Black linting

[zerothi]

I have long wanted to do some kind of linting.
Playing round turns out that there is no tool that really suits every wish.

I think it would be beneficial for sisl to use a linter to reduce the pain of going through code styles which is somewhat subjective, and also turns out to be too time-consuming.

With this PR we enforce the black coding style, whether we like it or not.

Comments would be greatly appreciated.

• Closes Using black code formatting style #413
• Tests added
• Ranned isort . at top--level
• Documentation for functionality, docs/
• Changes documented, CHANGELOG.md

[codecov]

Codecov Report

Attention: 354 lines in your changes are missing coverage. Please review.

Comparison is base (89dcede) 87.71% compared to head (80edad5) 87.73%.
Report is 7 commits behind head on main.

❗ Current head 80edad5 differs from pull request most recent head d79400d. Consider uploading reports for the commit d79400d to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #639      +/-   ##
==========================================
+ Coverage   87.71%   87.73%   +0.01%     
==========================================
  Files         362      362              
  Lines       48354    48436      +82     
==========================================
+ Hits        42415    42496      +81     
- Misses       5939     5940       +1     

Files	Coverage Δ
src/sisl/__init__.py	98.00% <100.00%> (ø)
src/sisl/_array.py	100.00% <100.00%> (ø)
src/sisl/_common.py	100.00% <ø> (ø)
src/sisl/_environ.py	87.17% <100.00%> (ø)
src/sisl/_internal.py	100.00% <ø> (ø)
src/sisl/_typing.py	100.00% <100.00%> (+100.00%)	⬆️
src/sisl/_typing_ext/numpy.py	50.00% <100.00%> (ø)
src/sisl/geom/_common.py	100.00% <100.00%> (ø)
src/sisl/geom/basic.py	97.10% <100.00%> (ø)
src/sisl/geom/bilayer.py	98.68% <100.00%> (+0.01%)	⬆️
... and 236 more

... and 65 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[github-advanced-security]

CodeQL found more than 10 potential problems in the proposed changes. Check the Files changed tab for more details.

[pfebrer]

Could you add this and isort to the PR CI? So that you don't need to run them locally and therefore we don't forget :)

[pfebrer]

Maybe using pre-commit? https://pre-commit.com/#intro

[pfebrer]

And actually if pre-commit is installed locally it's a great tool because it will run everything automatically when you do git commit.

[tfrederiksen]

Looks good to me!

[zerothi]

I have tried to make the test depend on the linter, this seems to not work since it is not part of the main branch.
https://docs.github.com/en/actions/using-workflows/events-that-trigger-workflows#workflow_run

[pfebrer]

I'm just checking pre-commit for another package and it works great with black, and probably isort too.

All you have to do is to have a .pre-commit-config.yaml file on the root folder which looks like:

# See https://pre-commit.com for more information
# See https://pre-commit.com/hooks.html for more hooks
repos:
-   repo: https://github.com/pre-commit/pre-commit-hooks
    rev: v4.5.0
    hooks:
    -   id: check-yaml
    -   id: end-of-file-fixer
    -   id: trailing-whitespace
# Using this mirror lets us use mypyc-compiled black, which is about 2x faster
- repo: https://github.com/psf/black-pre-commit-mirror
  rev: 23.10.1
  hooks:
    - id: black-jupyter

Then run pip install pre-commit and pre-commit install. Then each time you run git commit everything runs and the fixes are added to the commit.

It is much better than handling each fixer manually, give it a try :)

Then the checkbox can simply be removed, and run pre-commit on the CI to check that everything is fine, optionally automatically commiting the fixes.

[zerothi]

I'll merge and continue :)

I'm just checking pre-commit for another package and it works great with black, and probably isort too.

All you have to do is to have a .pre-commit-config.yaml file on the root folder which looks like:

# See https://pre-commit.com for more information
# See https://pre-commit.com/hooks.html for more hooks
repos:
-   repo: https://github.com/pre-commit/pre-commit-hooks
    rev: v4.5.0
    hooks:
    -   id: check-yaml
    -   id: end-of-file-fixer
    -   id: trailing-whitespace
# Using this mirror lets us use mypyc-compiled black, which is about 2x faster
- repo: https://github.com/psf/black-pre-commit-mirror
  rev: 23.10.1
  hooks:
    - id: black-jupyter

Then run pip install pre-commit and pre-commit install. Then each time you run git commit everything runs and the fixes are added to the commit.

It is much better than handling each fixer manually, give it a try :)

Then the checkbox can simply be removed, and run pre-commit on the CI to check that everything is fine, optionally automatically commiting the fixes.

I think this is a great advice, if anything we could encourage users to do it, but since this package is still in its infancy int terms of adoption, I would be afraid that it scares people away. If people find this useful, then use it locally, it will help us, but it shouldn't be a barrier for new devs :)

[pfebrer]

I think the separate fixers is more of a barrier than pre-commit, with the checkbox "Ran X, Y, ... at top level" being scary.

And the point is that in the end pre-commit should run on CI to make sure everything is fine, so the devs are not really required to use it.

[zerothi]

I think the separate fixers is more of a barrier than pre-commit, with the checkbox "Ran X, Y, ... at top level" being scary.

And the point is that in the end pre-commit should run on CI to make sure everything is fine, so the devs are not really required to use it.

Ah, you want that, that could also be a way out.

[pfebrer]

I think this is a great advice, if anything we could encourage users to do it

By the way having both approaches is possible. You just have the .pre-commit-config.yaml file in the root of the repo in case anyone wants to do it with pre-commit, but it can still be done manually.