Module talk:inc-pra-Deva-translit

Missing diaeresis after vowel letter
अउअ is being mistransliterated as 'aua' rather than as aüa, with the result that its declension using fails. It's also failing in Module:Brah-translit, so I suspect we have a general failure with initial aï and aü. Lets populate the testcases before we try fixing the code. --RichardW57 (talk) 21:29, 9 May 2022 (UTC)
 * I just copied the code  from other modules and don't quite understand how it handles the diaeresis, so adding other testcases before trying to fix it is probably a good idea.  Regarding the incorrect declension at, see also . Kutchkutch (talk) 01:06, 10 May 2022 (UTC)
 * c = consonant letter, d = optional dependent vowel, e = vowel letter. I think the options for e should be reduced to vowel letters I and U, unless we think we're likely to come to want to mark other vowels as well.  We could then simplify the code.  The code adds a combining diaeresis where appropriate in a coding style displaying disdain for anyone just using the basic editor provided by Wikimedia / Wiktionary. RichardW57m (talk) 12:24, 11 May 2022 (UTC)
 * I've now added handling, in a separate call, code for a = letter A or AA, b = letter I or U. I believe errors in declension have now vanished at the example above.  I've used a table diatrema for vowel plus diaeresis, which I hope is more readable.  I've also added testcases, which have now changed from fail to pass.
 * It will be at least a few hours before I fix the Brahmi transliterations. What other clones of (c, d, e) are there? RichardW57m (talk) 12:24, 11 May 2022 (UTC)

I've reduced the scope of the (c,d,e) so it only consumes something for e if that something is इ or उ. It was needed for वाअउ (Pischel §378, and findable on the web in Devanagari Prakrit text). --RichardW57 (talk) 04:33, 2 June 2022 (UTC)
 * Correction: The text I found the word in was a grammar written in Sanskrit, and mere mentions rather than uses. --10:14, 2 June 2022 (UTC)