We'd love to accept your patches! Before we can take them, we have to jump a couple of legal hurdles.
Please fill out either the individual or corporate Contributor License Agreement (CLA).
- If you are an individual writing original source code and you're sure you own the intellectual property, then you'll need to sign an individual CLA.
- If you work for a company that wants to allow you to contribute your work, then you'll need to sign a corporate CLA.
Follow either of the two links above to access the appropriate CLA and instructions for how to sign and return it. Once we receive it, we'll be able to accept your pull requests.
The Kubernetes Python client contains mostly files that are generated by the OpenAPI generator from this OpenAPI spec. In the repo there is also the utility part, which allows developers to create their own kubernetes clients (kubernetes/base). The base repo was once a submodule of the main repo, but is now integrated into the main repo. The archived code is available (here).
The following folders are automatically generated. You will need to submit a patch to the upstream Kubernetes repo kubernetes or the OpenAPI generator repo openapi-generator.
In this main repo, the following folders contain developer written codes and the patches should be submitted as pull requests here:
- kubernetes/base
- kubernetes/config
- kubernetes/dynamic
- kubernetes/e2e_test
- kubernetes/leaderelection
- kubernetes/stream
- kubernetes/utils
- kubernetes/watch
- examples
- scripts.
- Submit an issue describing your proposed change to the repo in question.
- The repo owners will respond to your issue promptly.
- If your proposed change is accepted, and you haven't already done so, sign a Contributor License Agreement (see details above).
- Fork the desired repo, develop and test your code changes. Add a test if possible.
- Submit a pull request.
If your patch depends on new packages, add those packages to requirements.txt and/or setup.py. If these package are for testing only, add those to test-requirements.txt.
Generally we would like to see one commit per pull request. However, if the pull request is reasonably large, the PR can be divided into several commits that make logical sense. The commit message should be clear and indicative of the aim of the fix. Sometimes multiple commits in a single pull request is acceptable if it meets the Kubernetes pull request guidelines.
If you have several commits in a pull request and have been asked to squash your commits, please use git reset --soft HEAD~N_COMMITS
and commit again to make your PR a single commit.
The symbolic links contained in this repo do not work for Windows operating systems. If you are a Windows developer, please run the fix inside the scripts folder or manually copy the content of the kubernetes/base folder into the kubernetes folder.
In addition to running the fix yourself and telling us that your fix works, you can demonstrate that your fix really works by using unit tests and end to end tests. Tests are mainly located in three places. You should put your tests into the places that they fit in.
- Generated tests by OpenAPI generator: these tests should pass and do not require modification.
- End to end tests: these are tests that can only be verified with a live kubernetes server.
- Base repo tests in the base repo, in which the test files are named
test_*.py
: These tests use the packageMock
and confirms the functionality of the base repo files.
We use an automatic coding style checker by using the diff
of the autopep8 output and your code file. To make sure that your code passes the coding style checker, run autopep8 --in-place --aggressive --aggressive your_code.py
before committing and submitting.
If you write a new end to end (e2e) test, or change behaviors that affect e2e tests, you should set up a local cluster and test them on your machine. The following steps will help you run the unit tests.
-
Acquire a local cluster. Minikube is a good choice for Windows and Linux developers. Alternatively if you are on Linux, you can clone the kubernetes repo and run install-etcd.sh and then local-up-cluster.sh to get a local cluster up and running.
-
Run the unit tests. In the root directory of the main repo, run
python -m unittest discover
. -
Check the test results and make corresponding fixes.