gh-148729: use memchr in SRE prefix scan#148733
Open
haampie wants to merge 2 commits intopython:mainfrom
Open
Conversation
eendebakpt
reviewed
Apr 18, 2026
eendebakpt
reviewed
Apr 18, 2026
For single byte characters use `memchr` instead of the equivalent hand-written while loop. This ensures that `re.search` is typically vectorized through libc for regexes starting with a `LITERAL`. In the no-match case this means 16 or 32 bytes per iterations instead of a single byte (ok, it was unrolled, but not auto-vectorized). Signed-off-by: Harmen Stoppels <harmenstoppels@gmail.com>
haampie
force-pushed
the
hs/feature/re-search-memchr
branch
from
f896a82 to
e488932
Compare
eendebakpt
approved these changes
Apr 18, 2026
Member
There was a problem hiding this comment.
This should be in Library.
Signed-off-by: Harmen Stoppels <harmenstoppels@gmail.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
For USC1 strings use
memchrinstead of the equivalenthand-written while loop.
This ensures that
re.searchis typically vectorized through libc forregexes starting with a
LITERAL. In the no-match case this means 16 or32 bytes per iteration instead of a single byte (with PGO it does unroll,
but that seems to be all).
memchrin SRE's prefix scanner #148729