incorrect displaying of Arabic text

[James Samir Ismail]

Has anybody experienced a problem with getting Arabic text to display correctly 
in their scores?  Here's a
 snippet that contains properly formatted UTF-8 Arabic:

\score 
{
        \relative d' 
        {
                c^\markup { "العربية" } d e f g^\markup { "english" } f e d
        }
}

I was surprised when I built the snippet because although I get Arabic 
characters, they are in the wrong 
form. In Arabic, characters potentially have different forms depending on where 
they fall in a word - 
initial, medial and final forms. The Arabic word in the source above displays 
correctly when viewed as 
a UTF-8 encoding. However, when the snippet is compiled, only the initial form 
of the Arabic characters 
display, which is incorrect. 

I'm saving the source as UTF-8, and Arabic characters do appear - just not he 
correct ones.  Is there 
something I'm not telling Lilypond, or is this a bug?


Thanks,
Samir