Closed agramfort closed 7 months ago
Attention: Patch coverage is 99.25000%
with 3 lines
in your changes are missing coverage. Please review.
Project coverage is 97.62%. Comparing base (
0deed68
) to head (bb56347
).
Files | Patch % | Lines |
---|---|---|
skada/datasets/_base.py | 80.00% | 2 Missing :warning: |
skada/datasets/_samples_generator.py | 96.42% | 1 Missing :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
@agramfort thanks for the PR, tests are not running anymore since last commit. https://github.com/scikit-adaptation/skada/actions/runs/8133112804
Could we have a quick test checking that style is valid (flake8 was not needed anymore but can we test it with ruff maybe?). To detect new contributors that did not install the pre-commit.
@rflamary @kachayev @tgnassou this one is good to go from my end
@agramfort This is cool, very needed tools! I have a question regarding testing/linting flow: testing job is setup to run pre-commit hooks, which contains ruff linter with --fix
flag. What happens if someone commits unlinted file? I guess it will just fix the code locally in the test container without reporting/failing? 🤔
@kachayev good question. I copied the practice of another repo that has been using this successfully for a few months. Let's see how it goes here. We can always adjust.
well now we know it does not fail since my PR definitely does not respect ruff: https://github.com/scikit-adaptation/skada/pull/135
That's interesting. In my PRs it actually fails because of files being changed (which is desirable) 🤔
ruff lint skada..........................................................Failed
- hook id: ruff
- files were modified by this hook
this is very weird no?
setting up precommit hooks with ruff + codespell
codespell found a few typos :)
feedback welcome