Re: Character encodings problem on hungarian text

["Nickolay V. Shmyrev" <nshmyrev yandex ru>]

В Вск, 10/07/2005 в 13:12 +0200, Khiraly пишет:
> Hi!
> 
> The hungarian accents(o" and u") are not working with evince.
> There is no problem with xpdf.
> 
> Here is an example:
> http://www.artisjus.hu/opencms/export/download/felhasznaloknak/aktualis_jogdijkozlemeny_U.pdf
> 
> Here is a screenshot from acrobat reader (only the problematic part):
> http://khiraly.4242.hu/tmp/adobe-good.png
> 
> Here is a screenshot from evince (the same part of the document):
> http://khiraly.4242.hu/tmp/evince-bad.png
> 
> Best regards, 
>  Khiraly
> 


Thanks, it's a poppler bug. You should file bug to poppler in
bugzilla.freedesktop.org. 

I don't know why it's working fine in xpdf, mine xpdf-3.0 also fails on
this document. The reason on it that poppler assumes that only
Adobe-Identity and Adobe-UCS fonts can miss maps to unicode while it's
not true. Here is small and dirty patch to poppler that fixes your
problem


--- poppler/GfxFont.cc	2005-07-07 00:02:45.000000000 +0400
+++ poppler/GfxFont.cc.new 2005-07-10 19:10:18.000000000 +0400
@@ -1157,7 +1157,7 @@ GfxCIDFont::GfxCIDFont(XRef *xref, char 
 
   // look for a ToUnicode CMap
   if (!(ctu = readToUnicodeCMap(fontDict, 16, NULL))) {
-
+#if 0
     // the "Adobe-Identity" and "Adobe-UCS" collections don't have
     // cidToUnicode files
     if (collection->cmp("Adobe-Identity") &&
@@ -1171,6 +1171,7 @@ GfxCIDFont::GfxCIDFont(XRef *xref, char 
 	goto err2;
       }
     }
+#endif
   }
 
   // encoding (i.e., CMap)