[poppler] New regressions found with current git master

[suzuki toshiya]

Dear Carlos,

Because Thomas and I are working for font substitution feature,
I'm interested in the regression of Anna-IptImgx330.pdf.
In fact, Anna-IptImgx330.pdf refers Helvetica and Times-Roman
that are not embedded in the PDF.

But, at present, I could not reproduce the regression. I think
"Text is not rendered" means that pdftoppm results will lose
the titles of each pictures (e.g. "Monotone.eps", "Duotone.eps" etc).
In my GNU/Linux and MinGW environment, they are rendered.

Could you explain more about your environment? I want to reproduce
the bug and try to fix it. Considering that the referred fonts are
not for CJK, the issue might be related with the base14font patch
applied in d6a1b7dcaeac1e49533519b9f8a279fd64d04c67.

Regards,
mpsuzuki

Carlos Garcia Campos wrote:
> I've just run the regtests with my pdfs again using current git
> master. The regressions I reported previously regarding fonts and
> gradients seem to be fixed now \o/. I've found 3 new regressions,
> though:
> 
> images/Anna-IptImgx330.pdf - Text is not rendered
> 
> This one can be found here
> http://acroeng.adobe.com/Test_Files/images/jpeg2000/
> 
> printing/pom_svsa_gesuch-lernfahrausweis.pdf - Some lines are rendered
> dashed
> 
> This one is attached.
> 
> transparency/[6]_aufwandsschaetzung.pdf - In this case a dashed
> rectangle is incorrectly rendered in pages 17, 18 and 19
> 
> This one is attached too.
> 
> I haven't had time to run a bisect yet, I'll try to do it this weekend.
> 
> Regards, 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> poppler mailing list
> poppler at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/poppler