Commit f2a59c14b95 for php
commit f2a59c14b951e41759aafb4844829e645b9b9edd
Author: Weilin Du <weilindu@php.net>
Date: Thu Oct 1 00:28:40 2026 +0800
ext/intl: Fix grapheme_strstr() and grapheme_stristr() match offsets after supplementary characters (#24012)
Convert the UTF-16 match offset to a code point count before using it
to locate the UTF-8 substring in grapheme_strstr() and grapheme_stristr().
Add regression tests for supplementary characters and combining sequences,
covering both functions and both beforeNeedle modes.
diff --git a/NEWS b/NEWS
index 2aa1831288a..4b9cd689113 100644
--- a/NEWS
+++ b/NEWS
@@ -65,6 +65,8 @@ PHP NEWS
(Ilia Alshanetsky)
. Fixed Collator attribute and strength methods not rejecting an
unconstructed Collator. (Ilia Alshanetsky)
+ . Fixed grapheme_strstr() and grapheme_stristr() returning incorrect substrings
+ when supplementary characters precede the match. (Weilin Du)
- Lexbor:
. Merge patches lexbor/lexbor@8a14bc0 and lexbor/lexbor@f67ce4b, fixing a
diff --git a/ext/intl/grapheme/grapheme_util.c b/ext/intl/grapheme/grapheme_util.c
index f0c79785483..66c3876a5bd 100644
--- a/ext/intl/grapheme/grapheme_util.c
+++ b/ext/intl/grapheme/grapheme_util.c
@@ -190,7 +190,7 @@ int32_t grapheme_strpos_utf16(char *haystack, size_t haystack_len, char *needle,
if(char_pos != USEARCH_DONE && ubrk_isBoundary(bi, char_pos)) {
ret_pos = grapheme_count_graphemes(bi, uhaystack,char_pos);
if(puchar_pos) {
- *puchar_pos = char_pos;
+ *puchar_pos = u_countChar32(uhaystack, char_pos);
}
} else {
ret_pos = -1;
diff --git a/ext/intl/tests/grapheme_strstr_supplementary_prefix.phpt b/ext/intl/tests/grapheme_strstr_supplementary_prefix.phpt
new file mode 100644
index 00000000000..f9f85880604
--- /dev/null
+++ b/ext/intl/tests/grapheme_strstr_supplementary_prefix.phpt
@@ -0,0 +1,27 @@
+--TEST--
+grapheme_strstr() and grapheme_stristr() preserve match boundaries after supplementary characters
+--EXTENSIONS--
+intl
+--FILE--
+<?php
+foreach (["\u{1F600}", "\u{1F600}\u{1F600}", "\u{1F600}e\u{0301}"] as $prefix) {
+ $haystack = $prefix . 'abc';
+ var_dump(grapheme_strstr($haystack, 'a'));
+ var_dump(grapheme_strstr($haystack, 'a', true));
+ var_dump(grapheme_stristr($haystack, 'A'));
+ var_dump(grapheme_stristr($haystack, 'A', true));
+}
+?>
+--EXPECT--
+string(3) "abc"
+string(4) "š"
+string(3) "abc"
+string(4) "š"
+string(3) "abc"
+string(8) "šš"
+string(3) "abc"
+string(8) "šš"
+string(3) "abc"
+string(7) "šeĢ"
+string(3) "abc"
+string(7) "šeĢ"