[poppler] poppler pdftops regressions

[William Bader]

Hi,
I am almost finished updating my "old" patches into the new version of pdftops.
The new OverprintPreview code interferes with it in PSOutput.cc, and I suspect that is why the patches were dropped.
The old patches are actually three sets of patches: part of the implementation of -binary, a bug fix where level2sep and level3sep wrote rgb color instead of cmyk color, and an optimization to write color images as mono if they contain only shades of gray.
I am redoing my old patches to use a single case block instead treating level2 and level3 in a different case block than level2sep and level3sep to avoid repeating code.
I still have all of the test files and scripts from when I wrote the original patches, so I can test the new patches well before I post them.
William

> From: aacid at kde.org
> To: poppler at lists.freedesktop.org
> Date: Fri, 24 Feb 2012 19:12:02 +0100
> Subject: Re: [poppler] poppler pdftops regressions
> 
> El Divendres, 24 de febrer de 2012, a les 09:33:03, Thomas Freitag va 
> escriure:
> > Am 23.02.2012 23:18, schrieb Albert Astals Cid:
> > > El Dijous, 23 de febrer de 2012, a les 14:36:29, William Bader va 
> escriure:
> > >> Hi,
> > > 
> > > Hi
> > > 
> > >> I am using a 32-bit i686 version of Fedora 14 Linux.
> > >> I copied poppler from git earlier today (Feb 23, 2012), and noticed
> > >> two
> > >> problems.
> > >> 
> > >> pdftops -level1 does not seem to handle bitmaps well.
> > >> My test files 020_ReadMe_Ghent_Output_Patch.pdf (Ghent Output Suite
> > >> document), bug-poppler27482-precision.pdf (Herbstfest bei MEIFORT ad),
> > >> elp-B83378-gray.pdf, and testimagemask.pdf (real estate ad with photos
> > >> of six agents at the bottom) showed like checkerboards. The size is
> > >> what I expect, and they don't get errors.  I suspect that the problem
> > >> is a change in the image code in the pdftops prolog.
> > >> 
> > >> pdftops -binary no longer has much effect.
> > >> It is present PSOutput.cc in the recent tarball release of 0.18.4, but
> > >> about half of the code is missing in the git version. Was that
> > >> intentional, or did a patch get applied to a branch that was not
> > >> later merged?
> > > 
> > > The patch was merged correctly but Thomas smashed your changes when
> > > merging xpdf303 changes and we did not realize
> > 
> > Sorry if I missed something. I'll have a look at it at the weekend, I'm
> > anyway working on a patch for CJK fonts in PSOutput. Are the "old"
> > patches from William available? 
> 
> Sure https://bugs.freedesktop.org/show_bug.cgi?id=34003 or 
> http://cgit.freedesktop.org/poppler/poppler/commit/?id=c3470145f95791167c19a438934a923eab8a93cf
> 
> > It's probably easier to look into the
> > patches and see what is missing now (of course together with comparing
> > code between tarball and git). Also it would be nice to have
> > elp-B83378-gray.pdf and testimagemask.pdf for testing, the other I have.
> > Didn't think that still someone use level1 option :-) Aren't that
> > printers been already in a museum?
> 
> William sent a patch for the level1 image regressionwhich i'm testing at the 
> moment.
> 
> Albert
> 
> > 
> > Thomas
> > 
> > >> Should I open bug reports for each of these issues?
> > > 
> > > Please, and if you can give patches to fix them even better :-)
> > > 
> > > Albert
> > > 
> > >> William
> > > 
> > > _______________________________________________
> > > poppler mailing list
> > > poppler at lists.freedesktop.org
> > > http://lists.freedesktop.org/mailman/listinfo/poppler
> > > 
> > > .
> > 
> > _______________________________________________
> > poppler mailing list
> > poppler at lists.freedesktop.org
> > http://lists.freedesktop.org/mailman/listinfo/poppler
> _______________________________________________
> poppler mailing list
> poppler at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/poppler
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/poppler/attachments/20120224/538fa51e/attachment.html>