Re: [Dev-luatex] Fwd: [poppler] [PATCH] Fixup LaTeX composed characters

On 25-3-2011 2:40, Martin Schröder wrote:

FYI. Don't ask me what's the problem with our PDF seems to be...

indeed. if we even consider this for luatex, it should be an option that is not applied by default ... tweaking an included pdf file should never happen automatically as overlapping can be intentional ... personally I think that this is more something to plug into (for instance) qpdf i.e. repair the problematic file using independent tools (actually, if it is added to pdftex one could use pdftex to copy the document into a another one that then gets the latex accents corrected) Hans

---------- Forwarded message ---------- From: Tim Brody Date: 2011/3/25 Subject: [poppler] [PATCH] Fixup LaTeX composed characters To: poppler@lists.freedesktop.org

Hi All,

Attached is a patch to address the previous problem I wrote about with pdflatex-produced PDFs that contain overlapping-diacritics/accents.

This patch contains: - a table of diacritic to Unicode combining character code-points - if an overlapping character is detected checks whether the first (in stream-sequence) character is in the table - pops the diacritic off the word - appends the diacritic to the character as a Unicode combining character

This does not fix \b{o} or \d{o} because TeX places them on the next line (so aren't detected as overlapping).

Yes, this is an issue with pdflatex but there are 100,000s of TeX-produced PDFs for which we don't have source for ...

----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------