[Libreoffice] [PUSHED] [PATCH] Fix Bug 34666

[Petr Mladek]

Christina Roßmanith píše v Pá 18. 03. 2011 v 20:51 +0100:
> Am 18.03.2011 14:08, schrieb Petr Mladek:
> >> Very cool, checked that - it's pushed! :)
> > Heh, I have tried the patch yesterday and did not see any difference in
> > the HTML import. I was not sure what it actually fixed, so I asked
> > Cedric for more details (he created the simplified test case). I havn't
> > got the response yet.
> The testdoc.html has a unknown tag which is closed inside the tag 
> <o:bla........../>. I tried to add a separate closing tag 
> <o:bla......></o:bla> and that worked. (*) But I thought, the goal is to 
> open testdoc.html properly not to change it that it can be opened. Then 
> I saw that the closing '/>' was eaten by ScanText('>') which eats 
> everything up to '>'. But if you inspect the last character of aToken 
> you can decide that the tag was closed immediately. What I didn't check 
> is whether testdoc.html is HTML-conformant. But see above (*)...

Makes perfect sense. My only problem was that I did not see any
difference when I opened the original testdoc.html in LO. Should an
extra comment appear there? Maybe, I am just blind. :-)

> Thanks. I expected some improving suggestions, so I'm not offended.

Great :-)


Best Regards,
Petr