.gitea/workflows | ||
.github/workflows | ||
codegen | ||
example | ||
src | ||
.gitignore | ||
build.zig | ||
build.zig.zon | ||
LICENSE | ||
Makefile | ||
README.md |
AWS SDK for Zig
Last Mach Nominated Zig Version:
Current executable size for the demo is 980k after compiling with -Doptimize=ReleaseSmall in x86_linux, and will vary based on services used. Tested targets:
- x86_64-linux
- riscv64-linux*
- aarch64-linux
- x86_64-windows**
- arm-linux
- aarch64-macos
- x86_64-macos
Tested targets are built, but not continuously tested, by CI.
* On Zig 0.12, riscv64-linux tests take a significant time to compile (each aws.zig test takes approximately 1min, 45 seconds to compile on Intel i9 10th gen)
** On Zig 0.12, x86_64-windows tests have one test skipped as LLVM consumes all available RAM on the system
Zig-Develop Branch
This branch is intended for use with the in-development version of Zig. This starts with 0.12.0-dev.3180+83e578a18. I will try to keep this branch up to date with latest, but with a special eye towards aligning with Mach Engine's Nominated Zig Versions. As nightly zig versions disappear off the downloads page (and back end server), we can use the mirroring that the Mach Engine participates in to pull these versions.
Building
zig build
should work. It will build the code generation project, fetch model
files from upstream AWS Go SDK v2, run the code generation, then build the main
project with the generated code. Testing can be done with zig test
.
Using
This is designed for use with the Zig 0.11 package manager, and exposes a module
called "aws". Set up build.zig.zon
and add the dependency/module to your project
as normal and the package manager should do its thing. A full example can be found
in /example.
Configuring the module and/or Running the demo
This library mimics the aws c libraries for it's work, so it operates like most
other 'AWS things'. /src/main.zig gives you a handful of examples
for working with services. For local testing or alternative endpoints, there's
no real standard, so there is code to look for AWS_ENDPOINT_URL
environment
variable that will supersede all other configuration.
Limitations
WebIdentityToken is not yet implemented.
TODO List:
- Json parsing is based on a fork of the 0.9.0 (maybe earlier?) json parser. This needs a re-visit. Note also that a json.zig file is embedded/copied from the codegen project, so that also needs a second look.
- Take a look to see about compilation speed. With codegen caching this is reasonable, but still takes longer than needed.
- Upgrade the model files. This is a simple tasks, but I'd like the first item on this list to be completed first.
- Implement sigv4a signing
- Implement jitter/exponential backoff
- Implement timeouts and other TODO's in the code
- Add option to cache signature keys
Services without TLS 1.3 support
All AWS services should support TLS 1.3 at this point, but there are many regions and several partitions, and not all of them have been tested, so your mileage may vary.
Dependency tree
No dependencies:
- aws_authentication: base structure for credentials (only one type)
- aws_http_base: contains basic structures for http requests/results
- case: provides functions to change casing
- date: provides limited date manipulation functions
- http_client_17015_issue: zig 0.11 http client, with changes
- json: custom version of earlier stdlib json parser
- xml: custom xml parser library
- url: custom url encoding
aws_credentials: Allows credential handling aws_authentication
aws_http: http_client_17015_issue aws_http_base aws_signing
aws_signing: handles signing of http requests aws_http_base aws_authentication date
aws: main usage point for libraries aws_http json url case date servicemodel xml_shaper aws_credentials aws_authentication
main: main entrypoint for demo executable aws
servicemodel: Provides access to all aws service generated models all generated model files
xml_shaper: Manages interface from xml to in memory structures xml date