Add methods exposing keys of BTreeMap and HashMap entries

[dstu]

As discussed in rfcs, it would make sense to add a keys() method to the Entry variants for both map types.

PR 32248 has code and tests.

[frewsxcv]

Was this completed in #32248?

[apasel422]

@frewsxcv Yes. We'll keep it open as the tracking issue.

[sgrif]

Had a use case for this in Diesel this morning. (Code comparison: https://gist.github.com/sgrif/68cbba9a87084da3a2e8c42a12e7d6cf). Worked great, and was exactly the missing piece. Looking forward to seeing this stabilized.

[sfackler]

I just noticed that while this is on both OccupiedEntry and VacantEntry, the convenience impl on Entry is missing.

[seanmonstar]

If I wanted to propose a fn take(self) -> (K, V), would it be fine to tag along with this issue, or should that require its own RFC?

[sfackler]

I think it'd be more associated with the key recovery tracking issue, but that's already landed. In any case, I don't think an RFC would be required - it can piggy back off of the existing key recovery one.

[sfackler]

cc @aturon

[alexcrichton]

🔔 This issue is now entering its cycle-long final comment period for stabilization 🔔

(including #33148)