StarDict doesn't search correctly for Unicode's combining diacritical marks

[GoogleCodeExporter]

I am creating a dictionary for Bolivian guarani, which is a language which uses 
Unicode combining diacritical marks. See: 
http://en.wikipedia.org/wiki/Combining_character

For example, to form the letter ï̵ (an i with a diaeresis and a line through 
the middle of it), I have to enter ï, then press CTRL+SHIFT+U, then enter 335 
to overlay the line). StarDict does not search correctly for words which 
contain this character, unless I type the entire word completely. 

For example, if I search for the word ï̵pati, StarDict has a lot of trouble 
finding the word. 

If I search for "ï̵", StarDict returns:
ïru, ïta, ïti

If I search for "ï̵p", StarDict returns:
ampuy, arpa, arpay, aypa, etc.

If I search for "ï̵pa", StarDict returns:
ampuy, arpa, arpay, aypa, etc.

If I search for "ï̵pa", StarDict returns:
alpaca, Amparo, Ampato, Apa, etc.

Only if search for "ï̵pati" does StarDict find the word.

To test this, see the attached dictionary.

Original issue reported on code.google.com by amosba...@gmail.com on 24 Sep 2012 at 1:48

Attachments:

• gu-es.zip