Replaces pylint with ruff #7438
Conversation
Ruff should be much faster so no need to have this complex script anymore
| 'pep8', | ||
| # spellcheck introduced in version 1.4.0 | ||
| 'pylint>=1.4.0', | ||
| 'pyenchant', |
There was a problem hiding this comment.
We are loosing spell checking here. I can bring in https://crates.io/crates/typos if desired.
|
As a follow up pull request, I would investigate all those pylint ignore comments. Some of them can be dropped now. |
| else: | ||
| # Split the string on whitespace to allow passing options in |
There was a problem hiding this comment.
I would prefer to disable warning that requires these changes. Often it's much more readable to have full else branch and a if statement there. If the main if and the else if checks are unrelated, this visual separation is a good way to hint to a logical separation.
There was a problem hiding this comment.
Ok I revisit these changes after #7436
| @@ -163,13 +163,13 @@ def test_compare(self): | |||
| b2 = self.makeService(attach_to_master=False, name='b', serviceid=21) # same args as 'b1' | |||
| b3 = self.makeService(attach_to_master=False, name='b', serviceid=20) # same id as 'a' | |||
|
|
|||
| self.assertTrue(a == a) # pylint: disable=comparison-with-itself | |||
| self.assertTrue(a == a) # noqa: PLR0124 | |||
There was a problem hiding this comment.
Is it possible to configure ruff to allow full name of the warning in places like here? The warning ID is completely nondescriptive and thus is blocker for retiring pylint.
There was a problem hiding this comment.
Turns out like this is not implemented in ruff. But it will be there eventually.
There was a problem hiding this comment.
I think it makes sense to have this PR in before ruff supports human-friendly rule names (astral-sh/ruff#1773). The speed benefits outweigh small additional confusion which will be addressed in the future anyway.
There was a problem hiding this comment.
I find those short labels easy to lookup in the ruff documentation: https://docs.astral.sh/ruff/rules/
Them being shorter actually makes them easier to copy.
There was a problem hiding this comment.
Agreed, but when reading code it's completely unclear what these numeric codes mean. Usually code is read far more times than it is written.
In any case, this PR can go after merge conflicts are resolved (do we need to rebase and just drop some commits?). We can revisit the rule names later.
Depends on #7436
Contributor Checklist:
newsfragmentsdirectory (and read theREADME.txtin that directory)