Make HashSet::replace work without raw table internals
#681
+158
−26
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
clarfonthey
commented
Jan 27, 2026
Comment on lines
+3800
to
+3801
| /// assert!(std::ptr::eq(replaced, old_key)); | ||
| /// assert!(std::ptr::eq(*entry.key(), new_key)); |
Contributor
Author
There was a problem hiding this comment.
This seemed like the best way to ensure the key was truly "replaced" instead of having a custom struct with a custom PartialEq implementation.
clarfonthey
force-pushed
the
replace-key
branch
from
49542db to
c232221
Compare
clarfonthey
force-pushed
the
replace-key
branch
from
c232221 to
8db923f
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The primary addition here is
OccupiedEntry::replace_key, which allows replacing the key in the map. Since this is an uncommon operation, I decided to make itunsafeand not check for the equivalence of the key, but that can potentially be changed.The other change that is made here is
VacantEntryRef::{insert_with_key,insert_entry_with_key}now haveuncheckedversions which do not check the equivalence of the key, to ensure the same behaviour with thereplacefunction never panicking.Unfortunately, an
UnsafeCellis required to truly make this API work without accessing internals, but it shouldn't ultimately matter, I think. The result is thefind_or_find_insert_indexmethod ofHashMapis no longer exported, and instead the entry API is used byHashSet. Now, the only remaining raw table API forHashSetisRawExtractIf, to be fixed later.