[Indic-computing-devel] Generic transliteration rule format for Indian languages

[<jk...@Fr...>]

knr> In an attempt to understand how transliteration could be done for
knr> majority of the Indian languages with generic principles, I have
knr> studied some of the South Indian languages and of course Hindi
knr> and came up with a generic framework to specify the
knr> transliteration rules (in the form of a set of grammar rules) for
knr> tranaliterating one 'word'.

Some requests:

A) Could you compare your approach with the other transliteration schemes
   available on the internet?
   
   E.g:-

   - Varamozhi [http://varamozhi.sourceforge.net/] (Malayalam)
   - ITRANS [http://www.aczone.com/itrans/] (Many languages)
   - The scheme used by Yudit [http://www.yudit.org/] for indian
     scripts
   - others ... 

B) The font ``composition'' scheme reminded me of Graphite
   [http://www.sil.org/computing/graphite/index.htm],
   except that Graphite is more declarative and hence easier
   to program in.

C) If you are intending to give code away, could you please clarify
   the license (BSD, GPL, MIT, public domain, other)?

D) If you could make your converter into a library linkable by
   other applications, we can then think of using it as part of
   larger applications.  Standalone converters don't compose well
   with other components.

Regards,
Koshy
<jk...@fr...>