Re: [patch] HTML parser bugfix
Jeremy Henty <onepoint-YprzHiG/[email protected]>
| Newsgroups | gmane.comp.web.dillo.devel |
|---|---|
| Message-ID | <[email protected]> |
123 wrote:
> Attached patch removes logic that tries to detect unterminated
> quoted attributes and fixes this bug.
I would be wary of applying this patch without further thought. I run
Dillo with a very similar patch and it breaks *many* more pages than
it fixes. It is unfortunately necessary to detect and recover from
quoting errors in attributes. For instance, the web is *full* of
pages with tags containing this:
foo="bar""
If you don't detect and ignore that extra double quote you will break
many pages that every other browser renders perfectly well.
It's not even that uncommon to see monstrosities like this:
foo=""bar"""
It is also common to see short attributes (often CSS lengths)
delimited with a mixture of single and double quotes, eg:
size="12pt'
For these reasons, sadly, it is IMHO not realistic to just drop the
misquote-detection logic.
Regards,
Jeremy Henty