[php-src] master: ext/standard: Optimize single-character searches in strcspn() and strpbrk() (#24087)
| From: | Weilin Du via GitHub | Date: | Sat, 03 Oct 2026 17:21:12 +0000 |
| Subject: | [php-src] master: ext/standard: Optimize single-character searches in strcspn() and strpbrk() (#24087) | ||
| Groups: | php.cvs | ||
| Request: | Send a blank email to php-cvs+get-141907@lists.php.net to get a copy of this message | ||
Author: Weilin Du (LamentXU123)
Committer: GitHub (web-flow)
Pusher: LamentXU123
Date: 2026-10-04T01:21:09+08:00
Commit: https://github.com/php/php-src/commit/c33572a7eef54ad5abec3cf270a0d8734c0fb9ae
Raw diff: https://github.com/php/php-src/commit/c33572a7eef54ad5abec3cf270a0d8734c0fb9ae.diff
ext/standard: Optimize single-character searches in strcspn() and strpbrk() (#24087)
Changed paths:
M UPGRADING
M ext/standard/string.c
Diff:
diff --git a/UPGRADING b/UPGRADING
index 9692bfd9b2cc..5fa66cfc8749 100644
--- a/UPGRADING
+++ b/UPGRADING
@@ -86,6 +86,8 @@ PHP 8.7 UPGRADE NOTES
========================================
- Standard:
+ . Improved performance of strcspn() and strpbrk() when searching longer strings
+ with a single-byte character mask.
. Improved performance of array_splice() when inserting without removing
elements.
. Improved performance of str_rot13().
diff --git a/ext/standard/string.c b/ext/standard/string.c
index d23c117321a4..8193b23dd8bc 100644
--- a/ext/standard/string.c
+++ b/ext/standard/string.c
@@ -1740,6 +1740,11 @@ static size_t php_strspn_strcspn_common(const char *haystack, const char
*charac
* We only compare in this case.
* Empirically tested that the table lookup approach is only beneficial if characters is longer
than 1 character. */
if (characters_end - characters == 1) {
+ /* Avoid memchr() overhead for short spans and an immediate match. */
+ if (!must_match && haystack_end - haystack >= 16 && *haystack != *characters)
{
+ const char *ptr = memchr(haystack, (unsigned char) *characters, haystack_end - haystack);
+ return ptr ? (size_t) (ptr - haystack) : (size_t) (haystack_end - haystack);
+ }
const char *ptr = haystack;
while (ptr < haystack_end && (*ptr == *characters) == must_match) {
ptr++;