Skip to content

Functionality of py-template

Please see template/functionality for base functionality, this template adjusts and extends them to be more useful for Python!

Below is a detailed list of functionalities and their source files (so you can modify them if you wish with ease), functionalities provided by template are not repeated here!

See specific use cases for more thorough description of possible adjustments

Local development

In order to have the exact same configuration as is run in pipelines described below do the following:

  1. clone your repository
  2. run poetry install --all-extras

Now you are able to autofix/lint/run tests in the exact same way!

Note: you might have to read the following sections to see commands being run, although most of it is specified via pyproject.toml.

Tooling configuration

All of the easily tunable parameters are located in pyproject.toml including:

  • Tests dependencies/settings
  • Automated fixers dependencies/settings
  • Linters dependencies/settings
  • Code statistics reporting
  • Documentation creation/styling
  • Semantic release

Automated code fixing/linting

After opening a PR your code will be subject to the following:

  • Whatever is possible will be fixed and pushed back to the branch (e.g. black formatting, removing unused imports etc.)
  • Linters are run on formatted version to find errors which we are unable to fix automatically
  • Code statistics are performed and saved as GitHub Job Summaries

Note: Please do a git pull from remote branch before pushing more code to the branch due to automated fixers!

Pipeline is defined in .github/workflows/python-code-checker.yml while settings are defined in pyproject.toml (code comment sections Fixers, Linters and Reporters)

Continuous Integration

We provide a dummy folder tests, where all of your pytests should reside (you might also describe testing procedures via generated ).

After every Pull Request the following pipeline is run:

  • Tests are run across matrix of python and os versions (last three minor versions and ubuntu-latest/macos-latest)
  • Coverage is calculated for every item and uploaded as a HTML artifact (also available within job's stdout)
  • Coverage is combined and uploaded as a HTML artifact (also available within job's stdout)

See .github/actions/python-test/action.yml for core functionality and .github/workflows/python-tests.yml for its usage in a workflow.

Continuous Deployment

After a PR is thoroughly checked (and merged), py-template will take care of deploying your python package to registry of your choice (see Setup section).

Please see .github/workflows/python-release.yml for the pipeline functionality.

Automated documentation generation

Every documented function will automatically be added to API Reference section, which is modeled after your source code. This action is performed after Continuous Deployment ran successfully.

One can see the script responsible for docs creation in docs/ .

In general, simply add code to src/<repo-name> source files and watch it automagically appear via mkdocs.yml

Please note we are also validating whether you have sufficiently documented your code via the following automated action: .github/workflows/python-docstrings-coverage.yml (anything below 100 won't pass, fortunately it is easily tunable :) ).

Caching (additional info, advanced)

We tried our best in order to improve pipelines efficiency as much as possible. In order to do that, we have created a two level caching mechanism:

  • pipx is cached on a per-python, per-os level
  • poetry is cached on a per-python, per-os level

This means initial pass of your actions might be a lot slower (even a few minutes), but afterwards everything will be simply downloaded from cache.

To see how we did it (or modify/improve it), check .github/actions/setup-poetry/action.yml .