Re: [xml] HTML Parser problems with chunk parser if HTML keywordsoverlap chunk border

["Cyrill Osterwalder" <Cyrill Osterwalder visonys com>]

> Why do you use PARSE_HTML_RECOVER ? The parser is already 
> doing recovery mode to some extend without them 
> (I mean the HTML parser :-).

Actually, the problem also seems to exist without PARSE_HTML_RECOVER,
otherwise the test with testHTML.c of the libxml2 package would not show
it, right? I will have to look at this again. I had the impression that
recovery mode is the trigger in htmlParseScript() to actually produce
the problem. But my testHTML.c example can be easily reproduced and it
does not use HTML_RECOVER. With the testHTML.c example it seems that
parsing fails if the CDATA end tag overlaps the chunk boundary. If
that's true even without PASRE_HTML_RECOVER, then it's just a matter of
luck if chunked parsing HTML with CDATA is successful.

> with '</ style' or '</foo> and expect taht to close the open tag, and
> 'style "</" style' and expect to not close it...

I see. I guess there's a reason why the slash in "</" should be quoted
in CDATA contents if not being the real end tag ;-) However, there's a
lot of HTML out "in the wild" containing unquoted "</" strings in CDATA
blocks.

> You can try, but it's all very messy IMHO, I will take 
> patches if not obviously broken

I will further look at it and get back to the list if I'm able to
produce anything useful.

Thanks,

Cyrill