eq-without-hash (PLW1641) does not check for __hash__ in superclasses

[DetachHead]

i noticed this is mentioned in the rule's docs but i'm raising it because i couldn't see an issue for it

[charliermarsh]

👍 Yeah this is why the rule is in the nursery -- as long as we don't support multi-file analysis, we can't implement it reliably. I'm gonna close as this is basically a known issue that will eventually be solved by multi-file analysis.

[DetachHead]

wouldn't it make sense to keep the issue open to track it if it's a known issue?

[charliermarsh]

We could consider creating a central tracking issue (or a tag like @zanieb did here). I just want to avoid piling up a bunch of known issues related to multi-file analysis that can't be worked on individually and make it harder to navigate the issue tracker.

[LefterisJP]

We could consider creating a central tracking issue (or a tag like @zanieb did here). I just want to avoid piling up a bunch of known issues related to multi-file analysis that can't be worked on individually and make it harder to navigate the issue tracker.

Also hit this one while enabling --preview. It's a useful rule. It found out 1 place we were missing __hash__ but then had 6 false positives due to superclasses.

Added # noqa: PLW1641 # hash implemented by superclass at the false positives for now and keeping the rule.

[DetachHead]

We could consider creating a central tracking issue

raised #7447

[KotlinIsland]

• see (📚) docs for PLW1641 known problems doesn't make sense #14883, it shouldn't look in superclasses