Solution for Extended Tamil

[James Kass]

On 2024-01-22 8:07 PM, Richard Wordingham via Unicode wrote:

 > Using GPOS for re-ordering seems to be a nightmare.

Which is why I didn't try it here.  You have more experience with modern 
typography tools than I do, so I would defer to your better judgment for 
possible approaches.

 > Where do you get this 'non-reordering' property from?

The canonical comb. class = 0, which means non-reordering.

Note that in the text quoted earlier from 
https://en.wiktionary.org/wiki/Module:sa-convert/testcases/Tamil it is 
said that "most forms of Extended Tamil...".  This suggests that there 
are other conventions which place the superscript digits elsewhere.  If, 
as suspected, an existing convention places the superscript digit at the 
end of the syllable, then the /de facto/ encoding sequence and default 
display might well be totally acceptable to the user community.