Fix false positive for superfluous-parens for return (a or b) in iterable

[jacobtylerwalls]

Type of Changes

	Type
✓	🐛 Bug fix

Description

Fix false positive for superfluous-parens for patterns like return (a or b) in iterable.

Closes #5803

[coveralls]

Pull Request Test Coverage Report for Build 2044359428

• 5 of 5 (100.0%) changed or added relevant lines in 1 file are covered.
• 103 unchanged lines in 8 files lost coverage.
• Overall coverage increased (+0.01%) to 94.184%

---

Files with Coverage Reduction	New Missed Lines	%
pylint/reporters/multi_reporter.py	1	98.21%
pylint/reporters/reports_handler_mix_in.py	2	94.74%
pylint/extensions/typing.py	4	97.08%
pylint/reporters/ureports/nodes.py	4	94.52%
pylint/utils/utils.py	13	86.9%
pylint/checkers/refactoring/refactoring_checker.py	15	98.25%
pylint/checkers/utils.py	29	95.97%
pylint/checkers/variables.py	35	96.49%

Totals
Change from base Build 2039774678:	0.01%
Covered Lines:	15336
Relevant Lines:	16283

---

💛 - Coveralls

[jacobtylerwalls]

@DanielNoord this is here to preserve the special case added in #4948, but when I looked into it, I had trouble figuring out why it was deemed a bugfix. (Looks like unnecessary parentheses to me!) Do you recall the thought behind it?

[DanielNoord]

Does this still pass:
D = [x for x in ((3, 4) if 1 > 0 else ((5, 6)))]

See #4907 (comment)

I think sometimes people want to create inner tuples and this was added so we don't flag those.

[jacobtylerwalls]

Yes, it still passes now, but I'm essentially suggesting we later revert #4948.

I think sometimes people want to create inner tuples and this was added so we don't flag those.

I'm not sure what inner tuple means here. ((5,6),) is a tuple of a tuple. ((5,6)) is superfluous parentheses, so it could just be disabled if people really prefer the extra parentheses. But I think maybe there's a nuance here I'm not getting?

[jacobtylerwalls]

Does this still pass:
D = [x for x in ((3, 4) if 1 > 0 else ((5, 6)))]

(This test is still in the suite)

[DanielNoord]

Hmm perhaps I was wrong to re-allow those.

I thought there might be use cases where you would want to create Tuple[Tuple[Tuple[int, ...]]] instead of Tuple[int, ...]?

[jacobtylerwalls]

Sure, but that's (((5,),),) rather than (((5,))).

Parentheses increase readability, so I understand why people use them, belt and suspenders and all, but I'm just saying those folks might not want the check enabled if they're in the practice of adding them.

[DanielNoord]

Hm, I'm fine with reverting then. But we should check #4907 for any legitimate use cases.

[jacobtylerwalls]

I'll capture this discussion on #5361

[Pierre-Sassoulas]

LGTM, I'm thinking the changelog should be in 2.13.1 in ChangeLog and in 2.14 in whatsnew if we backport this but let me know what you think. (edit: resolved in another discussion)

[DanielNoord]

Does this still pass:
D = [x for x in ((3, 4) if 1 > 0 else ((5, 6)))]

See #4907 (comment)

I think sometimes people want to create inner tuples and this was added so we don't flag those.

[DanielNoord]

@jacobtylerwalls Do we understand the coverage report?

[jacobtylerwalls]

@jacobtylerwalls Do we understand the coverage report?

From the edit history of the comment the lines lost coverage when I merged main 10 hours ago, so ... yes? :D (I don't really have a mental model of how Coveralls handles merges into PR branches.)