Skip to content

Latest commit

 

History

History
69 lines (43 loc) · 3.47 KB

DEVELOPMENT.md

File metadata and controls

69 lines (43 loc) · 3.47 KB
Raw

Development

Thanks for contributing to Stytch's Python library! If you run into trouble, find us in Slack.

Setup

  1. Install Python version 3.8 or greater.
  2. (Recommended, but optional) Create a new virtualenv.
  3. pip install -r requirements_dev.txt

Generating (and regenerating) the API

If you create, update, or delete an API method, you should rerun the codegen to keep this module up-to-date. There is a script at bin/generate-api.sh that will run the default codegen with sane options (applying default formatters, not overwriting existing manual implementation methods). This is the recommended way to allow the client to hit a new endpoint.

Adding a new manual implementation method

If you are adding a new manual method, first check if you need to add any new additional_imports within the spec file. Add your input and create the stub for your manual method. Run bin/generate-api.sh. The script will generate the new method stub, but your import may be missing -- this is because the auto-formatter determined that the method was not used. You can now write your manual method, and add a test in api/test/

You should be able to call bin/generate.sh again to ensure you've added all necessary imports. Verify the correctness of your changes by remembering to [run the unit tests][#testing].

Testing

Run the tests with the unittest module. To test everything, run python -m unittest from the project root directory.

Run the core unit tests with python -m unittest discover -s stytch/ -t ..

Run the codegen unit tests with python -m unittest discover -s codegen/ -t .

Run the integration tests with python -m unittest discover -s test/ -t .

The integration tests call the test API of Stytch to ensure you can call all methods. It requires setting up the following environment variables:

  • STYTCH_PYTHON_RUN_INTEGRATION_TESTS=1
  • STYTCH_PROJECT_ID=your-stytch-project-id
  • STYTCH_SECRET=your-stytch-secret

You can check test coverage with the following command:

coverage run -m unittest  && coverage report --omit="*test/*" --sort=cover

Note that if you didn't enable the integration tests that coverage will be much lower. If you have your STYTCH_PROJECT_ID and STYTCH_SECRET in your env, this command may be useful:

env STYTCH_PYTHON_RUN_INTEGRATION_TESTS=1 coverage run -m unittest && coverage html --omit="*test/*"

The above is also available as a convenience script at bin/generate-coverage.sh

You can run that and then see the exact coverage in a useful HTML page with line-by-line coverage.

Issues and Pull Requests

Please file issues in this repo. We don't have an issue template yet, but for now, say whatever you think is important!

If you have non-trivial changes you'd like us to incorporate, please open an issue first so we can discuss the changes before starting on a pull request. (It's fine to start with the PR for a typo or simple bug.) If we think the changes align with the direction of the project, we'll either ask you to open the PR or assign someone on the Stytch team to make the changes.

When you're ready for someone to look at your issue or PR, assign @stytchauth/client-libraries (GitHub should do this automatically). If we don't acknowledge it within one business day, please escalate it by tagging @stytchauth/engineering in a comment or letting us know in Slack.