Skip to content

ext/mbstring: Optimize valid UTF-8 in mb_substr() #24080

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
kamil-tekiela wants to merge 2 commits into
base: master
Choose a base branch
from
Open

ext/mbstring: Optimize valid UTF-8 in mb_substr() #24080

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b218ba6
Optimize valid UTF-8 in mb_substr()
kamil-tekiela Oct 2, 2026
6d9250e
Restrict optimization to only UTF-8
kamil-tekiela Oct 16, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
46 changes: 42 additions & 4 deletions ext/mbstring/mbstring.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1879,7 +1879,9 @@ static inline bool php_mb_is_no_encoding_utf8(enum mbfl_no_encoding no_enc)
return (no_enc >= mbfl_no_encoding_utf8 && no_enc <= mbfl_no_encoding_utf8_sb);
}

static unsigned char* offset_to_pointer_utf8(unsigned char *str, unsigned char *end, ssize_t offset) {
#define IS_UTF8_CONTINUATION_BYTE(c) (((c) & 0xC0) == 0x80)

static unsigned char* offset_to_pointer_utf8(unsigned char *str, unsigned char *end, ssize_t offset, bool is_valid_utf8) {
if (offset < 0) {
unsigned char *pos = end;
while (offset < 0) {
Expand All @@ -1888,14 +1890,29 @@ static unsigned char* offset_to_pointer_utf8(unsigned char *str, unsigned char *
}

unsigned char c = *--pos;
if (c < 0x80 || (c & 0xC0) != 0x80) {
if (!IS_UTF8_CONTINUATION_BYTE(c)) {
offset++;
}
}
return pos;
} else {
const unsigned char *u8_tbl = mbfl_encoding_utf8.mblen_table;
unsigned char *pos = str;
if (is_valid_utf8) {
while (offset > 256 && (size_t)(end - pos) >= 256) {
offset -= mb_fast_strlen_utf8(pos, 256);
pos += 256;
}
for (; pos < end; pos++) {
if (!IS_UTF8_CONTINUATION_BYTE(*pos)) {
if (offset == 0) {
return pos;
}
offset--;
}
}
return offset == 0 ? end : NULL;
}
while (offset-- > 0) {
if (pos >= end) {
return NULL;
Expand Down Expand Up @@ -1928,7 +1945,7 @@ static size_t mb_find_strpos(zend_string *haystack, zend_string *needle, const m
needle_u8 = needle;
}

offset_pointer = offset_to_pointer_utf8((unsigned char*)ZSTR_VAL(haystack_u8), (unsigned char*)ZSTR_VAL(haystack_u8) + ZSTR_LEN(haystack_u8), offset);
offset_pointer = offset_to_pointer_utf8((unsigned char*)ZSTR_VAL(haystack_u8), (unsigned char*)ZSTR_VAL(haystack_u8) + ZSTR_LEN(haystack_u8), offset, ZSTR_IS_VALID_UTF8(haystack_u8));
if (!offset_pointer) {
result = MBFL_ERROR_OFFSET;
goto out;
Expand All @@ -1946,7 +1963,7 @@ static size_t mb_find_strpos(zend_string *haystack, zend_string *needle, const m
found_pos = zend_memnrstr((const char*)offset_pointer, ZSTR_VAL(needle_u8), ZSTR_LEN(needle_u8), ZSTR_VAL(haystack_u8) + ZSTR_LEN(haystack_u8));
} else {
size_t needle_len = pointer_to_offset_utf8((unsigned char*)ZSTR_VAL(needle_u8), (unsigned char*)ZSTR_VAL(needle_u8) + ZSTR_LEN(needle_u8));
offset_pointer = offset_to_pointer_utf8(offset_pointer, (unsigned char*)ZSTR_VAL(haystack_u8) + ZSTR_LEN(haystack_u8), needle_len);
offset_pointer = offset_to_pointer_utf8(offset_pointer, (unsigned char*)ZSTR_VAL(haystack_u8) + ZSTR_LEN(haystack_u8), needle_len, ZSTR_IS_VALID_UTF8(haystack_u8));
if (!offset_pointer) {
offset_pointer = (unsigned char*)ZSTR_VAL(haystack_u8) + ZSTR_LEN(haystack_u8);
}
Expand Down Expand Up @@ -2185,6 +2202,27 @@ static zend_string* mb_get_substr(zend_string *input, size_t from, size_t len, c
return zend_string_init_fast((const char*)in, len);
}

if (enc == &mbfl_encoding_utf8 && ZSTR_IS_VALID_UTF8(input)) {
/* Valid UTF-8 can be sliced on codepoint boundaries without decoding it */
unsigned char *end = in + in_len;
unsigned char *start = offset_to_pointer_utf8(in, end, from, true);
if (!start) {
return zend_empty_string;
}
unsigned char *stop = end;
if (len < (size_t)(end - start)) {
stop = offset_to_pointer_utf8(start, end, len, true);
if (!stop) {
stop = end;
}
}
zend_string *result = zend_string_init_fast((const char*)start, stop - start);
if (!ZSTR_IS_INTERNED(result)) {
GC_ADD_FLAGS(result, IS_STR_VALID_UTF8);
}
return result;
}

return mb_get_substr_slow(in, in_len, from, len, enc);
}

Expand Down
Loading