Re: Pango Unicode coverage

[Pablo Saratxaga <pablo mandrakesoft com>]

Kaixo!

And Happy New Year.

On Mon, Dec 24, 2001 at 03:48:31PM -0500, Owen Taylor wrote:
 
> > Devanagari ?
> Yes. (X backend only)
> > Bengali ?
> Yes (X backend only)

What about using OpenType fonts ?

> > Ethiopic No
> Yes (?)

Yes, it has no special requirements, it is a precomposed syllabic encoding
like CJK or precomposed hangul, so the only requirement is to have a font.

In fact there are three kinds of scripts:
1. one position per code (that is, accented letters are
   precomposed). Almost all languages using latin or cyrillic alphabet
   are in this case; also are here greek sxcript, hebrew script, kana,
   pre-composed hangul, all CJK (and yi and similars), ethiopic,
   precomposed canadian syllabics (used by inuit languages etc),
   georgian, armenian.
2. composing chars that only need correct positioning of two or more
   glyphs; the only difficulty is the proper positioning, but even a
   naive positionning can give readable results (a bit ugly, but readable).
   in this case are Thai, Tamil using TSCII fonts, latin cyrillic and greek
   using non precomposed chars.
3. complex scripts needing complete substion of arrays of encoded chars
   by arrays of glyphs, the substitution being dependent on the script and
   on the font.

The 3. is the most complex to support, and needs also the help of people
with good knowledge of the languages involved.

If I understood correctly, Open Type fonts have the complex glyph substitution
mechanism embedded in them, which makes it much simpler for applications,
as they can handle it more or less like in case 1.

But what exactly is needed in the pangop side in order to take profit of
OpenType fonts ?

Thanks

-- 
Ki ça vos våye bén,
Pablo Saratxaga

http://www.srtxg.easynet.be/		PGP Key available, key ID: 0x8F0E4975