Apply NFKC normalization to unicode identifiers when storing bindings in the semantic model

[AlexWaygood]

Summary

Fixes #5003.

Python applies NFKC normalization to identifiers that use unicode characters. That means that F821 should not be emitted if ruff encounters the following snippet (but on main, it is), as from Python's perspective, these are all the same identifier:

𝒞 = 500
print(𝒞)
print(C + 𝒞)  # ruff says `C` isn't defined
print(C / 𝒞)
print(C == 𝑪 == 𝒞 == 𝓒 == 𝕮)  # ruff says `C`, `𝑪`, ... isn't defined

I fixed this false positive by changing the bindings field in ruff_python_semantic/scope.rs so that identifiers are always unicode-normalized according to NFKC before being stored in the hashmap. An alternative approach I played around with was to unicode-normalize identifiers in the AST itself. However, this would have had undesirable consequences: the formatter would have started eagerly normalizing unicode characters in identifiers when it reformatted a Python file.

Test Plan

cargo test

[codspeed-hq]

CodSpeed Performance Report

Merging #10381 will degrade performances by 5.17%

_{Comparing AlexWaygood:unicode-normalize-identifiers (0385a62) with main (e944c16)}

Summary

❌ 1 regressions
✅ 29 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Benchmarks breakdown

	Benchmark	main	AlexWaygood:unicode-normalize-identifiers	Change
❌	linter/default-rules[numpy/globals.py]	635 µs	669.6 µs	-5.17%

[AlexWaygood]

Ouch. I guess I need to start running benchmarks before filing PRs :)

[github-actions]

ruff-ecosystem results

Linter (stable)

✅ ecosystem check detected no linter changes.

Linter (preview)

✅ ecosystem check detected no linter changes.

Formatter (stable)

✅ ecosystem check detected no format changes.

Formatter (preview)

✅ ecosystem check detected no format changes.

[AlexWaygood]

Closing in favour of #10412