markers: replace regexp with scanning in StripMarkers and EscapeMarkers #37
Merged
dhartunian merged 2 commits intomasterfrom Mar 13, 2026
Merged
markers: replace regexp with scanning in StripMarkers and EscapeMarkers #37dhartunian merged 2 commits intomasterfrom
dhartunian merged 2 commits intomasterfrom
Conversation
Add exhaustive table-driven tests for StripMarkers (both string and byte variants) and EscapeMarkers covering edge cases including empty inputs, unicode content, consecutive markers, hash prefix markers, and boundary conditions. Add benchmarks to establish a performance baseline before optimizing these functions. Co-Authored-By: roachdev-claude <roachdev-claude-bot@cockroachlabs.com>
Contributor
|
The performance difference is great. |
Replace the regexp-based implementation of StripMarkers() and EscapeMarkers() with manual byte scanning using a shared stripMarkersBytes() function. The new implementation exploits the fact that all three marker characters (‹ U+2039, › U+203A, † U+2020) are 3-byte UTF-8 sequences sharing the prefix 0xE2 0x80. It scans for 0xE2 bytes and checks the subsequent two bytes to identify markers, then either removes them or replaces them with the escape character. Also removes the now-unused ReStripMarkers regexp and its regexp import. Benchmark results (Apple M1 Pro): name old time/op new time/op delta StripMarkers_String-10 738ns ± 2% 105ns ± 1% -85.83% (p=0.000 n=8) StripMarkers_Bytes-10 733ns ± 4% 58ns ± 1% -92.10% (p=0.000 n=8) StripMarkers_NoMarkers-10 660ns ± 1% 23ns ± 0% -96.53% (p=0.000 n=8) EscapeMarkers-10 377ns ± 2% 45ns ±18% -87.99% (p=0.000 n=8) name old alloc/op new alloc/op delta StripMarkers_String-10 185B ± 1% 144B ± 0% -22.16% (p=0.000 n=8) StripMarkers_Bytes-10 136B ± 0% 48B ± 0% -64.71% (p=0.000 n=8) StripMarkers_NoMarkers-10 144B ± 0% 0B ± 0% -100.00% (p=0.000 n=8) EscapeMarkers-10 40.0B ± 0% 24.0B ± 0% -40.00% (p=0.000 n=8) name old allocs/op new allocs/op delta StripMarkers_String-10 6.00 ± 0% 3.00 ± 0% -50.00% (p=0.000 n=8) StripMarkers_Bytes-10 5.00 ± 0% 1.00 ± 0% -80.00% (p=0.000 n=8) StripMarkers_NoMarkers-10 3.00 ± 0% 0.00 ± 0% -100.00% (p=0.000 n=8) EscapeMarkers-10 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=8) Co-Authored-By: roachdev-claude <roachdev-claude-bot@cockroachlabs.com>
dhartunian
force-pushed
the
davidh/strip-markers-perf
branch
from
cfc9aed to
ccab926
Compare
visheshbardia
approved these changes
Mar 13, 2026
Contributor
visheshbardia
left a comment
visheshbardia
left a comment
There was a problem hiding this comment.
@visheshbardia made 1 comment.
Reviewable status: 0 of 3 files reviewed, 1 unresolved discussion.
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
2 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.
Replace the regexp-based implementation of StripMarkers() and
EscapeMarkers() with manual byte scanning using a shared
stripMarkersBytes() function.
The new implementation exploits the fact that all three marker
characters (‹ U+2039, › U+203A, † U+2020) are 3-byte UTF-8 sequences
sharing the prefix 0xE2 0x80. It scans for 0xE2 bytes and checks the
subsequent two bytes to identify markers, then either removes them or
replaces them with the escape character.
Also removes the now-unused ReStripMarkers regexp and its regexp import.
Benchmark results (Apple M1 Pro):
Co-Authored-By: roachdev-claude roachdev-claude-bot@cockroachlabs.com
This change is