priyankoo,

i am interested in the assamese keyboard.  but i am more interested
in assamese fonts.  my concern with using an unmodified bengali
font is that the underlying code might be wrong.  for example, in
unicode "pra" in bengali is written as "pa"+hasanta+"ra" where the
"ra" is the bengali ra.  if the bengali unicode font has no table
entry to convert a "pa"+hasanta+"assamese ra" to "pra", assamese texts
written in this font will be wrongly coded, even though it will be
look right.  the problem recurs with the "reph" formation. 

i just visited your blog and see that this is indeed the problem.
the word "st petersburg" has the "bengali ra" encoded.  this issue
might have a bearing on natural language processing.  does chandan 
talukdar address this problem in his keyboard?

-- 
saurav

_______________________________________________
assam mailing list
[email protected]
http://assamnet.org/mailman/listinfo/assam_assamnet.org

Reply via email to