Re: kn-itrans.mim: About the updates in https://github.com/indic-transliteration/m17n-db-indic

[Vishvas Vasuki]

Thanks - 

I suppose that for kn, it is better to forgo
independent vowels LLi = ಌ and LLI = ೡ

but retain  ("LLI" (delete @-) "ೣ") and ("LLi" (delete @-) "ೄ").

The reason is that kaLLi and aLLi should intuitively yield ಕಳ್ಳಿ and ಅಳ್ಳಿ,
while kLLipti should yield ಕೢಪ್ತಿ (since ಳ್ಳ್ after a consonant is very unusual). 

As a simple unambiguous option, it would be good to retain ("LLL" "ಳ್ಳ್")

For devanAgari

LLi ऌ and LLI ॡ should be retained as well - I suppose there is no way around retaining and relying on LLL.

These should be recorded as comments in the files - lest we fall into the same thought cycle after a few years!

On Tue, 22 Aug 2023 at 15:49, Mike FABIAN <mfabian@redhat.com> wrote:

Mike FABIAN <mfabian@redhat.com> さんはかきました:

> "विश्वासो वासुकिजः (Vishvas Vasuki)" <vishvas.vasuki@gmail.com> さんはかきました:
>>>
>>> Isn’t it more natural to type only two "L" to get two "ಳ್"?
>>>
>>>
>> Yes it is.
>>
>> In case of devanAgarI, however, ळ्ळ् (LL) is rarer than ऌ (LLi) and ॡ
>> (LLI).
>> Is there a way to ensure that all three work in both kn and hi (that would
>> be ideal)?
>
> I think there must be a way to do that, I’ll try.

How about this?:

https://github.com/indic-transliteration/m17n-db-indic/pull/19

Please test it, I am not sure, but it seems to work for me.

This screenshot shows the behaviour after applying only the first patch

https://github.com/indic-transliteration/m17n-db-indic/pull/19/commits/6d5a3b485f0f71fdcb7e4eb2f160d2ae10e7f2ed


The only difference I can see to
http://aksharamukha.appspot.com/converter/ is that kLLI yields just ಕ
because ("LLI" (delete @-) "") just removes the U+0CCD KANNADA SIGN
VIRAMA from the ಕ್.

But in http://aksharamukha.appspot.com/converter/ the LLI not only
removes the VIRAMA but also adds U+0CE3 KANNADA VOWEL SIGN VOCALIC LL.

If I add this second patch:

https://github.com/indic-transliteration/m17n-db-indic/pull/19/commits/3610208c788772bedd88faa4c11f5cb58a2a434c

then kn-itrans.mim behaves the same.

Is that OK?

--
Mike FABIAN <mfabian@redhat.com>
睡眠不足はいい仕事の敵だ。

--

--
Vishvas /विश्वासः