[XeTeX] Xenotype font and Syriac scripting

[Jonathan Kew]

On 5 Jan 2010, at 18:50, Fr. Michael Gilmary wrote:

> Jonathan Kew wrote:
>> On 4 Jan 2010, at 21:52, Fr. Michael Gilmary wrote:
>> 
>>   
>> 
>>> Curiously, the "Kashida" code you've given before, Jonathan, breaks the scripting. But that's a minor concern.
>>>     
>>> 
>> 
>> This indicates (I believe) that the AAT tables in the font do not support the Zero-Width Joiner character (ZWJ, U+200D), which is needed to get the correct joining forms where the kashida rules are inserted. You could raise this with the font developer, who may simply not have considered this issue.
>>   
>> 
> 
> 
> I'll do that ... but when I opened the font via FontForge, without looking for any trouble, I got this warning:
> 
> The glyph named zerowidthjoiner is mapped to U+200D.
> But its name indicates it should be mapped to U+FEFF.

Odd. If that's an accurate report, it sounds like a FontForge bug. U+FEFF is ZERO WIDTH NO-BREAK SPACE, so I don't see any good reason why FF should want a glyph named zerowidthjoiner to be mapped there!

(Also, note that the mere presence of a glyph mapped to U+200D does not necessarily mean that the AAT tables are designed to handle it correctly.)

JK