Re: Decomposition of Indic vowels

From: Kenneth Whistler (kenw@sybase.com)
Date: Fri Oct 30 1998 - 19:10:54 EST


>
> I processed Unidata 2.1.5 (using the "ex" editor, plus sorting!)
> to produce a set of properties for the currently undecomposable Indic
> vowels. I excluded Thai, Lao, and Tibetan, which don't seem to
> work on the same principles.
>
> This is not a proposal, just a specimen. I don't have the
> necessary knowledge to make it a proposal.
>
> 093E;DEVANAGARI VOWEL SIGN AA;Mc;0;L;<vowel> 0906;;;;N;;;;;
...

What John Cowan is doing here is trying to express the relationship
between the dependent vowels:

093E;DEVANAGARI VOWEL SIGN AA;Mc;0;L;;;;;N;;;;;

and the corresponding independent vowels:

0906;DEVANAGARI LETTER AA;Lo;0;L;;;;;N;;;;;

While this is an interesting and valid relationship, it is
not a decomposition, and is not what the decomposition field
of the Unicode Character Database is intended for.

It would make more sense to simply publish a two-column chart
that shows the numerical relationship between all the independent
and dependent vowels in the Brahmi-based scripts in Unicode
(including those coming in Unicode 3.0). This would include
pairs that currently *do* have decompositions in Unicode, such
as:

0D4A;MALAYALAM VOWEL SIGN O;Mc;0;L;0D46 0D3E;;;;N;;;;;

and

0D12;MALAYALAM LETTER O;Lo;0;L;;;;;N;;;;;

or

0BCC;TAMIL VOWEL SIGN AU;Mc;0;L;0BC6 0BD7;;;;N;;;;;

and

0B94;TAMIL LETTER AU;Lo;0;L;0B92 0BD7;;;;N;;;;;

--Ken



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:42 EDT