keystone/doc/source/contributor/programming-exercises.rst

Programming Exercises for Interns and New Contributors

The keystone team participates in open source internship programs such as Outreachy and Google Summer of Code and welcomes contributions from students and developers of all skill levels. To help with formal applications for work programs or to give casual contributors a taste of what working on keystone is like, we've created a few exercises to showcase what we think are valuable development skills.

These exercises are samples, and code produced to solve them should most likely not be merged into keystone. However, you should still propose them to Gerrit to get practice with the code review system and to get feedback from the team. This is a good way to get used to the development workflow and get acquainted with the benefits of working in a collaborative development environment. Also feel free to talk to the keystone team <../getting-started/community> to get help with these exercises, and refer to the contributor documentation <index> for more context on the architecture and contributing guidelines for keystone.

The exercises provide some ideas of what you can do in keystone, but feel free to get creative.

Add a Parameter to an API

Add a string parameter named nickname to the Project API. The end result will be that you can use the new parameter when you create a new project using the POST /v3/projects API, update the parameter using the PATCH /v3/projects/{project_id} API, and the value displayed using the GET /v3/projects/{project_id}.

Refer to the API Change tutorial <api_change_tutorial>. In short, you will need to follow these steps:

1. Create a SQL migration to add the parameter to the database table (:pykeystone.common.sql.expand_repo.versions, :pykeystone.common.sql.data_migration_repo.versions, :pykeystone.common.sql.contract_repo.versions)
2. Add a SQL migration unit test (keystone/tests/unit/test_sql_upgrade.py)
3. Add the parameter to the SQL model for projects (:pykeystone.resource.backends.sql)
4. Add unit tests (keystone/tests/unit/resource/test_backend.py) for the manager (:pykeystone.resource.core) to show that the project can be created and updated with the new parameter using the provider mechanism
5. Add the parameter to the API schema (:pykeystone.resource.schema)
6. Add an API unit test (keystone/tests/unit/test_v3_resource.py)
7. Document the new parameter in the api-ref

Write an External Driver

Write an external driver named file that implements the Project API. The end result will be that you can set [resource]/driver = file in keystone.conf to have keystone load a list of project names from a text file, and querying keystone for projects will return projects with those names in the default domain.

Refer to the Developing Keystone Drivers <developing-drivers> tutorial. Your driver can start as an in-tree driver: create a class named Resource in keystone/resource/backends/file.py that implements :pykeystone.resource.backends.base.Resource. Once you have that working, break it out into a separate repository and create a Setuptools entrypoint to allow you to register it with keystone.

Write an Auth Plugin

Write an auth plugin named hacker that allows any existing user to authenticate if they provide a valid username and the password "hax0r". The end result will be that you can add hacker as an auth method in [auth]/methods in keystone.conf, and users will be able to get an unscoped token <../admin/tokens> using POST /v3/auth/tokens and providing "hacker" as the auth method, a valid username as the username, and "hax0r" as the password.

Refer to the auth-plugins documentation. You should create a class Hacker in keystone/auth/plugins/hacker.py that implements :pykeystone.auth.plugins.base.AuthMethodHandler. For bonus points, also add the plugin to keystoneauth so that Python clients can also use this auth method.