On transliteration of Jihvamuliya & Upadhmaniya

[sridatta1]

Jihvamuliya and Upadhmaniya are represented using following

1. Distinct signs for both
2. Ardhavisarga
3. A rare individual case of Ardhavisarga and Rotated ardhavisarga.

Distinct signs for both sounds are seen and in
Brahmi, Kannada, Tibetan, Devanagari, Bengali ( Vedic Extns code points), Soyombo, Newa, Sharada, Mongolian ( please add if some scripts are left out)

Ardhavisarga-
Ardhavisarga is seen and is added in the Script Extns for following scripts based on evidence by us
Script_Extensions=Beng Deva Gran Knda Mlym Nand Orya Sinh Telu Tirh
This covers all Major Indic scripts ( including Sri Lanka) used for Sanskrit.
In Gujarat only Devanagari is commonly used for Sanskrit even in the Gujarati language, script books, thus did not find any evidence.

Same with the Buddhist countries of SE Asia where I don’t know if there are any signs ( maybe one has to check in Sanskrit Vyakarana manuscripts if they exist)

Not sure of usages in Extended Tamil as the texts where these signs are generally seen ( like Vedic , Vyakarana) are all Grantha.
But the ardhavisarga glyph is generic, I guess can be used even where they were not seen traditionally.

While converting from scripts where Ardhavisarga is used needs to be based whether the next chars are ka/ kha and pa/pha ( any joiners / space needs or ignored )

But when the sign entered and used standalone in texts, then in target script it can be approximated to just visarga. Thus there are additional complexities.
In Devanagari, Bengali and Kannada an option use Ardhavisarga instead distinct signs needs to added.

[sridatta1]

For Ranjana- there are two cases,
I have seen Newa like Upadhmaniya in a Manuscript from Nepal.
And Landza uses Tibetan style glyphs thus the same is supported in the recently released font
https://github.com/EkType/Nithya-Ranjana So that font can be used for Ranjana by default.

[sridatta1]

Just like other Indic script, in combination with the next consonants, in Tibetan script- Subscripts are to be used.
ྈྐ in this way instead of ྈཀ
Ref doc https://www.unicode.org/L2/L2009/09032-n3568.pdf

[virtualvinodh]

Thanks a lot for this!

The bug in Tibetan is now fixed. I'll have a look at the others.

V

[sridatta1]

https://www.unicode.org/L2/L2024/24128-kawi-jihwamuliya-upadhmaniya.pdf

[sridatta1]

Now I think for major scripts in the output you can use Ardhavisarga, for the remaining scripts you can just approximate to Visarga.
Script_Extensions=Beng Deva Gran Knda Mlym Nand Orya Sinh Telu Tirh

[sridatta1]

Ardhavisarga can be supported in Tamil ( Extended ) as well