OpenType features for Armenian not working in PDF exported from MS Word

Adam JagoszAdam Jagosz Posts: 65
edited November 2016 in Font Technology
The features work correctly in MS Word, but upon export to pdf they fail. Analogical features for Latin and Cyrillic work correctly in the exported pdf file. The font file was crafted in FontForge, the pdf's were viewed in Chrome, Edge and Acrobat with the same results (except for blurry font rendering in Edge, lol). The script I used was as follows (excerpts):

languagesystem armn dflt;
languagesystem armn HYE;

lookup calt_armenian_Liwn {
    lookupflag 0;
        ignore sub uni053C' @arm_desc_left;
        sub uni053C' @armenian by uni053C.long;
} calt_armenian_Liwn;

feature calt {
script DFLT;
     language dflt ;
          lookup calt_armenian_Liwn;
script armn;     language dflt;
          lookup calt_armenian_Liwn;
    language HYE;
          lookup calt_armenian_Liwn;
} calt;


As you can see, I doubled things wherever possible, with no effect. I heard Adobe products had no support for the Serbian language, are there known issues with Armenian as well? Am I doing something wrong?

Comments

  • My PDF with Armenian OpenType features works just fine in both Adobe Acrobat Reader DC 2015 and Firefox 49.0.2. The font is made with FontCreator, and the PDF was made with Word 2016.

    Do contact me if you want me to run your font through FontCreator.
  • So, you're saying you get the right glyphs/behavior onscreen in Word, but when you create PDF from Word, you are no longer getting the correct behavior? Is that correct?

    You don't mention what version of Word, nor what platform you're on. I'm guessing some version of Windows, as you mention testing with Edge. I'm also guessing you used built-in PDF export from Word, not Adobe stuff?
  • Adam JagoszAdam Jagosz Posts: 65
    edited November 2016
    I'm on Windows 10, Word 2016. Using Word's export function. FF 2016-10-04.
    However, I managed to overcome the problem by copying all lookups to all language systems. I have several of them for SRB, MKD, BGR etc. Maybe the document was off in some place and detected Cyrillic as I messed around with it here and there? Or I messed something up in the font project that I managed to avoid this time.
    I didn't find the language system thing useful at all except for the locl feature in the first place...
  • Generally, Word does make some kind of language assumptions about text. It presumably defaults to your default language chosen when you installed it. For any text selection, you can both tell what language it is assuming, and override it, by going to Tools > Language... and see/set the language there.
Sign In or Register to comment.