shaba/libxml2
1
0
Fork 0
mirror of https://gitlab.gnome.org/GNOME/libxml2.git synced 2025-02-05 05:47:00 +03:00
Code

fixed a problem with xmlCheckUTF8 reported on the mailing list by Julius

Browse Source 
* xmlstring.c: fixed a problem with xmlCheckUTF8 reported on the
  mailing list by Julius Mittenzwei
This commit is contained in:
William M. Brack 2004-08-28 01:33:30 +00:00
parent da3336f2d1
commit 3ffe90ea1c
2 changed files with 31 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
ChangeLog
View File

@ -1,3 +1,8 @@
Fri Aug 27 18:32:24 PST 2004 William Brack <wbrack@mmm.com.hk>
* xmlstring.c: fixed a problem with xmlCheckUTF8 reported on the
mailing list by Julius Mittenzwei
Fri Aug 27 00:13:39 CEST 2004 Daniel Veillard <daniel@veillard.com> Fri Aug 27 00:13:39 CEST 2004 Daniel Veillard <daniel@veillard.com>
* libxml.spec.in: added BuildRequires: zlib-devel, fixes * libxml.spec.in: added BuildRequires: zlib-devel, fixes

29
xmlstring.c
View File

@ -764,13 +764,36 @@ xmlCheckUTF8(const unsigned char *utf)
int ix; int ix;
unsigned char c; unsigned char c;
for (ix = 0; (c = utf[ix]);) { /*
if (c & 0x80) { * utf is a string of 1, 2, 3 or 4 bytes. The valid strings
if ((c & 0xc0) != 0x80 || (utf[ix + 1] & 0xc0) != 0x80) * are as follows (in "bit format"):
* 0xxxxxxx valid 1-byte
* 110xxxxx 10xxxxxx valid 2-byte
* 1110xxxx 10xxxxxx 10xxxxxx valid 3-byte
* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx valid 4-byte
*/
for (ix = 0; (c = utf[ix]);) { /* string is 0-terminated */
if (c & 0x80) { /* if it is not a single byte */
/*
* We know the first byte starts with '1', so check
* the following bits and bytes.
*
* if the first byte does *not* start with 1 1, or the
* second byte does *not* start with 1 0 it's an error
*/
if (((c & 0xc0) != 0xc0) || ((utf[ix + 1] & 0xc0) != 0x80))
return(0); return(0);
/*
* if the first three bits are set then the 3rd byte *must* start
* with 1 0
*/
if ((c & 0xe0) == 0xe0) { if ((c & 0xe0) == 0xe0) {
if ((utf[ix + 2] & 0xc0) != 0x80) if ((utf[ix + 2] & 0xc0) != 0x80)
return(0); return(0);
/*
* if the first four bits are set then the fifth bit
* must not be set, and the 4th byte *must* start with 1 0
*/
if ((c & 0xf0) == 0xf0) { if ((c & 0xf0) == 0xf0) {
if ((c & 0xf8) != 0xf0 || (utf[ix + 3] & 0xc0) != 0x80) if ((c & 0xf8) != 0xf0 || (utf[ix + 3] & 0xc0) != 0x80)
return(0); return(0);