[XeTeX] Issue with Arabic Typography

[Jonathan Kew]

On 24 Feb 2009, at 16:44, maxwell at umiacs.umd.edu wrote:

>> Jonathan Kew wrote:
>>
>>> A complete solution would involve support for the OpenType
>>> justification table, although not all fonts provide this; it would
>>> apply to more than just Arabic script;
>>
>> Yes, it would be lovely for Syriac text and poetry as well!
>
> Does anyone (Jonathan K?) know whether this technique is used for
> justification in Urdu as well?  Urdu is written in the Nasta'liq (aka
> Nastaleeq) version of the Perso-Arabic script, which is quite  
> different
> from the Naskh version that most other languages written in the
> Perso-Arabic script use.

Well, you clearly can't justify Nastaliq script by adding straight  
"extenders" between the letters, that would be unspeakably ugly. But  
the concept of kashida (elongation) is certainly valid; a calligrapher  
will use longer or more compact forms of various letters as necessary  
in order to fill the line perfectly.

(This would also be true for high-quality Naskh and other styles; the  
use of straight extender glyphs is a "poor man's" version that can  
work with the highly regularized, simplified styles of typical fonts,  
but traditional Naskh calligraphy is much more complex than this and  
would present similar challenges to Nastaliq.)

In principle, a font could provide collections of such "justification  
alternates", and control their use with the AAT, OpenType, or Graphite  
justification layout tables, but I am not sure whether anyone has  
actually implemented this in a current font. Nor do I know whether  
there are applications that would be able to make use of it.....  
perhaps the middle-eastern versions of some high-end software?

JK