverwriteDuplicatedInlinedFields bool +} + +var _ ValueEncoder = &StructCodec{} +var _ ValueDecoder = &StructCodec{} + +// NewStructCodec returns a StructCodec that uses p for struct tag parsing. +func NewStructCodec(p StructTagParser, opts ...*bsonoptions.StructCodecOptions) (*StructCodec, error) { + if p == nil { + return nil, errors.New("a StructTagParser must be provided to NewStructCodec") + } + + structOpt := bsonoptions.MergeStructCodecOptions(opts...) + + codec := &StructCodec{ + cache: make(map[reflect.Type]*structDescription), + parser: p, + } + + if structOpt.DecodeZeroStruct != nil { + codec.DecodeZeroStruct = *structOpt.DecodeZeroStruct + } + if structOpt.DecodeDeepZeroInline != nil { + codec.DecodeDeepZeroInline = *structOpt.DecodeDeepZeroInline + } + if structOpt.EncodeOmitDefaultStruct != nil { + codec.EncodeOmitDefaultStruct = *structOpt.EncodeOmitDefaultStruct + } + if structOpt.OverwriteDuplicatedInlinedFields != nil { + codec.OverwriteDuplicatedInlinedFields = *structOpt.OverwriteDuplicatedInlinedFields + } + if structOpt.AllowUnexportedFields != nil { + codec.AllowUnexportedFields = *structOpt.AllowUnexportedFields + } + + return codec, nil +} + +// EncodeValue handles encoding generic struct types. +func (sc *StructCodec) EncodeValue(r EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Kind() != reflect.Struct { + return ValueEncoderError{Name: "StructCodec.EncodeValue", Kinds: []reflect.Kind{reflect.Struct}, Received: val} + } + + sd, err := sc.describeStruct(r.Registry, val.Type()) + if err != nil { + return err + } + + dw, err := vw.WriteDocument() + if err != nil { + return err + } + var rv reflect.Value + for _, desc := range sd.fl { + if desc.inline == nil { + rv = val.Field(desc.idx) + } else { + rv, err = fieldByIndexErr(val, desc.inline) + if err != nil { + continue + } + } + + desc.encoder, rv, err = defaultValueEncoders.lookupElementEncoder(r, desc.encoder, rv) + + if err != nil && err != errInvalidValue { + return err + } + + if err == errInvalidValue { + if desc.omitEmpty { + continue + } + vw2, err := dw.WriteDocumentElement(desc.name) + if err != nil { + return err + } + err = vw2.WriteNull() + if err != nil { + return err + } + continue + } + + if desc.encoder == nil { + return ErrNoEncoder{Type: rv.Type()} + } + + encoder := desc.encoder + + var isZero bool + rvInterface := rv.Interface() + if cz, ok := encoder.(CodecZeroer); ok { + isZero = cz.IsTypeZero(rvInterface) + } else if rv.Kind() == reflect.Interface { + // sc.isZero will not treat an interface rv as an interface, so we need to check for the zero interface separately. + isZero = rv.IsNil() + } else { + isZero = sc.isZero(rvInterface) + } + if desc.omitEmpty && isZero { + continue + } + + vw2, err := dw.WriteDocumentElement(desc.name) + if err != nil { + return err + } + + ectx := EncodeContext{Registry: r.Registry, MinSize: desc.minSize} + err = encoder.EncodeValue(ectx, vw2, rv) + if err != nil { + return err + } + } + + if sd.inlineMap >= 0 { + rv := val.Field(sd.inlineMap) + collisionFn := func(key string) bool { + _, exists := sd.fm[key] + return exists + } + + return defaultMapCodec.mapEncodeValue(r, dw, rv, collisionFn) + } + + return dw.WriteDocumentEnd() +} + +func newDecodeError(key string, original error) error { + de, ok := original.(*DecodeError) + if !ok { + return &DecodeError{ + keys: []string{key}, + wrapped: original, + } + } + + de.keys = append(de.keys, key) + return de +} + +// DecodeValue implements the Codec interface. +// By default, map types in val will not be cleared. If a map has existing key/value pairs, it will be extended with the new ones from vr. +// For slices, the decoder will set the length of the slice to zero and append all elements. The underlying array will not be cleared. +func (sc *StructCodec) DecodeValue(r DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Kind() != reflect.Struct { + return ValueDecoderError{Name: "StructCodec.DecodeValue", Kinds: []reflect.Kind{reflect.Struct}, Received: val} + } + + switch vrType := vr.Type(); vrType { + case bsontype.Type(0), bsontype.EmbeddedDocument: + case bsontype.Null: + if err := vr.ReadNull(); err != nil { + return err + } + + val.Set(reflect.Zero(val.Type())) + return nil + case bsontype.Undefined: + if err := vr.ReadUndefined(); err != nil { + return err + } + + val.Set(reflect.Zero(val.Type())) + return nil + default: + return fmt.Errorf("cannot decode %v into a %s", vrType, val.Type()) + } + + sd, err := sc.describeStruct(r.Registry, val.Type()) + if err != nil { + return err + } + + if sc.DecodeZeroStruct { + val.Set(reflect.Zero(val.Type())) + } + if sc.DecodeDeepZeroInline && sd.inline { + val.Set(deepZero(val.Type())) + } + + var decoder ValueDecoder + var inlineMap reflect.Value + if sd.inlineMap >= 0 { + inlineMap = val.Field(sd.inlineMap) + decoder, err = r.LookupDecoder(inlineMap.Type().Elem()) + if err != nil { + return err + } + } + + dr, err := vr.ReadDocument() + if err != nil { + return err + } + + for { + name, vr, err := dr.ReadElement() + if err == bsonrw.ErrEOD { + break + } + if err != nil { + return err + } + + fd, exists := sd.fm[name] + if !exists { + // if the original name isn't found in the struct description, try again with the name in lowercase + // this could match if a BSON tag isn't specified because by default, describeStruct lowercases all field + // names + fd, exists = sd.fm[strings.ToLower(name)] + } + + if !exists { + if sd.inlineMap < 0 { + // The encoding/json package requires a flag to return on error for non-existent fields. + // This functionality seems appropriate for the struct codec. + err = vr.Skip() + if err != nil { + return err + } + continue + } + + if inlineMap.IsNil() { + inlineMap.Set(reflect.MakeMap(inlineMap.Type())) + } + + elem := reflect.New(inlineMap.Type().Elem()).Elem() + r.Ancestor = inlineMap.Type() + err = decoder.DecodeValue(r, vr, elem) + if err != nil { + return err + } + inlineMap.SetMapIndex(reflect.ValueOf(name), elem) + continue + } + + var field reflect.Value + if fd.inline == nil { + field = val.Field(fd.idx) + } else { + field, err = getInlineField(val, fd.inline) + if err != nil { + return err + } + } + + if !field.CanSet() { // Being settable is a super set of being addressable. + innerErr := fmt.Errorf("field %v is not settable", field) + return newDecodeError(fd.name, innerErr) + } + if field.Kind() == reflect.Ptr && field.IsNil() { + field.Set(reflect.New(field.Type().Elem())) + } + field = field.Addr() + + dctx := DecodeContext{Registry: r.Registry, Truncate: fd.truncate || r.Truncate} + if fd.decoder == nil { + return newDecodeError(fd.name, ErrNoDecoder{Type: field.Elem().Type()}) + } + + err = fd.decoder.DecodeValue(dctx, vr, field.Elem()) + if err != nil { + return newDecodeError(fd.name, err) + } + } + + return nil +} + +func (sc *StructCodec) isZero(i interface{}) bool { + v := reflect.ValueOf(i) + + // check the value validity + if !v.IsValid() { + return true + } + + if z, ok := v.Interface().(Zeroer); ok && (v.Kind() != reflect.Ptr || !v.IsNil()) { + return z.IsZero() + } + + switch v.Kind() { + case reflect.Array, reflect.Map, reflect.Slice, reflect.String: + return v.Len() == 0 + case reflect.Bool: + return !v.Bool() + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64: + return v.Int() == 0 + case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr: + return v.Uint() == 0 + case reflect.Float32, reflect.Float64: + return v.Float() == 0 + case reflect.Interface, reflect.Ptr: + return v.IsNil() + case reflect.Struct: + if sc.EncodeOmitDefaultStruct { + vt := v.Type() + if vt == tTime { + return v.Interface().(time.Time).IsZero() + } + for i := 0; i < v.NumField(); i++ { + if vt.Field(i).PkgPath != "" && !vt.Field(i).Anonymous { + continue // Private field + } + fld := v.Field(i) + if !sc.isZero(fld.Interface()) { + return false + } + } + return true + } + } + + return false +} + +type structDescription struct { + fm map[string]fieldDescription + fl []fieldDescription + inlineMap int + inline bool +} + +type fieldDescription struct { + name string // BSON key name + fieldName string // struct field name + idx int + omitEmpty bool + minSize bool + truncate bool + inline []int + encoder ValueEncoder + decoder ValueDecoder +} + +type byIndex []fieldDescription + +func (bi byIndex) Len() int { return len(bi) } + +func (bi byIndex) Swap(i, j int) { bi[i], bi[j] = bi[j], bi[i] } + +func (bi byIndex) Less(i, j int) bool { + // If a field is inlined, its index in the top level struct is stored at inline[0] + iIdx, jIdx := bi[i].idx, bi[j].idx + if len(bi[i].inline) > 0 { + iIdx = bi[i].inline[0] + } + if len(bi[j].inline) > 0 { + jIdx = bi[j].inline[0] + } + if iIdx != jIdx { + return iIdx < jIdx + } + for k, biik := range bi[i].inline { + if k >= len(bi[j].inline) { + return false + } + if biik != bi[j].inline[k] { + return biik < bi[j].inline[k] + } + } + return len(bi[i].inline) < len(bi[j].inline) +} + +func (sc *StructCodec) describeStruct(r *Registry, t reflect.Type) (*structDescription, error) { + // We need to analyze the struct, including getting the tags, collecting + // information about inlining, and create a map of the field name to the field. + sc.l.RLock() + ds, exists := sc.cache[t] + sc.l.RUnlock() + if exists { + return ds, nil + } + + numFields := t.NumField() + sd := &structDescription{ + fm: make(map[string]fieldDescription, numFields), + fl: make([]fieldDescription, 0, numFields), + inlineMap: -1, + } + + var fields []fieldDescription + for i := 0; i < numFields; i++ { + sf := t.Field(i) + if sf.PkgPath != "" && (!sc.AllowUnexportedFields || !sf.Anonymous) { + // field is private or unexported fields aren't allowed, ignore + continue + } + + sfType := sf.Type + encoder, err := r.LookupEncoder(sfType) + if err != nil { + encoder = nil + } + decoder, err := r.LookupDecoder(sfType) + if err != nil { + decoder = nil + } + + description := fieldDescription{ + fieldName: sf.Name, + idx: i, + encoder: encoder, + decoder: decoder, + } + + stags, err := sc.parser.ParseStructTags(sf) + if err != nil { + return nil, err + } + if stags.Skip { + continue + } + description.name = stags.Name + description.omitEmpty = stags.OmitEmpty + description.minSize = stags.MinSize + description.truncate = stags.Truncate + + if stags.Inline { + sd.inline = true + switch sfType.Kind() { + case reflect.Map: + if sd.inlineMap >= 0 { + return nil, errors.New("(struct " + t.String() + ") multiple inline maps") + } + if sfType.Key() != tString { + return nil, errors.New("(struct " + t.String() + ") inline map must have a string keys") + } + sd.inlineMap = description.idx + case reflect.Ptr: + sfType = sfType.Elem() + if sfType.Kind() != reflect.Struct { + return nil, fmt.Errorf("(struct %s) inline fields must be a struct, a struct pointer, or a map", t.String()) + } + fallthrough + case reflect.Struct: + inlinesf, err := sc.describeStruct(r, sfType) + if err != nil { + return nil, err + } + for _, fd := range inlinesf.fl { + if fd.inline == nil { + fd.inline = []int{i, fd.idx} + } else { + fd.inline = append([]int{i}, fd.inline...) + } + fields = append(fields, fd) + + } + default: + return nil, fmt.Errorf("(struct %s) inline fields must be a struct, a struct pointer, or a map", t.String()) + } + continue + } + fields = append(fields, description) + } + + // Sort fieldDescriptions by name and use dominance rules to determine which should be added for each name + sort.Slice(fields, func(i, j int) bool { + x := fields + // sort field by name, breaking ties with depth, then + // breaking ties with index sequence. + if x[i].name != x[j].name { + return x[i].name < x[j].name + } + if len(x[i].inline) != len(x[j].inline) { + return len(x[i].inline) < len(x[j].inline) + } + return byIndex(x).Less(i, j) + }) + + for advance, i := 0, 0; i < len(fields); i += advance { + // One iteration per name. + // Find the sequence of fields with the name of this first field. + fi := fields[i] + name := fi.name + for advance = 1; i+advance < len(fields); advance++ { + fj := fields[i+advance] + if fj.name != name { + break + } + } + if advance == 1 { // Only one field with this name + sd.fl = append(sd.fl, fi) + sd.fm[name] = fi + continue + } + dominant, ok := dominantField(fields[i : i+advance]) + if !ok || !sc.OverwriteDuplicatedInlinedFields { + return nil, fmt.Errorf("struct %s has duplicated key %s", t.String(), name) + } + sd.fl = append(sd.fl, dominant) + sd.fm[name] = dominant + } + + sort.Sort(byIndex(sd.fl)) + + sc.l.Lock() + sc.cache[t] = sd + sc.l.Unlock() + + return sd, nil +} + +// dominantField looks through the fields, all of which are known to +// have the same name, to find the single field that dominates the +// others using Go's inlining rules. If there are multiple top-level +// fields, the boolean will be false: This condition is an error in Go +// and we skip all the fields. +func dominantField(fields []fieldDescription) (fieldDescription, bool) { + // The fields are sorted in increasing index-length order, then by presence of tag. + // That means that the first field is the dominant one. We need only check + // for error cases: two fields at top level. + if len(fields) > 1 && + len(fields[0].inline) == len(fields[1].inline) { + return fieldDescription{}, false + } + return fields[0], true +} + +func fieldByIndexErr(v reflect.Value, index []int) (result reflect.Value, err error) { + defer func() { + if recovered := recover(); recovered != nil { + switch r := recovered.(type) { + case string: + err = fmt.Errorf("%s", r) + case error: + err = r + } + } + }() + + result = v.FieldByIndex(index) + return +} + +func getInlineField(val reflect.Value, index []int) (reflect.Value, error) { + field, err := fieldByIndexErr(val, index) + if err == nil { + return field, nil + } + + // if parent of this element doesn't exist, fix its parent + inlineParent := index[:len(index)-1] + var fParent reflect.Value + if fParent, err = fieldByIndexErr(val, inlineParent); err != nil { + fParent, err = getInlineField(val, inlineParent) + if err != nil { + return fParent, err + } + } + fParent.Set(reflect.New(fParent.Type().Elem())) + + return fieldByIndexErr(val, index) +} + +// DeepZero returns recursive zero object +func deepZero(st reflect.Type) (result reflect.Value) { + result = reflect.Indirect(reflect.New(st)) + + if result.Kind() == reflect.Struct { + for i := 0; i < result.NumField(); i++ { + if f := result.Field(i); f.Kind() == reflect.Ptr { + if f.CanInterface() { + if ft := reflect.TypeOf(f.Interface()); ft.Elem().Kind() == reflect.Struct { + result.Field(i).Set(recursivePointerTo(deepZero(ft.Elem()))) + } + } + } + } + } + + return +} + +// recursivePointerTo calls reflect.New(v.Type) but recursively for its fields inside +func recursivePointerTo(v reflect.Value) reflect.Value { + v = reflect.Indirect(v) + result := reflect.New(v.Type()) + if v.Kind() == reflect.Struct { + for i := 0; i < v.NumField(); i++ { + if f := v.Field(i); f.Kind() == reflect.Ptr { + if f.Elem().Kind() == reflect.Struct { + result.Elem().Field(i).Set(recursivePointerTo(f)) + } + } + } + } + + return result +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/struct_tag_parser.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/struct_tag_parser.go new file mode 100644 index 0000000..6f406c1 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/struct_tag_parser.go @@ -0,0 +1,139 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "reflect" + "strings" +) + +// StructTagParser returns the struct tags for a given struct field. +type StructTagParser interface { + ParseStructTags(reflect.StructField) (StructTags, error) +} + +// StructTagParserFunc is an adapter that allows a generic function to be used +// as a StructTagParser. +type StructTagParserFunc func(reflect.StructField) (StructTags, error) + +// ParseStructTags implements the StructTagParser interface. +func (stpf StructTagParserFunc) ParseStructTags(sf reflect.StructField) (StructTags, error) { + return stpf(sf) +} + +// StructTags represents the struct tag fields that the StructCodec uses during +// the encoding and decoding process. +// +// In the case of a struct, the lowercased field name is used as the key for each exported +// field but this behavior may be changed using a struct tag. The tag may also contain flags to +// adjust the marshalling behavior for the field. +// +// The properties are defined below: +// +// OmitEmpty Only include the field if it's not set to the zero value for the type or to +// empty slices or maps. +// +// MinSize Marshal an integer of a type larger than 32 bits value as an int32, if that's +// feasible while preserving the numeric value. +// +// Truncate When unmarshaling a BSON double, it is permitted to lose precision to fit within +// a float32. +// +// Inline Inline the field, which must be a struct or a map, causing all of its fields +// or keys to be processed as if they were part of the outer struct. For maps, +// keys must not conflict with the bson keys of other struct fields. +// +// Skip This struct field should be skipped. This is usually denoted by parsing a "-" +// for the name. +// +// TODO(skriptble): Add tags for undefined as nil and for null as nil. +type StructTags struct { + Name string + OmitEmpty bool + MinSize bool + Truncate bool + Inline bool + Skip bool +} + +// DefaultStructTagParser is the StructTagParser used by the StructCodec by default. +// It will handle the bson struct tag. See the documentation for StructTags to see +// what each of the returned fields means. +// +// If there is no name in the struct tag fields, the struct field name is lowercased. +// The tag formats accepted are: +// +// "[][,[,]]" +// +// `(...) bson:"[][,[,]]" (...)` +// +// An example: +// +// type T struct { +// A bool +// B int "myb" +// C string "myc,omitempty" +// D string `bson:",omitempty" json:"jsonkey"` +// E int64 ",minsize" +// F int64 "myf,omitempty,minsize" +// } +// +// A struct tag either consisting entirely of '-' or with a bson key with a +// value consisting entirely of '-' will return a StructTags with Skip true and +// the remaining fields will be their default values. +var DefaultStructTagParser StructTagParserFunc = func(sf reflect.StructField) (StructTags, error) { + key := strings.ToLower(sf.Name) + tag, ok := sf.Tag.Lookup("bson") + if !ok && !strings.Contains(string(sf.Tag), ":") && len(sf.Tag) > 0 { + tag = string(sf.Tag) + } + return parseTags(key, tag) +} + +func parseTags(key string, tag string) (StructTags, error) { + var st StructTags + if tag == "-" { + st.Skip = true + return st, nil + } + + for idx, str := range strings.Split(tag, ",") { + if idx == 0 && str != "" { + key = str + } + switch str { + case "omitempty": + st.OmitEmpty = true + case "minsize": + st.MinSize = true + case "truncate": + st.Truncate = true + case "inline": + st.Inline = true + } + } + + st.Name = key + + return st, nil +} + +// JSONFallbackStructTagParser has the same behavior as DefaultStructTagParser +// but will also fallback to parsing the json tag instead on a field where the +// bson tag isn't available. +var JSONFallbackStructTagParser StructTagParserFunc = func(sf reflect.StructField) (StructTags, error) { + key := strings.ToLower(sf.Name) + tag, ok := sf.Tag.Lookup("bson") + if !ok { + tag, ok = sf.Tag.Lookup("json") + } + if !ok && !strings.Contains(string(sf.Tag), ":") && len(sf.Tag) > 0 { + tag = string(sf.Tag) + } + + return parseTags(key, tag) +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/time_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/time_codec.go new file mode 100644 index 0000000..ec7e30f --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/time_codec.go @@ -0,0 +1,127 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "fmt" + "reflect" + "time" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" +) + +const ( + timeFormatString = "2006-01-02T15:04:05.999Z07:00" +) + +// TimeCodec is the Codec used for time.Time values. +type TimeCodec struct { + UseLocalTimeZone bool +} + +var ( + defaultTimeCodec = NewTimeCodec() + + _ ValueCodec = defaultTimeCodec + _ typeDecoder = defaultTimeCodec +) + +// NewTimeCodec returns a TimeCodec with options opts. +func NewTimeCodec(opts ...*bsonoptions.TimeCodecOptions) *TimeCodec { + timeOpt := bsonoptions.MergeTimeCodecOptions(opts...) + + codec := TimeCodec{} + if timeOpt.UseLocalTimeZone != nil { + codec.UseLocalTimeZone = *timeOpt.UseLocalTimeZone + } + return &codec +} + +func (tc *TimeCodec) decodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + if t != tTime { + return emptyValue, ValueDecoderError{ + Name: "TimeDecodeValue", + Types: []reflect.Type{tTime}, + Received: reflect.Zero(t), + } + } + + var timeVal time.Time + switch vrType := vr.Type(); vrType { + case bsontype.DateTime: + dt, err := vr.ReadDateTime() + if err != nil { + return emptyValue, err + } + timeVal = time.Unix(dt/1000, dt%1000*1000000) + case bsontype.String: + // assume strings are in the isoTimeFormat + timeStr, err := vr.ReadString() + if err != nil { + return emptyValue, err + } + timeVal, err = time.Parse(timeFormatString, timeStr) + if err != nil { + return emptyValue, err + } + case bsontype.Int64: + i64, err := vr.ReadInt64() + if err != nil { + return emptyValue, err + } + timeVal = time.Unix(i64/1000, i64%1000*1000000) + case bsontype.Timestamp: + t, _, err := vr.ReadTimestamp() + if err != nil { + return emptyValue, err + } + timeVal = time.Unix(int64(t), 0) + case bsontype.Null: + if err := vr.ReadNull(); err != nil { + return emptyValue, err + } + case bsontype.Undefined: + if err := vr.ReadUndefined(); err != nil { + return emptyValue, err + } + default: + return emptyValue, fmt.Errorf("cannot decode %v into a time.Time", vrType) + } + + if !tc.UseLocalTimeZone { + timeVal = timeVal.UTC() + } + return reflect.ValueOf(timeVal), nil +} + +// DecodeValue is the ValueDecoderFunc for time.Time. +func (tc *TimeCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() || val.Type() != tTime { + return ValueDecoderError{Name: "TimeDecodeValue", Types: []reflect.Type{tTime}, Received: val} + } + + elem, err := tc.decodeType(dc, vr, tTime) + if err != nil { + return err + } + + val.Set(elem) + return nil +} + +// EncodeValue is the ValueEncoderFunc for time.TIme. +func (tc *TimeCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + if !val.IsValid() || val.Type() != tTime { + return ValueEncoderError{Name: "TimeEncodeValue", Types: []reflect.Type{tTime}, Received: val} + } + tt := val.Interface().(time.Time) + dt := primitive.NewDateTimeFromTime(tt) + return vw.WriteDateTime(int64(dt)) +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/types.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/types.go new file mode 100644 index 0000000..07f4b70 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/types.go @@ -0,0 +1,57 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "encoding/json" + "net/url" + "reflect" + "time" + + "go.mongodb.org/mongo-driver/bson/primitive" + "go.mongodb.org/mongo-driver/x/bsonx/bsoncore" +) + +var tBool = reflect.TypeOf(false) +var tFloat64 = reflect.TypeOf(float64(0)) +var tInt32 = reflect.TypeOf(int32(0)) +var tInt64 = reflect.TypeOf(int64(0)) +var tString = reflect.TypeOf("") +var tTime = reflect.TypeOf(time.Time{}) + +var tEmpty = reflect.TypeOf((*interface{})(nil)).Elem() +var tByteSlice = reflect.TypeOf([]byte(nil)) +var tByte = reflect.TypeOf(byte(0x00)) +var tURL = reflect.TypeOf(url.URL{}) +var tJSONNumber = reflect.TypeOf(json.Number("")) + +var tValueMarshaler = reflect.TypeOf((*ValueMarshaler)(nil)).Elem() +var tValueUnmarshaler = reflect.TypeOf((*ValueUnmarshaler)(nil)).Elem() +var tMarshaler = reflect.TypeOf((*Marshaler)(nil)).Elem() +var tUnmarshaler = reflect.TypeOf((*Unmarshaler)(nil)).Elem() +var tProxy = reflect.TypeOf((*Proxy)(nil)).Elem() + +var tBinary = reflect.TypeOf(primitive.Binary{}) +var tUndefined = reflect.TypeOf(primitive.Undefined{}) +var tOID = reflect.TypeOf(primitive.ObjectID{}) +var tDateTime = reflect.TypeOf(primitive.DateTime(0)) +var tNull = reflect.TypeOf(primitive.Null{}) +var tRegex = reflect.TypeOf(primitive.Regex{}) +var tCodeWithScope = reflect.TypeOf(primitive.CodeWithScope{}) +var tDBPointer = reflect.TypeOf(primitive.DBPointer{}) +var tJavaScript = reflect.TypeOf(primitive.JavaScript("")) +var tSymbol = reflect.TypeOf(primitive.Symbol("")) +var tTimestamp = reflect.TypeOf(primitive.Timestamp{}) +var tDecimal = reflect.TypeOf(primitive.Decimal128{}) +var tMinKey = reflect.TypeOf(primitive.MinKey{}) +var tMaxKey = reflect.TypeOf(primitive.MaxKey{}) +var tD = reflect.TypeOf(primitive.D{}) +var tA = reflect.TypeOf(primitive.A{}) +var tE = reflect.TypeOf(primitive.E{}) + +var tCoreDocument = reflect.TypeOf(bsoncore.Document{}) +var tCoreArray = reflect.TypeOf(bsoncore.Array{}) diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/uint_codec.go b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/uint_codec.go new file mode 100644 index 0000000..0b21ce9 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsoncodec/uint_codec.go @@ -0,0 +1,173 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsoncodec + +import ( + "fmt" + "math" + "reflect" + + "go.mongodb.org/mongo-driver/bson/bsonoptions" + "go.mongodb.org/mongo-driver/bson/bsonrw" + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +// UIntCodec is the Codec used for uint values. +type UIntCodec struct { + EncodeToMinSize bool +} + +var ( + defaultUIntCodec = NewUIntCodec() + + _ ValueCodec = defaultUIntCodec + _ typeDecoder = defaultUIntCodec +) + +// NewUIntCodec returns a UIntCodec with options opts. +func NewUIntCodec(opts ...*bsonoptions.UIntCodecOptions) *UIntCodec { + uintOpt := bsonoptions.MergeUIntCodecOptions(opts...) + + codec := UIntCodec{} + if uintOpt.EncodeToMinSize != nil { + codec.EncodeToMinSize = *uintOpt.EncodeToMinSize + } + return &codec +} + +// EncodeValue is the ValueEncoder for uint types. +func (uic *UIntCodec) EncodeValue(ec EncodeContext, vw bsonrw.ValueWriter, val reflect.Value) error { + switch val.Kind() { + case reflect.Uint8, reflect.Uint16: + return vw.WriteInt32(int32(val.Uint())) + case reflect.Uint, reflect.Uint32, reflect.Uint64: + u64 := val.Uint() + + // If ec.MinSize or if encodeToMinSize is true for a non-uint64 value we should write val as an int32 + useMinSize := ec.MinSize || (uic.EncodeToMinSize && val.Kind() != reflect.Uint64) + + if u64 <= math.MaxInt32 && useMinSize { + return vw.WriteInt32(int32(u64)) + } + if u64 > math.MaxInt64 { + return fmt.Errorf("%d overflows int64", u64) + } + return vw.WriteInt64(int64(u64)) + } + + return ValueEncoderError{ + Name: "UintEncodeValue", + Kinds: []reflect.Kind{reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uint}, + Received: val, + } +} + +func (uic *UIntCodec) decodeType(dc DecodeContext, vr bsonrw.ValueReader, t reflect.Type) (reflect.Value, error) { + var i64 int64 + var err error + switch vrType := vr.Type(); vrType { + case bsontype.Int32: + i32, err := vr.ReadInt32() + if err != nil { + return emptyValue, err + } + i64 = int64(i32) + case bsontype.Int64: + i64, err = vr.ReadInt64() + if err != nil { + return emptyValue, err + } + case bsontype.Double: + f64, err := vr.ReadDouble() + if err != nil { + return emptyValue, err + } + if !dc.Truncate && math.Floor(f64) != f64 { + return emptyValue, errCannotTruncate + } + if f64 > float64(math.MaxInt64) { + return emptyValue, fmt.Errorf("%g overflows int64", f64) + } + i64 = int64(f64) + case bsontype.Boolean: + b, err := vr.ReadBoolean() + if err != nil { + return emptyValue, err + } + if b { + i64 = 1 + } + case bsontype.Null: + if err = vr.ReadNull(); err != nil { + return emptyValue, err + } + case bsontype.Undefined: + if err = vr.ReadUndefined(); err != nil { + return emptyValue, err + } + default: + return emptyValue, fmt.Errorf("cannot decode %v into an integer type", vrType) + } + + switch t.Kind() { + case reflect.Uint8: + if i64 < 0 || i64 > math.MaxUint8 { + return emptyValue, fmt.Errorf("%d overflows uint8", i64) + } + + return reflect.ValueOf(uint8(i64)), nil + case reflect.Uint16: + if i64 < 0 || i64 > math.MaxUint16 { + return emptyValue, fmt.Errorf("%d overflows uint16", i64) + } + + return reflect.ValueOf(uint16(i64)), nil + case reflect.Uint32: + if i64 < 0 || i64 > math.MaxUint32 { + return emptyValue, fmt.Errorf("%d overflows uint32", i64) + } + + return reflect.ValueOf(uint32(i64)), nil + case reflect.Uint64: + if i64 < 0 { + return emptyValue, fmt.Errorf("%d overflows uint64", i64) + } + + return reflect.ValueOf(uint64(i64)), nil + case reflect.Uint: + if i64 < 0 || int64(uint(i64)) != i64 { // Can we fit this inside of an uint + return emptyValue, fmt.Errorf("%d overflows uint", i64) + } + + return reflect.ValueOf(uint(i64)), nil + default: + return emptyValue, ValueDecoderError{ + Name: "UintDecodeValue", + Kinds: []reflect.Kind{reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uint}, + Received: reflect.Zero(t), + } + } +} + +// DecodeValue is the ValueDecoder for uint types. +func (uic *UIntCodec) DecodeValue(dc DecodeContext, vr bsonrw.ValueReader, val reflect.Value) error { + if !val.CanSet() { + return ValueDecoderError{ + Name: "UintDecodeValue", + Kinds: []reflect.Kind{reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uint}, + Received: val, + } + } + + elem, err := uic.decodeType(dc, vr, val.Type()) + if err != nil { + return err + } + + val.SetUint(elem.Uint()) + return nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/byte_slice_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/byte_slice_codec_options.go new file mode 100644 index 0000000..b1256a4 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/byte_slice_codec_options.go @@ -0,0 +1,38 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +// ByteSliceCodecOptions represents all possible options for byte slice encoding and decoding. +type ByteSliceCodecOptions struct { + EncodeNilAsEmpty *bool // Specifies if a nil byte slice should encode as an empty binary instead of null. Defaults to false. +} + +// ByteSliceCodec creates a new *ByteSliceCodecOptions +func ByteSliceCodec() *ByteSliceCodecOptions { + return &ByteSliceCodecOptions{} +} + +// SetEncodeNilAsEmpty specifies if a nil byte slice should encode as an empty binary instead of null. Defaults to false. +func (bs *ByteSliceCodecOptions) SetEncodeNilAsEmpty(b bool) *ByteSliceCodecOptions { + bs.EncodeNilAsEmpty = &b + return bs +} + +// MergeByteSliceCodecOptions combines the given *ByteSliceCodecOptions into a single *ByteSliceCodecOptions in a last one wins fashion. +func MergeByteSliceCodecOptions(opts ...*ByteSliceCodecOptions) *ByteSliceCodecOptions { + bs := ByteSliceCodec() + for _, opt := range opts { + if opt == nil { + continue + } + if opt.EncodeNilAsEmpty != nil { + bs.EncodeNilAsEmpty = opt.EncodeNilAsEmpty + } + } + + return bs +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/empty_interface_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/empty_interface_codec_options.go new file mode 100644 index 0000000..6caaa00 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/empty_interface_codec_options.go @@ -0,0 +1,38 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +// EmptyInterfaceCodecOptions represents all possible options for interface{} encoding and decoding. +type EmptyInterfaceCodecOptions struct { + DecodeBinaryAsSlice *bool // Specifies if Old and Generic type binarys should default to []slice instead of primitive.Binary. Defaults to false. +} + +// EmptyInterfaceCodec creates a new *EmptyInterfaceCodecOptions +func EmptyInterfaceCodec() *EmptyInterfaceCodecOptions { + return &EmptyInterfaceCodecOptions{} +} + +// SetDecodeBinaryAsSlice specifies if Old and Generic type binarys should default to []slice instead of primitive.Binary. Defaults to false. +func (e *EmptyInterfaceCodecOptions) SetDecodeBinaryAsSlice(b bool) *EmptyInterfaceCodecOptions { + e.DecodeBinaryAsSlice = &b + return e +} + +// MergeEmptyInterfaceCodecOptions combines the given *EmptyInterfaceCodecOptions into a single *EmptyInterfaceCodecOptions in a last one wins fashion. +func MergeEmptyInterfaceCodecOptions(opts ...*EmptyInterfaceCodecOptions) *EmptyInterfaceCodecOptions { + e := EmptyInterfaceCodec() + for _, opt := range opts { + if opt == nil { + continue + } + if opt.DecodeBinaryAsSlice != nil { + e.DecodeBinaryAsSlice = opt.DecodeBinaryAsSlice + } + } + + return e +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/map_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/map_codec_options.go new file mode 100644 index 0000000..7a6a880 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/map_codec_options.go @@ -0,0 +1,67 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +// MapCodecOptions represents all possible options for map encoding and decoding. +type MapCodecOptions struct { + DecodeZerosMap *bool // Specifies if the map should be zeroed before decoding into it. Defaults to false. + EncodeNilAsEmpty *bool // Specifies if a nil map should encode as an empty document instead of null. Defaults to false. + // Specifies how keys should be handled. If false, the behavior matches encoding/json, where the encoding key type must + // either be a string, an integer type, or implement bsoncodec.KeyMarshaler and the decoding key type must either be a + // string, an integer type, or implement bsoncodec.KeyUnmarshaler. If true, keys are encoded with fmt.Sprint() and the + // encoding key type must be a string, an integer type, or a float. If true, the use of Stringer will override + // TextMarshaler/TextUnmarshaler. Defaults to false. + EncodeKeysWithStringer *bool +} + +// MapCodec creates a new *MapCodecOptions +func MapCodec() *MapCodecOptions { + return &MapCodecOptions{} +} + +// SetDecodeZerosMap specifies if the map should be zeroed before decoding into it. Defaults to false. +func (t *MapCodecOptions) SetDecodeZerosMap(b bool) *MapCodecOptions { + t.DecodeZerosMap = &b + return t +} + +// SetEncodeNilAsEmpty specifies if a nil map should encode as an empty document instead of null. Defaults to false. +func (t *MapCodecOptions) SetEncodeNilAsEmpty(b bool) *MapCodecOptions { + t.EncodeNilAsEmpty = &b + return t +} + +// SetEncodeKeysWithStringer specifies how keys should be handled. If false, the behavior matches encoding/json, where the +// encoding key type must either be a string, an integer type, or implement bsoncodec.KeyMarshaler and the decoding key +// type must either be a string, an integer type, or implement bsoncodec.KeyUnmarshaler. If true, keys are encoded with +// fmt.Sprint() and the encoding key type must be a string, an integer type, or a float. If true, the use of Stringer +// will override TextMarshaler/TextUnmarshaler. Defaults to false. +func (t *MapCodecOptions) SetEncodeKeysWithStringer(b bool) *MapCodecOptions { + t.EncodeKeysWithStringer = &b + return t +} + +// MergeMapCodecOptions combines the given *MapCodecOptions into a single *MapCodecOptions in a last one wins fashion. +func MergeMapCodecOptions(opts ...*MapCodecOptions) *MapCodecOptions { + s := MapCodec() + for _, opt := range opts { + if opt == nil { + continue + } + if opt.DecodeZerosMap != nil { + s.DecodeZerosMap = opt.DecodeZerosMap + } + if opt.EncodeNilAsEmpty != nil { + s.EncodeNilAsEmpty = opt.EncodeNilAsEmpty + } + if opt.EncodeKeysWithStringer != nil { + s.EncodeKeysWithStringer = opt.EncodeKeysWithStringer + } + } + + return s +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/slice_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/slice_codec_options.go new file mode 100644 index 0000000..ef965e4 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/slice_codec_options.go @@ -0,0 +1,38 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +// SliceCodecOptions represents all possible options for slice encoding and decoding. +type SliceCodecOptions struct { + EncodeNilAsEmpty *bool // Specifies if a nil slice should encode as an empty array instead of null. Defaults to false. +} + +// SliceCodec creates a new *SliceCodecOptions +func SliceCodec() *SliceCodecOptions { + return &SliceCodecOptions{} +} + +// SetEncodeNilAsEmpty specifies if a nil slice should encode as an empty array instead of null. Defaults to false. +func (s *SliceCodecOptions) SetEncodeNilAsEmpty(b bool) *SliceCodecOptions { + s.EncodeNilAsEmpty = &b + return s +} + +// MergeSliceCodecOptions combines the given *SliceCodecOptions into a single *SliceCodecOptions in a last one wins fashion. +func MergeSliceCodecOptions(opts ...*SliceCodecOptions) *SliceCodecOptions { + s := SliceCodec() + for _, opt := range opts { + if opt == nil { + continue + } + if opt.EncodeNilAsEmpty != nil { + s.EncodeNilAsEmpty = opt.EncodeNilAsEmpty + } + } + + return s +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/string_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/string_codec_options.go new file mode 100644 index 0000000..65964f4 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/string_codec_options.go @@ -0,0 +1,41 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +var defaultDecodeOIDAsHex = true + +// StringCodecOptions represents all possible options for string encoding and decoding. +type StringCodecOptions struct { + DecodeObjectIDAsHex *bool // Specifies if we should decode ObjectID as the hex value. Defaults to true. +} + +// StringCodec creates a new *StringCodecOptions +func StringCodec() *StringCodecOptions { + return &StringCodecOptions{} +} + +// SetDecodeObjectIDAsHex specifies if object IDs should be decoded as their hex representation. If false, a string made +// from the raw object ID bytes will be used. Defaults to true. +func (t *StringCodecOptions) SetDecodeObjectIDAsHex(b bool) *StringCodecOptions { + t.DecodeObjectIDAsHex = &b + return t +} + +// MergeStringCodecOptions combines the given *StringCodecOptions into a single *StringCodecOptions in a last one wins fashion. +func MergeStringCodecOptions(opts ...*StringCodecOptions) *StringCodecOptions { + s := &StringCodecOptions{&defaultDecodeOIDAsHex} + for _, opt := range opts { + if opt == nil { + continue + } + if opt.DecodeObjectIDAsHex != nil { + s.DecodeObjectIDAsHex = opt.DecodeObjectIDAsHex + } + } + + return s +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/struct_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/struct_codec_options.go new file mode 100644 index 0000000..78d1dd8 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/struct_codec_options.go @@ -0,0 +1,87 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +var defaultOverwriteDuplicatedInlinedFields = true + +// StructCodecOptions represents all possible options for struct encoding and decoding. +type StructCodecOptions struct { + DecodeZeroStruct *bool // Specifies if structs should be zeroed before decoding into them. Defaults to false. + DecodeDeepZeroInline *bool // Specifies if structs should be recursively zeroed when a inline value is decoded. Defaults to false. + EncodeOmitDefaultStruct *bool // Specifies if default structs should be considered empty by omitempty. Defaults to false. + AllowUnexportedFields *bool // Specifies if unexported fields should be marshaled/unmarshaled. Defaults to false. + OverwriteDuplicatedInlinedFields *bool // Specifies if fields in inlined structs can be overwritten by higher level struct fields with the same key. Defaults to true. +} + +// StructCodec creates a new *StructCodecOptions +func StructCodec() *StructCodecOptions { + return &StructCodecOptions{} +} + +// SetDecodeZeroStruct specifies if structs should be zeroed before decoding into them. Defaults to false. +func (t *StructCodecOptions) SetDecodeZeroStruct(b bool) *StructCodecOptions { + t.DecodeZeroStruct = &b + return t +} + +// SetDecodeDeepZeroInline specifies if structs should be zeroed before decoding into them. Defaults to false. +func (t *StructCodecOptions) SetDecodeDeepZeroInline(b bool) *StructCodecOptions { + t.DecodeDeepZeroInline = &b + return t +} + +// SetEncodeOmitDefaultStruct specifies if default structs should be considered empty by omitempty. A default struct has all +// its values set to their default value. Defaults to false. +func (t *StructCodecOptions) SetEncodeOmitDefaultStruct(b bool) *StructCodecOptions { + t.EncodeOmitDefaultStruct = &b + return t +} + +// SetOverwriteDuplicatedInlinedFields specifies if inlined struct fields can be overwritten by higher level struct fields with the +// same bson key. When true and decoding, values will be written to the outermost struct with a matching key, and when +// encoding, keys will have the value of the top-most matching field. When false, decoding and encoding will error if +// there are duplicate keys after the struct is inlined. Defaults to true. +func (t *StructCodecOptions) SetOverwriteDuplicatedInlinedFields(b bool) *StructCodecOptions { + t.OverwriteDuplicatedInlinedFields = &b + return t +} + +// SetAllowUnexportedFields specifies if unexported fields should be marshaled/unmarshaled. Defaults to false. +func (t *StructCodecOptions) SetAllowUnexportedFields(b bool) *StructCodecOptions { + t.AllowUnexportedFields = &b + return t +} + +// MergeStructCodecOptions combines the given *StructCodecOptions into a single *StructCodecOptions in a last one wins fashion. +func MergeStructCodecOptions(opts ...*StructCodecOptions) *StructCodecOptions { + s := &StructCodecOptions{ + OverwriteDuplicatedInlinedFields: &defaultOverwriteDuplicatedInlinedFields, + } + for _, opt := range opts { + if opt == nil { + continue + } + + if opt.DecodeZeroStruct != nil { + s.DecodeZeroStruct = opt.DecodeZeroStruct + } + if opt.DecodeDeepZeroInline != nil { + s.DecodeDeepZeroInline = opt.DecodeDeepZeroInline + } + if opt.EncodeOmitDefaultStruct != nil { + s.EncodeOmitDefaultStruct = opt.EncodeOmitDefaultStruct + } + if opt.OverwriteDuplicatedInlinedFields != nil { + s.OverwriteDuplicatedInlinedFields = opt.OverwriteDuplicatedInlinedFields + } + if opt.AllowUnexportedFields != nil { + s.AllowUnexportedFields = opt.AllowUnexportedFields + } + } + + return s +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/time_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/time_codec_options.go new file mode 100644 index 0000000..13496d1 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/time_codec_options.go @@ -0,0 +1,38 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +// TimeCodecOptions represents all possible options for time.Time encoding and decoding. +type TimeCodecOptions struct { + UseLocalTimeZone *bool // Specifies if we should decode into the local time zone. Defaults to false. +} + +// TimeCodec creates a new *TimeCodecOptions +func TimeCodec() *TimeCodecOptions { + return &TimeCodecOptions{} +} + +// SetUseLocalTimeZone specifies if we should decode into the local time zone. Defaults to false. +func (t *TimeCodecOptions) SetUseLocalTimeZone(b bool) *TimeCodecOptions { + t.UseLocalTimeZone = &b + return t +} + +// MergeTimeCodecOptions combines the given *TimeCodecOptions into a single *TimeCodecOptions in a last one wins fashion. +func MergeTimeCodecOptions(opts ...*TimeCodecOptions) *TimeCodecOptions { + t := TimeCodec() + for _, opt := range opts { + if opt == nil { + continue + } + if opt.UseLocalTimeZone != nil { + t.UseLocalTimeZone = opt.UseLocalTimeZone + } + } + + return t +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/uint_codec_options.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/uint_codec_options.go new file mode 100644 index 0000000..e08b7f1 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonoptions/uint_codec_options.go @@ -0,0 +1,38 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonoptions + +// UIntCodecOptions represents all possible options for uint encoding and decoding. +type UIntCodecOptions struct { + EncodeToMinSize *bool // Specifies if all uints except uint64 should be decoded to minimum size bsontype. Defaults to false. +} + +// UIntCodec creates a new *UIntCodecOptions +func UIntCodec() *UIntCodecOptions { + return &UIntCodecOptions{} +} + +// SetEncodeToMinSize specifies if all uints except uint64 should be decoded to minimum size bsontype. Defaults to false. +func (u *UIntCodecOptions) SetEncodeToMinSize(b bool) *UIntCodecOptions { + u.EncodeToMinSize = &b + return u +} + +// MergeUIntCodecOptions combines the given *UIntCodecOptions into a single *UIntCodecOptions in a last one wins fashion. +func MergeUIntCodecOptions(opts ...*UIntCodecOptions) *UIntCodecOptions { + u := UIntCodec() + for _, opt := range opts { + if opt == nil { + continue + } + if opt.EncodeToMinSize != nil { + u.EncodeToMinSize = opt.EncodeToMinSize + } + } + + return u +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/copier.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/copier.go new file mode 100644 index 0000000..5cdf646 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/copier.go @@ -0,0 +1,445 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonrw + +import ( + "fmt" + "io" + + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" + "go.mongodb.org/mongo-driver/x/bsonx/bsoncore" +) + +// Copier is a type that allows copying between ValueReaders, ValueWriters, and +// []byte values. +type Copier struct{} + +// NewCopier creates a new copier with the given registry. If a nil registry is provided +// a default registry is used. +func NewCopier() Copier { + return Copier{} +} + +// CopyDocument handles copying a document from src to dst. +func CopyDocument(dst ValueWriter, src ValueReader) error { + return Copier{}.CopyDocument(dst, src) +} + +// CopyDocument handles copying one document from the src to the dst. +func (c Copier) CopyDocument(dst ValueWriter, src ValueReader) error { + dr, err := src.ReadDocument() + if err != nil { + return err + } + + dw, err := dst.WriteDocument() + if err != nil { + return err + } + + return c.copyDocumentCore(dw, dr) +} + +// CopyArrayFromBytes copies the values from a BSON array represented as a +// []byte to a ValueWriter. +func (c Copier) CopyArrayFromBytes(dst ValueWriter, src []byte) error { + aw, err := dst.WriteArray() + if err != nil { + return err + } + + err = c.CopyBytesToArrayWriter(aw, src) + if err != nil { + return err + } + + return aw.WriteArrayEnd() +} + +// CopyDocumentFromBytes copies the values from a BSON document represented as a +// []byte to a ValueWriter. +func (c Copier) CopyDocumentFromBytes(dst ValueWriter, src []byte) error { + dw, err := dst.WriteDocument() + if err != nil { + return err + } + + err = c.CopyBytesToDocumentWriter(dw, src) + if err != nil { + return err + } + + return dw.WriteDocumentEnd() +} + +type writeElementFn func(key string) (ValueWriter, error) + +// CopyBytesToArrayWriter copies the values from a BSON Array represented as a []byte to an +// ArrayWriter. +func (c Copier) CopyBytesToArrayWriter(dst ArrayWriter, src []byte) error { + wef := func(_ string) (ValueWriter, error) { + return dst.WriteArrayElement() + } + + return c.copyBytesToValueWriter(src, wef) +} + +// CopyBytesToDocumentWriter copies the values from a BSON document represented as a []byte to a +// DocumentWriter. +func (c Copier) CopyBytesToDocumentWriter(dst DocumentWriter, src []byte) error { + wef := func(key string) (ValueWriter, error) { + return dst.WriteDocumentElement(key) + } + + return c.copyBytesToValueWriter(src, wef) +} + +func (c Copier) copyBytesToValueWriter(src []byte, wef writeElementFn) error { + // TODO(skriptble): Create errors types here. Anything thats a tag should be a property. + length, rem, ok := bsoncore.ReadLength(src) + if !ok { + return fmt.Errorf("couldn't read length from src, not enough bytes. length=%d", len(src)) + } + if len(src) < int(length) { + return fmt.Errorf("length read exceeds number of bytes available. length=%d bytes=%d", len(src), length) + } + rem = rem[:length-4] + + var t bsontype.Type + var key string + var val bsoncore.Value + for { + t, rem, ok = bsoncore.ReadType(rem) + if !ok { + return io.EOF + } + if t == bsontype.Type(0) { + if len(rem) != 0 { + return fmt.Errorf("document end byte found before end of document. remaining bytes=%v", rem) + } + break + } + + key, rem, ok = bsoncore.ReadKey(rem) + if !ok { + return fmt.Errorf("invalid key found. remaining bytes=%v", rem) + } + + // write as either array element or document element using writeElementFn + vw, err := wef(key) + if err != nil { + return err + } + + val, rem, ok = bsoncore.ReadValue(rem, t) + if !ok { + return fmt.Errorf("not enough bytes available to read type. bytes=%d type=%s", len(rem), t) + } + err = c.CopyValueFromBytes(vw, t, val.Data) + if err != nil { + return err + } + } + return nil +} + +// CopyDocumentToBytes copies an entire document from the ValueReader and +// returns it as bytes. +func (c Copier) CopyDocumentToBytes(src ValueReader) ([]byte, error) { + return c.AppendDocumentBytes(nil, src) +} + +// AppendDocumentBytes functions the same as CopyDocumentToBytes, but will +// append the result to dst. +func (c Copier) AppendDocumentBytes(dst []byte, src ValueReader) ([]byte, error) { + if br, ok := src.(BytesReader); ok { + _, dst, err := br.ReadValueBytes(dst) + return dst, err + } + + vw := vwPool.Get().(*valueWriter) + defer vwPool.Put(vw) + + vw.reset(dst) + + err := c.CopyDocument(vw, src) + dst = vw.buf + return dst, err +} + +// AppendArrayBytes copies an array from the ValueReader to dst. +func (c Copier) AppendArrayBytes(dst []byte, src ValueReader) ([]byte, error) { + if br, ok := src.(BytesReader); ok { + _, dst, err := br.ReadValueBytes(dst) + return dst, err + } + + vw := vwPool.Get().(*valueWriter) + defer vwPool.Put(vw) + + vw.reset(dst) + + err := c.copyArray(vw, src) + dst = vw.buf + return dst, err +} + +// CopyValueFromBytes will write the value represtend by t and src to dst. +func (c Copier) CopyValueFromBytes(dst ValueWriter, t bsontype.Type, src []byte) error { + if wvb, ok := dst.(BytesWriter); ok { + return wvb.WriteValueBytes(t, src) + } + + vr := vrPool.Get().(*valueReader) + defer vrPool.Put(vr) + + vr.reset(src) + vr.pushElement(t) + + return c.CopyValue(dst, vr) +} + +// CopyValueToBytes copies a value from src and returns it as a bsontype.Type and a +// []byte. +func (c Copier) CopyValueToBytes(src ValueReader) (bsontype.Type, []byte, error) { + return c.AppendValueBytes(nil, src) +} + +// AppendValueBytes functions the same as CopyValueToBytes, but will append the +// result to dst. +func (c Copier) AppendValueBytes(dst []byte, src ValueReader) (bsontype.Type, []byte, error) { + if br, ok := src.(BytesReader); ok { + return br.ReadValueBytes(dst) + } + + vw := vwPool.Get().(*valueWriter) + defer vwPool.Put(vw) + + start := len(dst) + + vw.reset(dst) + vw.push(mElement) + + err := c.CopyValue(vw, src) + if err != nil { + return 0, dst, err + } + + return bsontype.Type(vw.buf[start]), vw.buf[start+2:], nil +} + +// CopyValue will copy a single value from src to dst. +func (c Copier) CopyValue(dst ValueWriter, src ValueReader) error { + var err error + switch src.Type() { + case bsontype.Double: + var f64 float64 + f64, err = src.ReadDouble() + if err != nil { + break + } + err = dst.WriteDouble(f64) + case bsontype.String: + var str string + str, err = src.ReadString() + if err != nil { + return err + } + err = dst.WriteString(str) + case bsontype.EmbeddedDocument: + err = c.CopyDocument(dst, src) + case bsontype.Array: + err = c.copyArray(dst, src) + case bsontype.Binary: + var data []byte + var subtype byte + data, subtype, err = src.ReadBinary() + if err != nil { + break + } + err = dst.WriteBinaryWithSubtype(data, subtype) + case bsontype.Undefined: + err = src.ReadUndefined() + if err != nil { + break + } + err = dst.WriteUndefined() + case bsontype.ObjectID: + var oid primitive.ObjectID + oid, err = src.ReadObjectID() + if err != nil { + break + } + err = dst.WriteObjectID(oid) + case bsontype.Boolean: + var b bool + b, err = src.ReadBoolean() + if err != nil { + break + } + err = dst.WriteBoolean(b) + case bsontype.DateTime: + var dt int64 + dt, err = src.ReadDateTime() + if err != nil { + break + } + err = dst.WriteDateTime(dt) + case bsontype.Null: + err = src.ReadNull() + if err != nil { + break + } + err = dst.WriteNull() + case bsontype.Regex: + var pattern, options string + pattern, options, err = src.ReadRegex() + if err != nil { + break + } + err = dst.WriteRegex(pattern, options) + case bsontype.DBPointer: + var ns string + var pointer primitive.ObjectID + ns, pointer, err = src.ReadDBPointer() + if err != nil { + break + } + err = dst.WriteDBPointer(ns, pointer) + case bsontype.JavaScript: + var js string + js, err = src.ReadJavascript() + if err != nil { + break + } + err = dst.WriteJavascript(js) + case bsontype.Symbol: + var symbol string + symbol, err = src.ReadSymbol() + if err != nil { + break + } + err = dst.WriteSymbol(symbol) + case bsontype.CodeWithScope: + var code string + var srcScope DocumentReader + code, srcScope, err = src.ReadCodeWithScope() + if err != nil { + break + } + + var dstScope DocumentWriter + dstScope, err = dst.WriteCodeWithScope(code) + if err != nil { + break + } + err = c.copyDocumentCore(dstScope, srcScope) + case bsontype.Int32: + var i32 int32 + i32, err = src.ReadInt32() + if err != nil { + break + } + err = dst.WriteInt32(i32) + case bsontype.Timestamp: + var t, i uint32 + t, i, err = src.ReadTimestamp() + if err != nil { + break + } + err = dst.WriteTimestamp(t, i) + case bsontype.Int64: + var i64 int64 + i64, err = src.ReadInt64() + if err != nil { + break + } + err = dst.WriteInt64(i64) + case bsontype.Decimal128: + var d128 primitive.Decimal128 + d128, err = src.ReadDecimal128() + if err != nil { + break + } + err = dst.WriteDecimal128(d128) + case bsontype.MinKey: + err = src.ReadMinKey() + if err != nil { + break + } + err = dst.WriteMinKey() + case bsontype.MaxKey: + err = src.ReadMaxKey() + if err != nil { + break + } + err = dst.WriteMaxKey() + default: + err = fmt.Errorf("Cannot copy unknown BSON type %s", src.Type()) + } + + return err +} + +func (c Copier) copyArray(dst ValueWriter, src ValueReader) error { + ar, err := src.ReadArray() + if err != nil { + return err + } + + aw, err := dst.WriteArray() + if err != nil { + return err + } + + for { + vr, err := ar.ReadValue() + if err == ErrEOA { + break + } + if err != nil { + return err + } + + vw, err := aw.WriteArrayElement() + if err != nil { + return err + } + + err = c.CopyValue(vw, vr) + if err != nil { + return err + } + } + + return aw.WriteArrayEnd() +} + +func (c Copier) copyDocumentCore(dw DocumentWriter, dr DocumentReader) error { + for { + key, vr, err := dr.ReadElement() + if err == ErrEOD { + break + } + if err != nil { + return err + } + + vw, err := dw.WriteDocumentElement(key) + if err != nil { + return err + } + + err = c.CopyValue(vw, vr) + if err != nil { + return err + } + } + + return dw.WriteDocumentEnd() +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/doc.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/doc.go new file mode 100644 index 0000000..750b0d2 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/doc.go @@ -0,0 +1,9 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +// Package bsonrw contains abstractions for reading and writing +// BSON and BSON like types from sources. +package bsonrw // import "go.mongodb.org/mongo-driver/bson/bsonrw" diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_parser.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_parser.go new file mode 100644 index 0000000..54c76bf --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_parser.go @@ -0,0 +1,806 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonrw + +import ( + "encoding/base64" + "encoding/hex" + "errors" + "fmt" + "io" + "strings" + + "go.mongodb.org/mongo-driver/bson/bsontype" +) + +const maxNestingDepth = 200 + +// ErrInvalidJSON indicates the JSON input is invalid +var ErrInvalidJSON = errors.New("invalid JSON input") + +type jsonParseState byte + +const ( + jpsStartState jsonParseState = iota + jpsSawBeginObject + jpsSawEndObject + jpsSawBeginArray + jpsSawEndArray + jpsSawColon + jpsSawComma + jpsSawKey + jpsSawValue + jpsDoneState + jpsInvalidState +) + +type jsonParseMode byte + +const ( + jpmInvalidMode jsonParseMode = iota + jpmObjectMode + jpmArrayMode +) + +type extJSONValue struct { + t bsontype.Type + v interface{} +} + +type extJSONObject struct { + keys []string + values []*extJSONValue +} + +type extJSONParser struct { + js *jsonScanner + s jsonParseState + m []jsonParseMode + k string + v *extJSONValue + + err error + canonical bool + depth int + maxDepth int + + emptyObject bool + relaxedUUID bool +} + +// newExtJSONParser returns a new extended JSON parser, ready to to begin +// parsing from the first character of the argued json input. It will not +// perform any read-ahead and will therefore not report any errors about +// malformed JSON at this point. +func newExtJSONParser(r io.Reader, canonical bool) *extJSONParser { + return &extJSONParser{ + js: &jsonScanner{r: r}, + s: jpsStartState, + m: []jsonParseMode{}, + canonical: canonical, + maxDepth: maxNestingDepth, + } +} + +// peekType examines the next value and returns its BSON Type +func (ejp *extJSONParser) peekType() (bsontype.Type, error) { + var t bsontype.Type + var err error + initialState := ejp.s + + ejp.advanceState() + switch ejp.s { + case jpsSawValue: + t = ejp.v.t + case jpsSawBeginArray: + t = bsontype.Array + case jpsInvalidState: + err = ejp.err + case jpsSawComma: + // in array mode, seeing a comma means we need to progress again to actually observe a type + if ejp.peekMode() == jpmArrayMode { + return ejp.peekType() + } + case jpsSawEndArray: + // this would only be a valid state if we were in array mode, so return end-of-array error + err = ErrEOA + case jpsSawBeginObject: + // peek key to determine type + ejp.advanceState() + switch ejp.s { + case jpsSawEndObject: // empty embedded document + t = bsontype.EmbeddedDocument + ejp.emptyObject = true + case jpsInvalidState: + err = ejp.err + case jpsSawKey: + if initialState == jpsStartState { + return bsontype.EmbeddedDocument, nil + } + t = wrapperKeyBSONType(ejp.k) + + // if $uuid is encountered, parse as binary subtype 4 + if ejp.k == "$uuid" { + ejp.relaxedUUID = true + t = bsontype.Binary + } + + switch t { + case bsontype.JavaScript: + // just saw $code, need to check for $scope at same level + _, err = ejp.readValue(bsontype.JavaScript) + if err != nil { + break + } + + switch ejp.s { + case jpsSawEndObject: // type is TypeJavaScript + case jpsSawComma: + ejp.advanceState() + + if ejp.s == jpsSawKey && ejp.k == "$scope" { + t = bsontype.CodeWithScope + } else { + err = fmt.Errorf("invalid extended JSON: unexpected key %s in CodeWithScope object", ejp.k) + } + case jpsInvalidState: + err = ejp.err + default: + err = ErrInvalidJSON + } + case bsontype.CodeWithScope: + err = errors.New("invalid extended JSON: code with $scope must contain $code before $scope") + } + } + } + + return t, err +} + +// readKey parses the next key and its type and returns them +func (ejp *extJSONParser) readKey() (string, bsontype.Type, error) { + if ejp.emptyObject { + ejp.emptyObject = false + return "", 0, ErrEOD + } + + // advance to key (or return with error) + switch ejp.s { + case jpsStartState: + ejp.advanceState() + if ejp.s == jpsSawBeginObject { + ejp.advanceState() + } + case jpsSawBeginObject: + ejp.advanceState() + case jpsSawValue, jpsSawEndObject, jpsSawEndArray: + ejp.advanceState() + switch ejp.s { + case jpsSawBeginObject, jpsSawComma: + ejp.advanceState() + case jpsSawEndObject: + return "", 0, ErrEOD + case jpsDoneState: + return "", 0, io.EOF + case jpsInvalidState: + return "", 0, ejp.err + default: + return "", 0, ErrInvalidJSON + } + case jpsSawKey: // do nothing (key was peeked before) + default: + return "", 0, invalidRequestError("key") + } + + // read key + var key string + + switch ejp.s { + case jpsSawKey: + key = ejp.k + case jpsSawEndObject: + return "", 0, ErrEOD + case jpsInvalidState: + return "", 0, ejp.err + default: + return "", 0, invalidRequestError("key") + } + + // check for colon + ejp.advanceState() + if err := ensureColon(ejp.s, key); err != nil { + return "", 0, err + } + + // peek at the value to determine type + t, err := ejp.peekType() + if err != nil { + return "", 0, err + } + + return key, t, nil +} + +// readValue returns the value corresponding to the Type returned by peekType +func (ejp *extJSONParser) readValue(t bsontype.Type) (*extJSONValue, error) { + if ejp.s == jpsInvalidState { + return nil, ejp.err + } + + var v *extJSONValue + + switch t { + case bsontype.Null, bsontype.Boolean, bsontype.String: + if ejp.s != jpsSawValue { + return nil, invalidRequestError(t.String()) + } + v = ejp.v + case bsontype.Int32, bsontype.Int64, bsontype.Double: + // relaxed version allows these to be literal number values + if ejp.s == jpsSawValue { + v = ejp.v + break + } + fallthrough + case bsontype.Decimal128, bsontype.Symbol, bsontype.ObjectID, bsontype.MinKey, bsontype.MaxKey, bsontype.Undefined: + switch ejp.s { + case jpsSawKey: + // read colon + ejp.advanceState() + if err := ensureColon(ejp.s, ejp.k); err != nil { + return nil, err + } + + // read value + ejp.advanceState() + if ejp.s != jpsSawValue || !ejp.ensureExtValueType(t) { + return nil, invalidJSONErrorForType("value", t) + } + + v = ejp.v + + // read end object + ejp.advanceState() + if ejp.s != jpsSawEndObject { + return nil, invalidJSONErrorForType("} after value", t) + } + default: + return nil, invalidRequestError(t.String()) + } + case bsontype.Binary, bsontype.Regex, bsontype.Timestamp, bsontype.DBPointer: + if ejp.s != jpsSawKey { + return nil, invalidRequestError(t.String()) + } + // read colon + ejp.advanceState() + if err := ensureColon(ejp.s, ejp.k); err != nil { + return nil, err + } + + ejp.advanceState() + if t == bsontype.Binary && ejp.s == jpsSawValue { + // convert relaxed $uuid format + if ejp.relaxedUUID { + defer func() { ejp.relaxedUUID = false }() + uuid, err := ejp.v.parseSymbol() + if err != nil { + return nil, err + } + + // RFC 4122 defines the length of a UUID as 36 and the hyphens in a UUID as appearing + // in the 8th, 13th, 18th, and 23rd characters. + // + // See https://tools.ietf.org/html/rfc4122#section-3 + valid := len(uuid) == 36 && + string(uuid[8]) == "-" && + string(uuid[13]) == "-" && + string(uuid[18]) == "-" && + string(uuid[23]) == "-" + if !valid { + return nil, fmt.Errorf("$uuid value does not follow RFC 4122 format regarding length and hyphens") + } + + // remove hyphens + uuidNoHyphens := strings.Replace(uuid, "-", "", -1) + if len(uuidNoHyphens) != 32 { + return nil, fmt.Errorf("$uuid value does not follow RFC 4122 format regarding length and hyphens") + } + + // convert hex to bytes + bytes, err := hex.DecodeString(uuidNoHyphens) + if err != nil { + return nil, fmt.Errorf("$uuid value does not follow RFC 4122 format regarding hex bytes: %v", err) + } + + ejp.advanceState() + if ejp.s != jpsSawEndObject { + return nil, invalidJSONErrorForType("$uuid and value and then }", bsontype.Binary) + } + + base64 := &extJSONValue{ + t: bsontype.String, + v: base64.StdEncoding.EncodeToString(bytes), + } + subType := &extJSONValue{ + t: bsontype.String, + v: "04", + } + + v = &extJSONValue{ + t: bsontype.EmbeddedDocument, + v: &extJSONObject{ + keys: []string{"base64", "subType"}, + values: []*extJSONValue{base64, subType}, + }, + } + + break + } + + // convert legacy $binary format + base64 := ejp.v + + ejp.advanceState() + if ejp.s != jpsSawComma { + return nil, invalidJSONErrorForType(",", bsontype.Binary) + } + + ejp.advanceState() + key, t, err := ejp.readKey() + if err != nil { + return nil, err + } + if key != "$type" { + return nil, invalidJSONErrorForType("$type", bsontype.Binary) + } + + subType, err := ejp.readValue(t) + if err != nil { + return nil, err + } + + ejp.advanceState() + if ejp.s != jpsSawEndObject { + return nil, invalidJSONErrorForType("2 key-value pairs and then }", bsontype.Binary) + } + + v = &extJSONValue{ + t: bsontype.EmbeddedDocument, + v: &extJSONObject{ + keys: []string{"base64", "subType"}, + values: []*extJSONValue{base64, subType}, + }, + } + break + } + + // read KV pairs + if ejp.s != jpsSawBeginObject { + return nil, invalidJSONErrorForType("{", t) + } + + keys, vals, err := ejp.readObject(2, true) + if err != nil { + return nil, err + } + + ejp.advanceState() + if ejp.s != jpsSawEndObject { + return nil, invalidJSONErrorForType("2 key-value pairs and then }", t) + } + + v = &extJSONValue{t: bsontype.EmbeddedDocument, v: &extJSONObject{keys: keys, values: vals}} + + case bsontype.DateTime: + switch ejp.s { + case jpsSawValue: + v = ejp.v + case jpsSawKey: + // read colon + ejp.advanceState() + if err := ensureColon(ejp.s, ejp.k); err != nil { + return nil, err + } + + ejp.advanceState() + switch ejp.s { + case jpsSawBeginObject: + keys, vals, err := ejp.readObject(1, true) + if err != nil { + return nil, err + } + v = &extJSONValue{t: bsontype.EmbeddedDocument, v: &extJSONObject{keys: keys, values: vals}} + case jpsSawValue: + if ejp.canonical { + return nil, invalidJSONError("{") + } + v = ejp.v + default: + if ejp.canonical { + return nil, invalidJSONErrorForType("object", t) + } + return nil, invalidJSONErrorForType("ISO-8601 Internet Date/Time Format as described in RFC-3339", t) + } + + ejp.advanceState() + if ejp.s != jpsSawEndObject { + return nil, invalidJSONErrorForType("value and then }", t) + } + default: + return nil, invalidRequestError(t.String()) + } + case bsontype.JavaScript: + switch ejp.s { + case jpsSawKey: + // read colon + ejp.advanceState() + if err := ensureColon(ejp.s, ejp.k); err != nil { + return nil, err + } + + // read value + ejp.advanceState() + if ejp.s != jpsSawValue { + return nil, invalidJSONErrorForType("value", t) + } + v = ejp.v + + // read end object or comma and just return + ejp.advanceState() + case jpsSawEndObject: + v = ejp.v + default: + return nil, invalidRequestError(t.String()) + } + case bsontype.CodeWithScope: + if ejp.s == jpsSawKey && ejp.k == "$scope" { + v = ejp.v // this is the $code string from earlier + + // read colon + ejp.advanceState() + if err := ensureColon(ejp.s, ejp.k); err != nil { + return nil, err + } + + // read { + ejp.advanceState() + if ejp.s != jpsSawBeginObject { + return nil, invalidJSONError("$scope to be embedded document") + } + } else { + return nil, invalidRequestError(t.String()) + } + case bsontype.EmbeddedDocument, bsontype.Array: + return nil, invalidRequestError(t.String()) + } + + return v, nil +} + +// readObject is a utility method for reading full objects of known (or expected) size +// it is useful for extended JSON types such as binary, datetime, regex, and timestamp +func (ejp *extJSONParser) readObject(numKeys int, started bool) ([]string, []*extJSONValue, error) { + keys := make([]string, numKeys) + vals := make([]*extJSONValue, numKeys) + + if !started { + ejp.advanceState() + if ejp.s != jpsSawBeginObject { + return nil, nil, invalidJSONError("{") + } + } + + for i := 0; i < numKeys; i++ { + key, t, err := ejp.readKey() + if err != nil { + return nil, nil, err + } + + switch ejp.s { + case jpsSawKey: + v, err := ejp.readValue(t) + if err != nil { + return nil, nil, err + } + + keys[i] = key + vals[i] = v + case jpsSawValue: + keys[i] = key + vals[i] = ejp.v + default: + return nil, nil, invalidJSONError("value") + } + } + + ejp.advanceState() + if ejp.s != jpsSawEndObject { + return nil, nil, invalidJSONError("}") + } + + return keys, vals, nil +} + +// advanceState reads the next JSON token from the scanner and transitions +// from the current state based on that token's type +func (ejp *extJSONParser) advanceState() { + if ejp.s == jpsDoneState || ejp.s == jpsInvalidState { + return + } + + jt, err := ejp.js.nextToken() + + if err != nil { + ejp.err = err + ejp.s = jpsInvalidState + return + } + + valid := ejp.validateToken(jt.t) + if !valid { + ejp.err = unexpectedTokenError(jt) + ejp.s = jpsInvalidState + return + } + + switch jt.t { + case jttBeginObject: + ejp.s = jpsSawBeginObject + ejp.pushMode(jpmObjectMode) + ejp.depth++ + + if ejp.depth > ejp.maxDepth { + ejp.err = nestingDepthError(jt.p, ejp.depth) + ejp.s = jpsInvalidState + } + case jttEndObject: + ejp.s = jpsSawEndObject + ejp.depth-- + + if ejp.popMode() != jpmObjectMode { + ejp.err = unexpectedTokenError(jt) + ejp.s = jpsInvalidState + } + case jttBeginArray: + ejp.s = jpsSawBeginArray + ejp.pushMode(jpmArrayMode) + case jttEndArray: + ejp.s = jpsSawEndArray + + if ejp.popMode() != jpmArrayMode { + ejp.err = unexpectedTokenError(jt) + ejp.s = jpsInvalidState + } + case jttColon: + ejp.s = jpsSawColon + case jttComma: + ejp.s = jpsSawComma + case jttEOF: + ejp.s = jpsDoneState + if len(ejp.m) != 0 { + ejp.err = unexpectedTokenError(jt) + ejp.s = jpsInvalidState + } + case jttString: + switch ejp.s { + case jpsSawComma: + if ejp.peekMode() == jpmArrayMode { + ejp.s = jpsSawValue + ejp.v = extendJSONToken(jt) + return + } + fallthrough + case jpsSawBeginObject: + ejp.s = jpsSawKey + ejp.k = jt.v.(string) + return + } + fallthrough + default: + ejp.s = jpsSawValue + ejp.v = extendJSONToken(jt) + } +} + +var jpsValidTransitionTokens = map[jsonParseState]map[jsonTokenType]bool{ + jpsStartState: { + jttBeginObject: true, + jttBeginArray: true, + jttInt32: true, + jttInt64: true, + jttDouble: true, + jttString: true, + jttBool: true, + jttNull: true, + jttEOF: true, + }, + jpsSawBeginObject: { + jttEndObject: true, + jttString: true, + }, + jpsSawEndObject: { + jttEndObject: true, + jttEndArray: true, + jttComma: true, + jttEOF: true, + }, + jpsSawBeginArray: { + jttBeginObject: true, + jttBeginArray: true, + jttEndArray: true, + jttInt32: true, + jttInt64: true, + jttDouble: true, + jttString: true, + jttBool: true, + jttNull: true, + }, + jpsSawEndArray: { + jttEndObject: true, + jttEndArray: true, + jttComma: true, + jttEOF: true, + }, + jpsSawColon: { + jttBeginObject: true, + jttBeginArray: true, + jttInt32: true, + jttInt64: true, + jttDouble: true, + jttString: true, + jttBool: true, + jttNull: true, + }, + jpsSawComma: { + jttBeginObject: true, + jttBeginArray: true, + jttInt32: true, + jttInt64: true, + jttDouble: true, + jttString: true, + jttBool: true, + jttNull: true, + }, + jpsSawKey: { + jttColon: true, + }, + jpsSawValue: { + jttEndObject: true, + jttEndArray: true, + jttComma: true, + jttEOF: true, + }, + jpsDoneState: {}, + jpsInvalidState: {}, +} + +func (ejp *extJSONParser) validateToken(jtt jsonTokenType) bool { + switch ejp.s { + case jpsSawEndObject: + // if we are at depth zero and the next token is a '{', + // we can consider it valid only if we are not in array mode. + if jtt == jttBeginObject && ejp.depth == 0 { + return ejp.peekMode() != jpmArrayMode + } + case jpsSawComma: + switch ejp.peekMode() { + // the only valid next token after a comma inside a document is a string (a key) + case jpmObjectMode: + return jtt == jttString + case jpmInvalidMode: + return false + } + } + + _, ok := jpsValidTransitionTokens[ejp.s][jtt] + return ok +} + +// ensureExtValueType returns true if the current value has the expected +// value type for single-key extended JSON types. For example, +// {"$numberInt": v} v must be TypeString +func (ejp *extJSONParser) ensureExtValueType(t bsontype.Type) bool { + switch t { + case bsontype.MinKey, bsontype.MaxKey: + return ejp.v.t == bsontype.Int32 + case bsontype.Undefined: + return ejp.v.t == bsontype.Boolean + case bsontype.Int32, bsontype.Int64, bsontype.Double, bsontype.Decimal128, bsontype.Symbol, bsontype.ObjectID: + return ejp.v.t == bsontype.String + default: + return false + } +} + +func (ejp *extJSONParser) pushMode(m jsonParseMode) { + ejp.m = append(ejp.m, m) +} + +func (ejp *extJSONParser) popMode() jsonParseMode { + l := len(ejp.m) + if l == 0 { + return jpmInvalidMode + } + + m := ejp.m[l-1] + ejp.m = ejp.m[:l-1] + + return m +} + +func (ejp *extJSONParser) peekMode() jsonParseMode { + l := len(ejp.m) + if l == 0 { + return jpmInvalidMode + } + + return ejp.m[l-1] +} + +func extendJSONToken(jt *jsonToken) *extJSONValue { + var t bsontype.Type + + switch jt.t { + case jttInt32: + t = bsontype.Int32 + case jttInt64: + t = bsontype.Int64 + case jttDouble: + t = bsontype.Double + case jttString: + t = bsontype.String + case jttBool: + t = bsontype.Boolean + case jttNull: + t = bsontype.Null + default: + return nil + } + + return &extJSONValue{t: t, v: jt.v} +} + +func ensureColon(s jsonParseState, key string) error { + if s != jpsSawColon { + return fmt.Errorf("invalid JSON input: missing colon after key \"%s\"", key) + } + + return nil +} + +func invalidRequestError(s string) error { + return fmt.Errorf("invalid request to read %s", s) +} + +func invalidJSONError(expected string) error { + return fmt.Errorf("invalid JSON input; expected %s", expected) +} + +func invalidJSONErrorForType(expected string, t bsontype.Type) error { + return fmt.Errorf("invalid JSON input; expected %s for %s", expected, t) +} + +func unexpectedTokenError(jt *jsonToken) error { + switch jt.t { + case jttInt32, jttInt64, jttDouble: + return fmt.Errorf("invalid JSON input; unexpected number (%v) at position %d", jt.v, jt.p) + case jttString: + return fmt.Errorf("invalid JSON input; unexpected string (\"%v\") at position %d", jt.v, jt.p) + case jttBool: + return fmt.Errorf("invalid JSON input; unexpected boolean literal (%v) at position %d", jt.v, jt.p) + case jttNull: + return fmt.Errorf("invalid JSON input; unexpected null literal at position %d", jt.p) + case jttEOF: + return fmt.Errorf("invalid JSON input; unexpected end of input at position %d", jt.p) + default: + return fmt.Errorf("invalid JSON input; unexpected %c at position %d", jt.v.(byte), jt.p) + } +} + +func nestingDepthError(p, depth int) error { + return fmt.Errorf("invalid JSON input; nesting too deep (%d levels) at position %d", depth, p) +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_reader.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_reader.go new file mode 100644 index 0000000..35832d7 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_reader.go @@ -0,0 +1,644 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 + +package bsonrw + +import ( + "fmt" + "io" + "sync" + + "go.mongodb.org/mongo-driver/bson/bsontype" + "go.mongodb.org/mongo-driver/bson/primitive" +) + +// ExtJSONValueReaderPool is a pool for ValueReaders that read ExtJSON. +type ExtJSONValueReaderPool struct { + pool sync.Pool +} + +// NewExtJSONValueReaderPool instantiates a new ExtJSONValueReaderPool. +func NewExtJSONValueReaderPool() *ExtJSONValueReaderPool { + return &ExtJSONValueReaderPool{ + pool: sync.Pool{ + New: func() interface{} { + return new(extJSONValueReader) + }, + }, + } +} + +// Get retrieves a ValueReader from the pool and uses src as the underlying ExtJSON. +func (bvrp *ExtJSONValueReaderPool) Get(r io.Reader, canonical bool) (ValueReader, error) { + vr := bvrp.pool.Get().(*extJSONValueReader) + return vr.reset(r, canonical) +} + +// Put inserts a ValueReader into the pool. If the ValueReader is not a ExtJSON ValueReader nothing +// is inserted into the pool and ok will be false. +func (bvrp *ExtJSONValueReaderPool) Put(vr ValueReader) (ok bool) { + bvr, ok := vr.(*extJSONValueReader) + if !ok { + return false + } + + bvr, _ = bvr.reset(nil, false) + bvrp.pool.Put(bvr) + return true +} + +type ejvrState struct { + mode mode + vType bsontype.Type + depth int +} + +// extJSONValueReader is for reading extended JSON. +type extJSONValueReader struct { + p *extJSONParser + + stack []ejvrState + frame int +} + +// NewExtJSONValueReader creates a new ValueReader from a given io.Reader +// It will interpret the JSON of r as canonical or relaxed according to the +// given canonical flag +func NewExtJSONValueReader(r io.Reader, canonical bool) (ValueReader, error) { + return newExtJSONValueReader(r, canonical) +} + +func newExtJSONValueReader(r io.Reader, canonical bool) (*extJSONValueReader, error) { + ejvr := new(extJSONValueReader) + return ejvr.reset(r, canonical) +} + +func (ejvr *extJSONValueReader) reset(r io.Reader, canonical bool) (*extJSONValueReader, error) { + p := newExtJSONParser(r, canonical) + typ, err := p.peekType() + + if err != nil { + return nil, ErrInvalidJSON + } + + var m mode + switch typ { + case bsontype.EmbeddedDocument: + m = mTopLevel + case bsontype.Array: + m = mArray + default: + m = mValue + } + + stack := make([]ejvrState, 1, 5) + stack[0] = ejvrState{ + mode: m, + vType: typ, + } + return &extJSONValueReader{ + p: p, + stack: stack, + }, nil +} + +func (ejvr *extJSONValueReader) advanceFrame() { + if ejvr.frame+1 >= len(ejvr.stack) { // We need to grow the stack + length := len(ejvr.stack) + if length+1 >= cap(ejvr.stack) { + // double it + buf := make([]ejvrState, 2*cap(ejvr.stack)+1) + copy(buf, ejvr.stack) + ejvr.stack = buf + } + ejvr.stack = ejvr.stack[:length+1] + } + ejvr.frame++ + + // Clean the stack + ejvr.stack[ejvr.frame].mode = 0 + ejvr.stack[ejvr.frame].vType = 0 + ejvr.stack[ejvr.frame].depth = 0 +} + +func (ejvr *extJSONValueReader) pushDocument() { + ejvr.advanceFrame() + + ejvr.stack[ejvr.frame].mode = mDocument + ejvr.stack[ejvr.frame].depth = ejvr.p.depth +} + +func (ejvr *extJSONValueReader) pushCodeWithScope() { + ejvr.advanceFrame() + + ejvr.stack[ejvr.frame].mode = mCodeWithScope +} + +func (ejvr *extJSONValueReader) pushArray() { + ejvr.advanceFrame() + + ejvr.stack[ejvr.frame].mode = mArray +} + +func (ejvr *extJSONValueReader) push(m mode, t bsontype.Type) { + ejvr.advanceFrame() + + ejvr.stack[ejvr.frame].mode = m + ejvr.stack[ejvr.frame].vType = t +} + +func (ejvr *extJSONValueReader) pop() { + switch ejvr.stack[ejvr.frame].mode { + case mElement, mValue: + ejvr.frame-- + case mDocument, mArray, mCodeWithScope: + ejvr.frame -= 2 // we pop twice to jump over the vrElement: vrDocument -> vrElement -> vrDocument/TopLevel/etc... + } +} + +func (ejvr *extJSONValueReader) skipObject() { + // read entire object until depth returns to 0 (last ending } or ] seen) + depth := 1 + for depth > 0 { + ejvr.p.advanceState() + + // If object is empty, raise depth and continue. When emptyObject is true, the + // parser has already read both the opening and closing brackets of an empty + // object ("{}"), so the next valid token will be part of the parent document, + // not part of the nested document. + // + // If there is a comma, there are remaining fields, emptyObject must be set back + // to false, and comma must be skipped with advanceState(). + if ejvr.p.emptyObject { + if ejvr.p.s == jpsSawComma { + ejvr.p.emptyObject = false + ejvr.p.advanceState() + } + depth-- + continue + } + + switch ejvr.p.s { + case jpsSawBeginObject, jpsSawBeginArray: + depth++ + case jpsSawEndObject, jpsSawEndArray: + depth-- + } + } +} + +func (ejvr *extJSONValueReader) invalidTransitionErr(destination mode, name string, modes []mode) error { + te := TransitionError{ + name: name, + current: ejvr.stack[ejvr.frame].mode, + destination: destination, + modes: modes, + action: "read", + } + if ejvr.frame != 0 { + te.parent = ejvr.stack[ejvr.frame-1].mode + } + return te +} + +func (ejvr *extJSONValueReader) typeError(t bsontype.Type) error { + return fmt.Errorf("positioned on %s, but attempted to read %s", ejvr.stack[ejvr.frame].vType, t) +} + +func (ejvr *extJSONValueReader) ensureElementValue(t bsontype.Type, destination mode, callerName string, addModes ...mode) error { + switch ejvr.stack[ejvr.frame].mode { + case mElement, mValue: + if ejvr.stack[ejvr.frame].vType != t { + return ejvr.typeError(t) + } + default: + modes := []mode{mElement, mValue} + if addModes != nil { + modes = append(modes, addModes...) + } + return ejvr.invalidTransitionErr(destination, callerName, modes) + } + + return nil +} + +func (ejvr *extJSONValueReader) Type() bsontype.Type { + return ejvr.stack[ejvr.frame].vType +} + +func (ejvr *extJSONValueReader) Skip() error { + switch ejvr.stack[ejvr.frame].mode { + case mElement, mValue: + default: + return ejvr.invalidTransitionErr(0, "Skip", []mode{mElement, mValue}) + } + + defer ejvr.pop() + + t := ejvr.stack[ejvr.frame].vType + switch t { + case bsontype.Array, bsontype.EmbeddedDocument, bsontype.CodeWithScope: + // read entire array, doc or CodeWithScope + ejvr.skipObject() + default: + _, err := ejvr.p.readValue(t) + if err != nil { + return err + } + } + + return nil +} + +func (ejvr *extJSONValueReader) ReadArray() (ArrayReader, error) { + switch ejvr.stack[ejvr.frame].mode { + case mTopLevel: // allow reading array from top level + case mArray: + return ejvr, nil + default: + if err := ejvr.ensureElementValue(bsontype.Array, mArray, "ReadArray", mTopLevel, mArray); err != nil { + return nil, err + } + } + + ejvr.pushArray() + + return ejvr, nil +} + +func (ejvr *extJSONValueReader) ReadBinary() (b []byte, btype byte, err error) { + if err := ejvr.ensureElementValue(bsontype.Binary, 0, "ReadBinary"); err != nil { + return nil, 0, err + } + + v, err := ejvr.p.readValue(bsontype.Binary) + if err != nil { + return nil, 0, err + } + + b, btype, err = v.parseBinary() + + ejvr.pop() + return b, btype, err +} + +func (ejvr *extJSONValueReader) ReadBoolean() (bool, error) { + if err := ejvr.ensureElementValue(bsontype.Boolean, 0, "ReadBoolean"); err != nil { + return false, err + } + + v, err := ejvr.p.readValue(bsontype.Boolean) + if err != nil { + return false, err + } + + if v.t != bsontype.Boolean { + return false, fmt.Errorf("expected type bool, but got type %s", v.t) + } + + ejvr.pop() + return v.v.(bool), nil +} + +func (ejvr *extJSONValueReader) ReadDocument() (DocumentReader, error) { + switch ejvr.stack[ejvr.frame].mode { + case mTopLevel: + return ejvr, nil + case mElement, mValue: + if ejvr.stack[ejvr.frame].vType != bsontype.EmbeddedDocument { + return nil, ejvr.typeError(bsontype.EmbeddedDocument) + } + + ejvr.pushDocument() + return ejvr, nil + default: + return nil, ejvr.invalidTransitionErr(mDocument, "ReadDocument", []mode{mTopLevel, mElement, mValue}) + } +} + +func (ejvr *extJSONValueReader) ReadCodeWithScope() (code string, dr DocumentReader, err error) { + if err = ejvr.ensureElementValue(bsontype.CodeWithScope, 0, "ReadCodeWithScope"); err != nil { + return "", nil, err + } + + v, err := ejvr.p.readValue(bsontype.CodeWithScope) + if err != nil { + return "", nil, err + } + + code, err = v.parseJavascript() + + ejvr.pushCodeWithScope() + return code, ejvr, err +} + +func (ejvr *extJSONValueReader) ReadDBPointer() (ns string, oid primitive.ObjectID, err error) { + if err = ejvr.ensureElementValue(bsontype.DBPointer, 0, "ReadDBPointer"); err != nil { + return "", primitive.NilObjectID, err + } + + v, err := ejvr.p.readValue(bsontype.DBPointer) + if err != nil { + return "", primitive.NilObjectID, err + } + + ns, oid, err = v.parseDBPointer() + + ejvr.pop() + return ns, oid, err +} + +func (ejvr *extJSONValueReader) ReadDateTime() (int64, error) { + if err := ejvr.ensureElementValue(bsontype.DateTime, 0, "ReadDateTime"); err != nil { + return 0, err + } + + v, err := ejvr.p.readValue(bsontype.DateTime) + if err != nil { + return 0, err + } + + d, err := v.parseDateTime() + + ejvr.pop() + return d, err +} + +func (ejvr *extJSONValueReader) ReadDecimal128() (primitive.Decimal128, error) { + if err := ejvr.ensureElementValue(bsontype.Decimal128, 0, "ReadDecimal128"); err != nil { + return primitive.Decimal128{}, err + } + + v, err := ejvr.p.readValue(bsontype.Decimal128) + if err != nil { + return primitive.Decimal128{}, err + } + + d, err := v.parseDecimal128() + + ejvr.pop() + return d, err +} + +func (ejvr *extJSONValueReader) ReadDouble() (float64, error) { + if err := ejvr.ensureElementValue(bsontype.Double, 0, "ReadDouble"); err != nil { + return 0, err + } + + v, err := ejvr.p.readValue(bsontype.Double) + if err != nil { + return 0, err + } + + d, err := v.parseDouble() + + ejvr.pop() + return d, err +} + +func (ejvr *extJSONValueReader) ReadInt32() (int32, error) { + if err := ejvr.ensureElementValue(bsontype.Int32, 0, "ReadInt32"); err != nil { + return 0, err + } + + v, err := ejvr.p.readValue(bsontype.Int32) + if err != nil { + return 0, err + } + + i, err := v.parseInt32() + + ejvr.pop() + return i, err +} + +func (ejvr *extJSONValueReader) ReadInt64() (int64, error) { + if err := ejvr.ensureElementValue(bsontype.Int64, 0, "ReadInt64"); err != nil { + return 0, err + } + + v, err := ejvr.p.readValue(bsontype.Int64) + if err != nil { + return 0, err + } + + i, err := v.parseInt64() + + ejvr.pop() + return i, err +} + +func (ejvr *extJSONValueReader) ReadJavascript() (code string, err error) { + if err = ejvr.ensureElementValue(bsontype.JavaScript, 0, "ReadJavascript"); err != nil { + return "", err + } + + v, err := ejvr.p.readValue(bsontype.JavaScript) + if err != nil { + return "", err + } + + code, err = v.parseJavascript() + + ejvr.pop() + return code, err +} + +func (ejvr *extJSONValueReader) ReadMaxKey() error { + if err := ejvr.ensureElementValue(bsontype.MaxKey, 0, "ReadMaxKey"); err != nil { + return err + } + + v, err := ejvr.p.readValue(bsontype.MaxKey) + if err != nil { + return err + } + + err = v.parseMinMaxKey("max") + + ejvr.pop() + return err +} + +func (ejvr *extJSONValueReader) ReadMinKey() error { + if err := ejvr.ensureElementValue(bsontype.MinKey, 0, "ReadMinKey"); err != nil { + return err + } + + v, err := ejvr.p.readValue(bsontype.MinKey) + if err != nil { + return err + } + + err = v.parseMinMaxKey("min") + + ejvr.pop() + return err +} + +func (ejvr *extJSONValueReader) ReadNull() error { + if err := ejvr.ensureElementValue(bsontype.Null, 0, "ReadNull"); err != nil { + return err + } + + v, err := ejvr.p.readValue(bsontype.Null) + if err != nil { + return err + } + + if v.t != bsontype.Null { + return fmt.Errorf("expected type null but got type %s", v.t) + } + + ejvr.pop() + return nil +} + +func (ejvr *extJSONValueReader) ReadObjectID() (primitive.ObjectID, error) { + if err := ejvr.ensureElementValue(bsontype.ObjectID, 0, "ReadObjectID"); err != nil { + return primitive.ObjectID{}, err + } + + v, err := ejvr.p.readValue(bsontype.ObjectID) + if err != nil { + return primitive.ObjectID{}, err + } + + oid, err := v.parseObjectID() + + ejvr.pop() + return oid, err +} + +func (ejvr *extJSONValueReader) ReadRegex() (pattern string, options string, err error) { + if err = ejvr.ensureElementValue(bsontype.Regex, 0, "ReadRegex"); err != nil { + return "", "", err + } + + v, err := ejvr.p.readValue(bsontype.Regex) + if err != nil { + return "", "", err + } + + pattern, options, err = v.parseRegex() + + ejvr.pop() + return pattern, options, err +} + +func (ejvr *extJSONValueReader) ReadString() (string, error) { + if err := ejvr.ensureElementValue(bsontype.String, 0, "ReadString"); err != nil { + return "", err + } + + v, err := ejvr.p.readValue(bsontype.String) + if err != nil { + return "", err + } + + if v.t != bsontype.String { + return "", fmt.Errorf("expected type string but got type %s", v.t) + } + + ejvr.pop() + return v.v.(string), nil +} + +func (ejvr *extJSONValueReader) ReadSymbol() (symbol string, err error) { + if err = ejvr.ensureElementValue(bsontype.Symbol, 0, "ReadSymbol"); err != nil { + return "", err + } + + v, err := ejvr.p.readValue(bsontype.Symbol) + if err != nil { + return "", err + } + + symbol, err = v.parseSymbol() + + ejvr.pop() + return symbol, err +} + +func (ejvr *extJSONValueReader) ReadTimestamp() (t uint32, i uint32, err error) { + if err = ejvr.ensureElementValue(bsontype.Timestamp, 0, "ReadTimestamp"); err != nil { + return 0, 0, err + } + + v, err := ejvr.p.readValue(bsontype.Timestamp) + if err != nil { + return 0, 0, err + } + + t, i, err = v.parseTimestamp() + + ejvr.pop() + return t, i, err +} + +func (ejvr *extJSONValueReader) ReadUndefined() error { + if err := ejvr.ensureElementValue(bsontype.Undefined, 0, "ReadUndefined"); err != nil { + return err + } + + v, err := ejvr.p.readValue(bsontype.Undefined) + if err != nil { + return err + } + + err = v.parseUndefined() + + ejvr.pop() + return err +} + +func (ejvr *extJSONValueReader) ReadElement() (string, ValueReader, error) { + switch ejvr.stack[ejvr.frame].mode { + case mTopLevel, mDocument, mCodeWithScope: + default: + return "", nil, ejvr.invalidTransitionErr(mElement, "ReadElement", []mode{mTopLevel, mDocument, mCodeWithScope}) + } + + name, t, err := ejvr.p.readKey() + + if err != nil { + if err == ErrEOD { + if ejvr.stack[ejvr.frame].mode == mCodeWithScope { + _, err := ejvr.p.peekType() + if err != nil { + return "", nil, err + } + } + + ejvr.pop() + } + + return "", nil, err + } + + ejvr.push(mElement, t) + return name, ejvr, nil +} + +func (ejvr *extJSONValueReader) ReadValue() (ValueReader, error) { + switch ejvr.stack[ejvr.frame].mode { + case mArray: + default: + return nil, ejvr.invalidTransitionErr(mValue, "ReadValue", []mode{mArray}) + } + + t, err := ejvr.p.peekType() + if err != nil { + if err == ErrEOA { + ejvr.pop() + } + + return nil, err + } + + ejvr.push(mValue, t) + return ejvr, nil +} diff --git a/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_tables.go b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_tables.go new file mode 100644 index 0000000..ba39c96 --- /dev/null +++ b/vendor/go.mongodb.org/mongo-driver/bson/bsonrw/extjson_tables.go @@ -0,0 +1,223 @@ +// Copyright (C) MongoDB, Inc. 2017-present. +// +// Licensed under the Apache License, Version 2.0 (the "License"); you may +// not use this file except in compliance with the License. You may obtain +// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 +// +// Based on github.com/golang/go by The Go Authors +// See THIRD-PARTY-NOTICES for original license terms. + +package bsonrw + +import "unicode/utf8" + +// safeSet holds the value true if the ASCII character with the given array +// position can be represented inside a JSON string without any further +// escaping. +// +// All values are true except for the ASCII control characters (0-31), the +// double quote ("), and the backslash character ("\"). +var safeSet = [utf8.RuneSelf]bool{ + ' ': true, + '!': true, + '"': false, + '#': true, + '$': true, + '%': true, + '&': true, + '\'': true, + '(': true, + ')': true, + '*': true, + '+': true, + ',': true, + '-': true, + '.': true, + '/': true, + '0': true, + '1': true, + '2': true, + '3': true, + '4': true, + '5': true, + '6': true, + '7': true, + '8': true, + '9': true, + ':': true, + ';': true, + '<': true, + '=': true, + '>': true, + '?': true, + '@': true, + 'A': true, + 'B': true, + 'C': true, + 'D': true, + 'E': true, + 'F': true, + 'G': true, + 'H': true, + 'I': true, + 'J': true, + 'K': true, + 'L': true, + 'M': true, + 'N': true, + 'O': true, + 'P': true, + 'Q': true, + 'R': true, + 'S': true, + 'T': true, + 'U': true, + 'V': true, + 'W': true, + 'X': true, + 'Y': true, + 'Z': true, + '[': true, + '\\': false, + ']': true, + '^': true, + '_': true, + '`': true, + 'a': true, + 'b': true, + 'c': true, + 'd': true, + 'e': true, + 'f': true, + 'g': true, + 'h': true, + 'i': true, + 'j': true, + 'k': true, + 'l': true, + 'm': true, + 'n': true, + 'o': true, + 'p': true, + 'q': true, + 'r': true, + 's': true, + 't': true, + 'u': true, + 'v': true, + 'w': true, + 'x': true, + 'y': true, + 'z': true, + '{': true, + '|': true, + '}': true, + '~': true, + '\u007f': true, +} + +// htmlSafeSet holds the value true if the ASCII character with the given +// array position can be safely represented inside a JSON string, embedded +// inside of HTML