No std hashing

[Person-93]

Closes #6329

I've added a new trait in impl_: PyHashable. I used ahash::randome_state::RandomState as the build hasher but that can be replaced by any BuildHasher that can be constructed at compile time.

[codspeed-hq]

Merging this PR will not alter performance

✅ 141 untouched benchmarks
⏩ 6 skipped benchmarks¹

---

_{Comparing Person-93:no_std_hashing (cbdfc5b) with main (98cf22d)}

Footnotes

1. 6 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[Person-93]

I'm having some trouble with the UI tests. They are still failing even after running it with UI_TEST=bless.

[davidhewitt]

Thanks for iterating through the ways forward here. I'm not entirely sure about the consequences of adding the ahash dependency or switching the hashing to it.

Would an acceptable alternative be to make the #[pyclass(hash)] option require the std feature for now? We might yet proceed with this option later, but deferring by gating the functionality might be good enough to unblock you?

[Person-93]

Would an acceptable alternative be to make the #[pyclass(hash)] option require the std feature for now? We might yet proceed with this option later, but deferring by gating the functionality might be good enough to unblock you?

Yes, this would be a good way to unblock this for now.

[Person-93]

I'm not entirely sure about the consequences of adding the ahash dependency or switching the hashing to it.

I haven't really thought it through thoroughly either, I just wanted to see if something like this could be done so that a properly thought out choice could be implemented later.