lib/libxml2
1
0
Fork 0
mirror of https://gitlab.gnome.org/GNOME/libxml2.git synced 2025-07-29 11:41:22 +03:00
Code Activity

applied a reworked version of Usamah Malik patch to avoid growing the

Browse Source 
* HTMLparser.c: applied a reworked version of Usamah Malik patch
  to avoid growing the parser stack in some autoclose cases, should
  fix #361221
Daniel
This commit is contained in:
Daniel Veillard
2006-10-27 12:53:28 +00:00
parent 17970a72ca
commit 890fd9f9f3
2 changed files with 64 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
ChangeLog
View File

@ -1,3 +1,9 @@
Fri Oct 27 14:54:07 CEST 2006 Daniel Veillard <daniel@veillard.com>
* HTMLparser.c: applied a reworked version of Usamah Malik patch
to avoid growing the parser stack in some autoclose cases, should
fix #361221
Thu Oct 26 10:54:40 CEST 2006 Daniel Veillard <daniel@veillard.com> Thu Oct 26 10:54:40 CEST 2006 Daniel Veillard <daniel@veillard.com>
* xpath.c: William spotted an obvious bug * xpath.c: William spotted an obvious bug

58
HTMLparser.c
View File

@ -2205,6 +2205,38 @@ htmlParseHTMLName(htmlParserCtxtPtr ctxt) {
return(xmlDictLookup(ctxt->dict, loc, i)); return(xmlDictLookup(ctxt->dict, loc, i));
} }
/**
* htmlParseHTMLName_nonInvasive:
* @ctxt: an HTML parser context
*
* parse an HTML tag or attribute name, note that we convert it to lowercase
* since HTML names are not case-sensitive, this doesn't consume the data
* from the stream, it's a look-ahead
*
* Returns the Tag Name parsed or NULL
*/
static const xmlChar *
htmlParseHTMLName_nonInvasive(htmlParserCtxtPtr ctxt) {
int i = 0;
xmlChar loc[HTML_PARSER_BUFFER_SIZE];
if (!IS_ASCII_LETTER(NXT(1)) && (NXT(1) != '_') &&
(NXT(1) != ':')) return(NULL);
while ((i < HTML_PARSER_BUFFER_SIZE) &&
((IS_ASCII_LETTER(NXT(1+i))) || (IS_ASCII_DIGIT(NXT(1+i))) ||
(NXT(1+i) == ':') || (NXT(1+i) == '-') || (NXT(1+i) == '_'))) {
if ((NXT(1+i) >= 'A') && (NXT(1+i) <= 'Z')) loc[i] = NXT(1+i) + 0x20;
else loc[i] = NXT(1+i);
i++;
}
return(xmlDictLookup(ctxt->dict, loc, i));
}
/** /**
* htmlParseName: * htmlParseName:
* @ctxt: an HTML parser context * @ctxt: an HTML parser context
@ -3775,6 +3807,7 @@ static void
htmlParseContent(htmlParserCtxtPtr ctxt) { htmlParseContent(htmlParserCtxtPtr ctxt) {
xmlChar *currentNode; xmlChar *currentNode;
int depth; int depth;
const xmlChar *name;
currentNode = xmlStrdup(ctxt->name); currentNode = xmlStrdup(ctxt->name);
depth = ctxt->nameNr; depth = ctxt->nameNr;
@ -3795,6 +3828,31 @@ htmlParseContent(htmlParserCtxtPtr ctxt) {
continue; /* while */ continue; /* while */
} }
else if ((CUR == '<') &&
((IS_ASCII_LETTER(NXT(1))) ||
(NXT(1) == '_') || (NXT(1) == ':'))) {
name = htmlParseHTMLName_nonInvasive(ctxt);
if (name == NULL) {
htmlParseErr(ctxt, XML_ERR_NAME_REQUIRED,
"htmlParseStartTag: invalid element name\n",
NULL, NULL);
/* Dump the bogus tag like browsers do */
while ((IS_CHAR_CH(CUR)) && (CUR != '>'))
NEXT;
if (currentNode != NULL)
xmlFree(currentNode);
return;
}
if (ctxt->name != NULL) {
if (htmlCheckAutoClose(name, ctxt->name) == 1) {
htmlAutoClose(ctxt, name);
continue;
}
}
}
/* /*
* Has this node been popped out during parsing of * Has this node been popped out during parsing of
* the next element * the next element