CI: Make pep8 validation work with multiple files simultaneously

[dontgoto]

The validation is bottlenecked due to the repeated sys calls, subprocess spawning, and pep8 calls that only validate a single file each. Running the validation on a list of files is approximately as fast as running it on a single file. This PR makes the first step, enabling the pep8 validation of a list of files, but for now still running it only on a single file. It will be followed by a few more PRs. The end result will be a complete refactoring of validate_docstrings.py, that will then take about 10 seconds to validate all docstrings.

• All code checks passed.
• Added type annotations to new arguments/methods/functions.

@datapythonista

[datapythonista]

I'm travelling and I could only revew with the phone, but looks good.

Quick question, would it make sense to use ruff instead of flake8 for this?

[dontgoto]

Thanks for the quick review!

Quick question, would it make sense to use ruff instead of flake8 for this?

I think it would be slightly faster, but someone would have to configure ruff to behave the same way as we intend flake8 to behave here. On my machine, the whole code_check.sh docstrings validation finishes in about 10 seconds, 7s of those are separate calls to numpydoc.validate (I put in a PR there that brings this down to 5.5s), and the remaining 3s are split between our validate.pyand pep8.

So I think even with ruff we would at best get to 5.5s instead of 8.5s if we keep numpydoc.validate. Might still be worth an issue to discuss the (future) role of ruff here and in other places.

[datapythonista]

Yep, trying to improve 10s doesn't make sense. I know we are using ruff in the CI nowz bit I didn't check the exact state. The question was more about consistency, I didn't check if we already replaced flake8 with ruff.

[dontgoto]

I see ruff being used as part of the pre-commit hooks. And there it let a docstring error that I introduced in a non-replaced docstring through. It's also replaced flake8 for the docstring formatting, but back when I introduced my error, only the code_check.sh showed up as failed.

[github-actions]

This pull request is stale because it has been open for thirty days with no activity. Please update and respond to this comment if you're still interested in working on this.

[mroeschke]

Thanks for the PR but appears there not much appetite for this change so closing