Assamese

[scossu]

Add Assamese support via Aksharamukha.

Note that AM has ALA-LC support for Assamese, but it does not show as 100% accurate from tests (see attached).

Assamese Romanization Testing.docx

[virtualvinodh]

I had a look at the issue. It's because the ALA-LC table for Assamese doesn't include a 'v' but rather 'w'.

[virtualvinodh]

Aksharamukha tries to guess /y/ and /ẏ/ contextually by default. You'd have to disable it and enable strict transliteration by activating "Preserve source".

[scossu]

Thanks for looking into this, Vinodh.

There is a way to set Aksharamukha options on a per-script basis in the configuration. Are you suggesting that I should turn "preserve source" on by default for the whole Assamese script (vs. leaving the option up to the user on a case-by-case basis)?

[scossu]

Assamese is now available for LC transliteration in Aksharamukha and integrated in SS. However, the test samples attached in this ticket do not show how the expected transliteration should be. Hence, leaving open and on hold until we get more input (it has been solicited).