[Freetype] how does freetype handle combining character for unicode

[Ziying Sherwin]

We installed php 4.2.2 with freetype 1.3.1 on our solaris 2.8 machine.  We 
are using function imagettftext to convert UTF8 string to image file for
Internet display. However, it seems that this function does not handle the 
combining characters properly.  For instance, the sequence of unicode 
characters U+0061(a) and U+0308 (combining diacritics) is display as two 
characters: a and a square indicating not displayable. The correct result should
be a character identical to U+00E4.

We are wondering where the normalization is handled in this case, freetype
library or inside imagettftext php function? Is it a bug in freetype library?
If not, how can we fix it? 

Thanks,

Ziying Sherwin

P.S. I am not on the mailing list, please send your response to 
address@hidden