RUF027 no longer has false negatives with string literals inside of method calls

[snowsignal]

Fixes #9857.

Summary

Statements like logging.info("Today it is: {day}") will no longer be ignored by RUF027. As before, statements like "Today it is: {day}".format(day="Tuesday") will continue to be ignored.

Test Plan

The snapshot tests were expanded to include new cases. Additionally, the snapshot tests have been split in two to separate positive cases from negative cases.

[github-actions]

ruff-ecosystem results

Linter (stable)

✅ ecosystem check detected no linter changes.

Linter (preview)

ℹ️ ecosystem check detected linter changes. (+8 -0 violations, +0 -0 fixes in 3 projects; 40 projects unchanged)

RasaHQ/rasa (+4 -0 violations, +0 -0 fixes)

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

+ tests/core/nlg/test_response.py:372:9: RUF027 Possible f-string without an `f` prefix
+ tests/shared/core/test_slot_mappings.py:72:9: RUF027 Possible f-string without an `f` prefix
+ tests/shared/core/training_data/story_writer/test_yaml_story_writer.py:201:9: RUF027 Possible f-string without an `f` prefix
+ tests/test_model_training.py:922:9: RUF027 Possible f-string without an `f` prefix

ibis-project/ibis (+1 -0 violations, +0 -0 fixes)

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

+ ibis/expr/types/generic.py:752:37: RUF027 Possible f-string without an `f` prefix

rotki/rotki (+3 -0 violations, +0 -0 fixes)

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

+ rotkehlchen/db/dbhandler.py:2800:23: RUF027 Possible f-string without an `f` prefix
+ rotkehlchen/exchanges/bitpanda.py:275:47: RUF027 Possible f-string without an `f` prefix
+ rotkehlchen/exchanges/kraken.py:419:21: RUF027 Possible f-string without an `f` prefix

Changes by rule (1 rules affected)

code	total	+ violation	- violation	+ fix	- fix
RUF027	8	8	0	0	0

[charliermarsh]

The Rasa checks look like false positives (hard to avoid...), the Ibis check looks like a true positive (I think), the Pandas check looks like a false positive (maybe avoidable?), the Rotki checks look like they might be true positives.

[charliermarsh]

Perhaps we could avoid the Pandas false positive by also ignoring calls with positional arguments that use the exact same name...

[snowsignal]

@charliermarsh I added some logic to ignore literals with matching regular arguments in addition to keyword arguments.

[codspeed-hq]

CodSpeed Performance Report

Merging #9865 will improve performances by 11.53%

_{Comparing jane/linter/RUF027/method-call-fix (787ae4e) with main (daae28e)}

Summary

⚡ 10 improvements
✅ 20 untouched benchmarks

Benchmarks breakdown

	Benchmark	main	jane/linter/RUF027/method-call-fix	Change
⚡	linter/all-rules[numpy/ctypeslib.py]	20.6 ms	18.9 ms	+9.19%
⚡	linter/all-rules[large/dataset.py]	92.5 ms	82.9 ms	+11.53%
⚡	linter/all-with-preview-rules[numpy/globals.py]	3.1 ms	2.9 ms	+5.7%
⚡	linter/all-with-preview-rules[numpy/ctypeslib.py]	23.3 ms	21.5 ms	+8.31%
⚡	linter/all-with-preview-rules[pydantic/types.py]	51 ms	47 ms	+8.66%
⚡	linter/all-with-preview-rules[large/dataset.py]	105.1 ms	95.3 ms	+10.23%
⚡	linter/all-with-preview-rules[unicode/pypinyin.py]	11.8 ms	11.2 ms	+5.21%
⚡	linter/all-rules[numpy/globals.py]	2.8 ms	2.6 ms	+6.35%
⚡	linter/all-rules[unicode/pypinyin.py]	10.8 ms	10.2 ms	+5.8%
⚡	linter/all-rules[pydantic/types.py]	43.9 ms	39.6 ms	+10.71%