[Pkg-haskell-maintainers] Bug#739448: pandoc: endless loop converting from html

[John MacFarlane]

+++ Jonas Smedegaard [Feb 19 14 21:41 ]:
> reopen 739448
> thanks
> 
> Quoting Thomas Viehweger (2014-02-19 20:49:51)
> >>>> pandoc -f html -t epub -o 131212.epub \
> >>>> http://lwn.net/Articles/575838/bigpage?format=printable
> >>>
> >>> I can reproduce the error.
> >>>
> >>> Seems it is not tied to epub (trying with markdown does the same) 
> >>> but instead with impurity of html (it works after cleaning up with 
> >>> tidy).
> >>
> >> Investigating closer - using weblint (Debian package weblint-perl) 
> >> reveals that the html page is broken: contains several non-closed tr 
> >> and td tags.
> >
> > thanks for the quick response. tidy-ing the html file really helps.
> >
> > But.. the previous version of pandoc in testing (1.9.?) also behaves 
> > as expected (no endless loop).
> >
> > In my opinion pandoc should not go into an endless loop - no matter 
> > what I feed it with.
> > It might produce garbage or abort. But an endless loop is not 
> > acceptable (for me).
> 
> Good point.  Reopening the bugreport.
> 
> Btw: A closed bugreport can still be posted to (I guess that's why you 
> contacted me discretely, and I took the liberty to reply in public).  
> Only when _archived_ can you no longer post to it.

I've added a bug report upstream:
https://github.com/jgm/pandoc/issues/1167