Continuous Integration Pre-Commit

What is it? Why do it?

Continuous integration Pre-Commit is the practice of running a set of hooks that enforce code styling consistency whenever new changes are proposed.

pre-commit hooks should be setup locally, so one can apply the required fixes before committing. That is an assumption not all developers comply with, and therefore, it’s useful to have a GitHub workflow to help enforcing code style and keeping it consistent during development.

Using pre-commit.ci lite, we incorporate a bot into our project that performs as many automatic fixes as possible and reports any code style issues that it could not resolve but should be fixed.

For each pull request, our pre-commit.ci workflow will:

• Clear outputs from Jupyter notebooks

• Analyze the src code style and report code that doesn’t adhere, using the options you selected for your tooling set:

  • “ruff linting”: checks for linting rules, and import order

  • “ruff auto-formatter”: auto-formats code

  • “pylint”: checks for compliance with pylint rules

  • “black”: auto-formats code (including notebooks)

  • “isort”: Sort imports using isort

• Analyze type hints (if mypy type checking is enabled)

Note

• Only a small subset of hooks declared in the .pre-commit-config.yaml file is relevant for the CI pipeline. Some checks are simply not relevant, while others are performed in other github workflows instead.

  For more information on the list of available hooks visit Pre-Commit.

How to activate

The template will automatically generate the pre-commit-ci.yml workflow file. To finish the configuration of this feature and activate the automatic fixes on pull requests, you need to install the pre-commit application for your relevant GitHub repository.