Scikit-HEP repo-review

This tool can check the style of a repository. Use like this:

pipx run 'scikit-hep-repo-review[cli]' <path to repository>

This will produce a list of results - green checkmarks mean this rule is followed, red x’s mean the rule is not. A yellow warning sign means that the check was skipped because a previous required check failed. Some checks will fail, that’s okay - the goal is bring all possible issues to your attention, not to force compliance with arbitrary checks. Eventually there might be a way to mark checks as ignored.

For example, GH101 expects all your action files to have a nice name: field. If you are happy with the file-based names you see in CI, you should feel free to simply ignore this check (just visually ignore it for the moment, a way to specify ignored checks will likely be added eventually).

All checks are mentioned at least in some way in the Scikit-HEP Developer Guidelines. You should read that first - if you are not attempting to follow them, some of the checks might not work. For example, the guidelines specify pytest configuration be placed in pyproject.toml. If you place it somewhere else, then all the pytest checks will be skipped.

You are not required to be in Scikit-HEP to find this useful, however - examples of repositories that at least partially follow the guidelines include pypa/cibuildwheel, pypa/build, and pybind/pybind11.

Development

This repository is intended to be fun to develop - it requires and uses Python 3.10, and makes a few potentially... uncommon decisions. You might not want to design your library this way, but this is not a library. It's a fun and enjoyable app.

There are a few key designs that are very useful and make this possible. First, all paths are handled as Traversables. This allows a simple Traversable implementation based on open_url to provide a web interface for use in the webapp. It also would allow zipfile.Path to work just as well, too - no need to extract.

Checks can request fixtures (like pytest) as arguments. Check files can add new fixtures as needed. The end of every check file has a list of checks (usually pulled from subclasses of a common ancestor) and a list of fixtures it provides.

Check files do not depend on the main library, and can be extended (similar to Flake8). You register new check files via entry-points - so extending this is with custom checks or custom fixtures is easy and trivial. There's no need to subclass or do anything with the base library - no dependency required.

Checks are as simple as possible so they are easy to write. A check is a class with the name (1-2 letters + number) and a docstring (the check message). It should define a set of requires with any checks it depends on (by name), and have a check classmethod. The docstring of this method is the failure message, and supports substitution. Arguments to this method are fixtures, and package is the built-in one providing the package directory as a Traversable. Any other fixtures are available by name. A new fixture is given the package Traversable, and can produce anything (recommended to be cached via functools.cache).

The runner will topologically sort the checks, and checks that do not run will get a None result and the check classmethod will not run. The front-end (Rich powered CLI or Pyodide webapp) will render the markdown-formatted check docstring only if the result is False.

Links

This project inspired Try-PyHF, an interface for a High Energy Physics package in Scikit-HEP.

This project inspired abSENSE, an web interface to abSENSE.