- Getting started with the Kythe codebase
- Code review
- Linting and Testing
- Contribution Ideas
Getting started with the Kythe codebase
All changes to Kythe must go through code review before being submitted, and each individual or corporate contributor must sign an appropriate Contributor License Agreement. Once your CLA is submitted (or if you already submitted one for another Google project), make a commit adding yourself to the AUTHORS and CONTRIBUTORS files. This commit can be part of your first Pull Request.
The Kythe team has chosen to use GitHub Pull Requests for code review.
Linting and Testing
./tools/git/setup_hooks.sh to initialize the standard Git hooks used by
the Kythe developers. This ensures that each commit passes all unit/lint tests
and that commit descriptions follow the Conventional
before a requesting code review. If necessary, use
git commit --no-verify
to bypass these checks. In addition to the tools necessary to build Kythe, the
following are necessary to run the installed hooks:
Each Pull Request will also require a gambit of tests run by Kythe’s instance of
Buildbot. All non-maintainers
will require a Kythe team member to add the
buildbot label to each Pull
Request to verify the PR should be checked by Buildbot.
Kythe C++ code follows the Google style guide. You can run
clang-format to do
If you forgot to do this for a commit, you can amend it easily:
Likewise, all Java code should be formatted by the latest version of
google-java-format and Go code should pass through
New Extractors and Indexers
Kythe is built on the idea of having a common set of tools across programming languages so Kythe is always happy to add another language to its family.
Build System Integration
In order to use Kythe’s compilation extractors, they must be given precise information about how a compilation is processed. Currently, Kythe has built-in support for Bazel and rudimentary support for CMake. Contributing support for more build systems like Gradle will greatly help the ease of use for Kythe and increase the breadth of what it can index.
Kythe emits a lot of data and there are many ways to interpret/display it all. Kythe has provided a sample UI, but it currently only scratches the surface of Kythe’s data. Other ideas for visualizers include an interactive graph, a documentation browser and a source file hierarchical overview.