[poppler] poppler/Annot.cc poppler/Annot.h

[Albert Astals Cid]

A Dilluns, 7 de març de 2011, Carlos Garcia Campos va escriure:
> Excerpts from Albert Astals Cid's message of dom mar 06 21:15:58 +0100 2011:
> > A Dissabte, 5 de març de 2011, Carlos Garcia Campos va escriure:
> > >  poppler/Annot.cc |   91
> > > 
> > > ++++++++++++++++++------------------------------------- poppler/Annot.h
> > >  |
> > > 
> > >    4 +-
> > >  
> > >  2 files changed, 34 insertions(+), 61 deletions(-)
> > > 
> > > New commits:
> > > commit 128dcaf282f418d8c45cea4df6ad4d3167b3e39c
> > > Author: Carlos Garcia Campos <carlosgc at gnome.org>
> > > Date:   Sat Mar 5 13:12:49 2011 +0100
> > > 
> > >     annots: Add AnnotBorder::parseDashArray() to parse dash arrays
> > >     
> > >     The code was duplicated in AnnotBorderArray and AnnotBorderBS
> > 
> > As far as i understand this introduces a small behaviour change when
> > if (array->get(3, &obj1)->isArray())
> > is not true in AnnotBorderArray since previously it was setting correct
> > to false which in turn set width to 0 and now that does nto happen.
> > 
> > Is this in purpose?
> 
> Yes, I don't think we shouldn't draw any border just because the dash
> array is invalid. This is for consistency with AnnotBorderBS that
> creates a default dash array when when dict->lookup("D",&obj1)->isArray()
> fails. Since the default for an AnnotBorderArray is drawing a solid
> rectangle, we simply ignore the dash array if it's not valid. Reading
> the code again I think we should only use the default dash array for
> AnnotBorderBS when border style is dashed. I'll change that.

Some of those changes broke 
http://launchpadlibrarian.net/19906672/ClassSchedule_2091_ENGL.pdf can you 
please fix it?

Thanks,
  Albert

> 
> > Albert
> 
> Regards,