clones/php-src
1
0
Fork 0
mirror of https://github.com/php/php-src.git synced 2024-09-23 10:57:26 +00:00
Code Issues Packages Projects Releases Wiki Activity

- Bug #49785: take 5. What the hell happened to me...

Browse Source
This commit is contained in:
Moriyoshi Koizumi 2009-10-13 05:18:37 +00:00
parent 5d34329dab
commit 20737bac6a
2 changed files with 21 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
ext/standard/html.c
View File

@ -538,19 +538,18 @@ inline static unsigned int get_next_char(enum entity_charset charset,
c = str[pos]; c = str[pos];
if (c < 0x80) { if (c < 0x80) {
MB_WRITE(c); MB_WRITE(c);
this_char = c; this_char = c;
pos++; pos++;
} else if (c < 0xc0) { } else if (c < 0xc0) {
MB_FAILURE(pos); MB_FAILURE(pos);
} else if (c < 0xe0) { } else if (c < 0xe0) {
CHECK_LEN(pos, 2); CHECK_LEN(pos, 2);
if (str[pos + 1] < 0x80 || str[pos + 1] > 0xbf) { if (str[pos + 1] < 0x80 || str[pos + 1] > 0xbf) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
this_char = ((c & 0x1f) << 6) | (str[pos + 1] & 0x3f); this_char = ((c & 0x1f) << 6) | (str[pos + 1] & 0x3f);
if (this_char < 0x80) { if (this_char < 0x80) {
*status = FAILURE; MB_FAILURE(pos);
return 0;
} }
MB_WRITE((unsigned char)c); MB_WRITE((unsigned char)c);
MB_WRITE((unsigned char)str[pos + 1]); MB_WRITE((unsigned char)str[pos + 1]);
@ -558,14 +557,14 @@ inline static unsigned int get_next_char(enum entity_charset charset,
} else if (c < 0xf0) { } else if (c < 0xf0) {
CHECK_LEN(pos, 3); CHECK_LEN(pos, 3);
if (str[pos + 1] < 0x80 || str[pos + 1] > 0xbf) { if (str[pos + 1] < 0x80 || str[pos + 1] > 0xbf) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
if (str[pos + 2] < 0x80 || str[pos + 2] > 0xbf) { if (str[pos + 2] < 0x80 || str[pos + 2] > 0xbf) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
this_char = ((c & 0x0f) << 12) | ((str[pos + 1] & 0x3f) << 6) | (str[pos + 2] & 0x3f); this_char = ((c & 0x0f) << 12) | ((str[pos + 1] & 0x3f) << 6) | (str[pos + 2] & 0x3f);
if (this_char < 0x800) { if (this_char < 0x800) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
MB_WRITE((unsigned char)c); MB_WRITE((unsigned char)c);
MB_WRITE((unsigned char)str[pos + 1]); MB_WRITE((unsigned char)str[pos + 1]);
@ -574,17 +573,17 @@ inline static unsigned int get_next_char(enum entity_charset charset,
} else if (c < 0xf8) { } else if (c < 0xf8) {
CHECK_LEN(pos, 4); CHECK_LEN(pos, 4);
if (str[pos + 1] < 0x80 || str[pos + 1] > 0xbf) { if (str[pos + 1] < 0x80 || str[pos + 1] > 0xbf) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
if (str[pos + 2] < 0x80 || str[pos + 2] > 0xbf) { if (str[pos + 2] < 0x80 || str[pos + 2] > 0xbf) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
if (str[pos + 3] < 0x80 || str[pos + 3] > 0xbf) { if (str[pos + 3] < 0x80 || str[pos + 3] > 0xbf) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
this_char = ((c & 0x07) << 18) | ((str[pos + 1] & 0x3f) << 12) | ((str[pos + 2] & 0x3f) << 6) | (str[pos + 3] & 0x3f); this_char = ((c & 0x07) << 18) | ((str[pos + 1] & 0x3f) << 12) | ((str[pos + 2] & 0x3f) << 6) | (str[pos + 3] & 0x3f);
if (this_char < 0x10000) { if (this_char < 0x10000) {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
MB_WRITE((unsigned char)c); MB_WRITE((unsigned char)c);
MB_WRITE((unsigned char)str[pos + 1]); MB_WRITE((unsigned char)str[pos + 1]);
@ -592,7 +591,7 @@ inline static unsigned int get_next_char(enum entity_charset charset,
MB_WRITE((unsigned char)str[pos + 3]); MB_WRITE((unsigned char)str[pos + 3]);
pos += 4; pos += 4;
} else { } else {
MB_FAILURE(pos); MB_FAILURE(pos);
} }
} }
break; break;

10
ext/standard/tests/strings/bug49785.phpt
View File

@ -35,6 +35,12 @@ var_dump(_bin2hex(htmlentities("\xf7\xbf\xff\xbf", ENT_QUOTES, "UTF-8")));
var_dump(_bin2hex(htmlentities("\xf7\xbf\xbf\xff", ENT_QUOTES, "UTF-8"))); var_dump(_bin2hex(htmlentities("\xf7\xbf\xbf\xff", ENT_QUOTES, "UTF-8")));
var_dump(_bin2hex(htmlentities("\xf8\x88\x80\x80\x80", ENT_QUOTES, "UTF-8"))); var_dump(_bin2hex(htmlentities("\xf8\x88\x80\x80\x80", ENT_QUOTES, "UTF-8")));
echo "--\n";
// UTF-8: with ENT_IGNORE
var_dump(_bin2hex(htmlentities("\xc0\xa0\xc2\x80", ENT_QUOTES | ENT_IGNORE, "UTF-8")));
var_dump(_bin2hex(htmlentities("\xe0\x80\x80\xe0\xa0\x80", ENT_QUOTES | ENT_IGNORE, "UTF-8")));
var_dump(_bin2hex(htmlentities("\xf0\x80\x80\x80\xf0\x90\x80\x80", ENT_QUOTES | ENT_IGNORE, "UTF-8")));
echo "--\n"; echo "--\n";
// UTF-8: alternative (invalid) UTF-8 sequence // UTF-8: alternative (invalid) UTF-8 sequence
var_dump(_bin2hex(htmlspecialchars("\xc0\xa6", ENT_QUOTES, 'UTF-8'))); var_dump(_bin2hex(htmlspecialchars("\xc0\xa6", ENT_QUOTES, 'UTF-8')));
@ -155,6 +161,10 @@ string(0) ""
string(0) "" string(0) ""
string(0) "" string(0) ""
-- --
string(4) "c280"
string(6) "e0a080"
string(8) "f0908080"
--
string(0) "" string(0) ""
string(0) "" string(0) ""
string(0) "" string(0) ""