[XeTeX] Old Persian

[Jonathan Kew]

On 10 Sep 2004, at 5:00 pm, Yves Codet wrote:

> A late reply accompanying a new question. Yes, the above code gives 
> the expected result with XeTeX. And the question is: is it possible to 
> create a key layout with KeyLayoutMaker for syllabaries on higher 
> planes? If so what should be input to KeyLayoutMaker, code points or 
> surrogate pairs? I was thinking of Cypriot (of which a friend of mine 
> is a specialist) and I was wondering if the input file should look 
> like this:
>
> a	10800
> e	10801
> i	10802
> ...
>
> or this:
>
> a	D802 DC00
> e	D802 DC01
> i	D802 DC02
> ...
>

Hello Yves,

I wouldn't recommend KeyLayoutMaker for this job--first, because it 
doesn't currently support sequence output or non-Plane 0 characters 
(sorry; that wouldn't be a particularly difficult enhancement, though); 
and second, because its focus is on layouts with large numbers of 
dead-key sequences, such as the Ethiopic or Cree examples. So I think 
it will generate a dead-key state for every keystroke that's used in 
the input, even though in this case they're unnecessary.

If I get a chance, I may look at extending KeyLayoutMaker to deal with 
these issues. (For that matter, it's just a Perl script, anyone is 
welcome to take a look and contribute improvements!)

For a simple layout where single keystrokes generate output directly 
(either single characters or sequences), I'd suggest you look at the 
GUI tool Ukelele <http://scripts.sil.org/ukelele>. In conjunction with 
the Character Palette, it should be simple to put the Cypriot 
characters on whichever keystrokes you wish.

Jonathan