[red-knot] Refactor KnownFunction::takes_expression_arguments()

[AlexWaygood]

Summary

Refactors the KnownFunction::takes_expression_arguments(). By introducing a custom type rather than passing around a u32 everywhere, we improve readability, the documentation of the code, and type safety, without introducing any allocation.

Test Plan

cargo test -p red_knot_python_semantic

[github-actions]

ruff-ecosystem results

Linter (stable)

✅ ecosystem check detected no linter changes.

Linter (preview)

✅ ecosystem check detected no linter changes.

[MichaReiser]

Nice. Using bitset! over a hand written bitset would probably have been an alternative solution.

[MichaReiser]

I don't really have a better suggestion but the name is somewhat vague. ParameterKindness? ParameterValueKind?

[AlexWaygood]

I agree it's not a great name but I can't think of anything better, and I'm not sure I love either of these 😆

[AlexWaygood]

Nice. Using bitset! over a hand written bitset would probably have been an alternative solution.

Yeah -- I considered it, but ultimately concluded it wouldn't buy us much here, since there's only a limited number of special-cased functions anyway

[carljm]

Pretty sure this will use more memory than a bitset (I don't think Rust will pack a slice of one-bit enums into a bitset automatically?) but yeah, it shouldn't matter.

[MichaReiser]

Pretty sure this will use more memory than a bitset (I don't think Rust will pack a slice of one-bit enums into a bitset automatically?) but yeah, it shouldn't matter.

Yeah, it should be 2 words but we never store the data structure anywhere so...

[AlexWaygood]

Pretty sure this will use more memory than a bitset (I don't think Rust will pack a slice of one-bit enums into a bitset automatically?) but yeah, it shouldn't matter.

I also agree that I don't think it matters, but if it makes you happier I've switched to a representation where it's now smaller than on main (1 byte instead of 4 :-)