Consecutive <meta> tags aren't handled properly

[AnyOldName3]

The best example of this is probably the stub that gets made for certain URLs that redirect https://github.com/mitchcapper/httrack/blob/master/src/htsparse.c#L3610. When that gets parsed, the Content-Type <meta> tag is skipped, which then means the <meta> tag detection never notices the next one that does the redirect. The URL is still detected, but because the parser hasn't noticed it's in a <meta> tag, it doesn't stop at the closing ", and thinks that's still part of the URL, only stopping at the closing >. This then means a bogus URL gets archived and the redirect page ends up malformed, so doesn't redirect or display the later link to manually click on properly.

Everything seems to work if I add a continue here https://github.com/mitchcapper/httrack/blob/master/src/htsparse.c#L715 to go back to the start of the parsing loop when the position pointer's been advanced past the tag that's being skipped.

[emmguyot]

I've made a quick check and I doubt that the "continue" is the good way.

The problem isn't about consecutive meta tags in general. I have some and it works fine.
The problem looks to be about a meta tag right after the meta content-type.

So I guess it's more about intag_ctype that is set to 1 and shouldn't because we want to skip the tag content-type.

I need to make some tests to be sure and fix it in my merge request.

[emmguyot]

I've fixed the issue and it is now in the waiting merge request xroche#227