Re: [xml] CDATA sections do not count line numbers via SAX2 API

On Mon, 22 Dec 2003 16:48:57 +0000, Dave Beckett <dave beckett bristol ac uk> wrote:
... after dealing with the cdata, the following is at parser.c line 9719:
                  SKIP(base + 3);.

where base is returned from  xmlParseLookupSequence
looking for the closing ]]>

Earlier on it says
 *   SKIP(n) Skip n xmlChar, and must also be used only to skip ASCII defined
 *           strings without newlines within the parser.

However, that SKIP does (may) contain newlines.  There's
another nearby one at 9705 that might also be suspicious.

Well, I tried fixing that by making a SKIPL(x) that checks for
newlines in the skipped chars.  This is only needed for SKIP(x)
when x is not a constant, both places I see here are near
the CDATA counting.

So the attached patch adds a SKIPL macro and uses it;
with that, the answer I get is consistent.


Attachment: parser.c.patch
Description: Text document

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]