[ruff] Add support for more re patterns (RUF055)

[Garrett-R]

Summary

Implements some of #14738, by adding support for 6 new patterns:

re.search("abc", s) is None       # ⇒ "abc" not in s
re.search("abc", s) is not None   # ⇒ "abc" in s

re.match("abc", s) is None       # ⇒ not s.startswith("abc")  
re.match("abc", s) is not None   # ⇒ s.startswith("abc")

re.fullmatch("abc", s) is None       # ⇒ s != "abc"
re.fullmatch("abc", s) is not None   # ⇒ s == "abc"

Test Plan

cargo nextest run
cargo insta review

And ran the fix on my startup's repo.

Note

One minor limitation here:

if not re.match('abc', s) is None:
    pass

will get fixed to this (technically correct, just not nice):

if not not s.startswith('abc'):
    pass

This seems fine given that Ruff has this covered: the initial code should be caught by E714 and the fixed code should be caught by SIM208.

[github-actions]

ruff-ecosystem results

Linter (stable)

✅ ecosystem check detected no linter changes.

Linter (preview)

ℹ️ ecosystem check detected linter changes. (+2 -0 violations, +2 -0 fixes in 2 projects; 53 projects unchanged)

ibis-project/ibis (+2 -0 violations, +0 -0 fixes)

ruff check --no-cache --exit-zero --ignore RUF9 --no-fix --output-format concise --preview

+ ibis/backends/tests/test_client.py:1080:9: RUF055 [*] Plain string pattern passed to `re` function
+ ibis/backends/tests/test_client.py:1101:9: RUF055 [*] Plain string pattern passed to `re` function

astropy/astropy (+0 -0 violations, +2 -0 fixes)

ruff check --no-cache --exit-zero --ignore RUF9 --no-fix --output-format concise --preview

- astropy/coordinates/tests/test_frames.py:352:11: RUF055 Plain string pattern passed to `re` function
+ astropy/coordinates/tests/test_frames.py:352:11: RUF055 [*] Plain string pattern passed to `re` function

Changes by rule (1 rules affected)

code	total	+ violation	- violation	+ fix	- fix
RUF055	4	2	0	2	0

[ntBre]

Thanks for doing this! I suggested some changes, but I think the overall approach is spot-on.

[Garrett-R]

Thanks for the thorough review! I've applied all suggestions in this commit: 92e84ee

Then I came up with solution to the unsafe fix issue you found in this commit: 3d25918

[ntBre]

Thanks, this looks great! Just a couple more much smaller changes, and I think this is good to merge.

[Garrett-R]

Thanks, good ideas! Implemented all those 😎

[ntBre]

This looks great. Thanks again for your work on this!

[ntBre]

Oh and just to acknowledge the not not case, I agree that it's unfortunate, but your observation that the original code should be caught by E714, which I believe is enabled by default, is good enough for me. I think we'd have to look at the grandparent expression to detect this anyway, and I think it's reasonable to draw the line at the first parent, especially in light of the other rules.

[Garrett-R]

For sure, thanks for the super helpful review, learned a bunch! 😎