Re: replacement for pfx2ttf.fontforge
- From: Werner LEMBERG <wl gnu org>
- To: hanwen lilypond org
- Cc: gtk-i18n-list gnome org, lilypond-devel gnu org
- Subject: Re: replacement for pfx2ttf.fontforge
- Date: Fri, 29 Dec 2006 15:51:57 +0100 (CET)
> > I can add the I+J->IJ stuff for OpenType script `latn', language
> > `NLD ' (Dutch), if you want me to do that. Similarly, I can
> > disable the f+i->fi stuff for languages `TRK ' (Turkish) and
> > `PTG ' (Portuguese) -- I'm not sure about the latter, however.
>
> I thought that the idea was to put this knowledge in Pango rather
> than in the font, so font coders don't have to duplicate this info a
> zillion times for each font.
This is tricky. Think of the Turkish case where you must not have
f+i->fi because this language has a dotless i also. Instead, there
might be a hypothetical rule
f + i -> f + <small horizontal space> + i
within the font to pronounce visibility. This small horizontal space
is definitely font dependent.
I think that ligature rules which actually produce a visible result
(this is, a ligature glyph) should be within the font -- this is done
so in both Type 1 and OpenType fonts.
On the other hand, input ligatures as provided by TeX like
` + ` -> <doublequote glyph>
should not be part of the font.
Behdad, what do you think?
Werner
[
Date Prev][Date Next] [
Thread Prev][Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]