Indic fonts

[Rudra Banerjee]

Hi,
Is there any plan to support indic scrips?

I found only few appearance of this topic in Google search(one of them was by me on 2012), and hence clearly not a priority.

One clue is gedit can do it fine, which is rather minimalist editor.

[Tony Mechelynck]

As you probably know better than I do, the problem with Indic scripts,
especially within a fixed-cell character grid like Vim's, is how to
make the transition from the sequence of individual consonantic and
vocalic Unicode codepoints to the sequence of glyphs to be presented
in the successive screen cells. A similar but maybe simpler problem
was solved some years ago about the Arabic script by Nadim Shaikli:
Arabic is a very cursive script where each consonant can have up to
four shapes depending (I'm simplifying) on its position within a word,
and in addition there are established separate glyphs to be used OT1H
for the sequences LLH (lillāh) and ALLH (Allāh) and OTOH for the
sequence LA (laam-alif) which is treated as a single letter. Nadim
solved it for the Arabic language but IIUC it could be used for Urdu
with hardly any change.

I think Bram wouldn't be adverse to accepting a new subsystem for
Indic scripts if someone (probably a programmer from the Indian
subcontinent) could develop it and show that it is indeed usable to
let Vim handle the family of Brahmi-derived scripts (including of
course devanāgari but not limited to it); but no one so far has
responded to the challenge. If you are able and willing, or if you
know someone who is, feel free to try.

Best regards,
Tony.

[Rudra Banerjee]

Thanks for your reply. I would have love to make my hand dirty, except I have minimal idea of how the fonts are displayed. Like, I thought, it should not be a problem with encode=utf-8, but clearly ots not. As you said, probably due to fixed cell grid of vim.
Being said that, I have started looking at how gedit has solved the problem. But most probably, they don't use fixed cell grid as vim/gvim.
Anywhere to start? Can you give me any suggestions?

[Tony Mechelynck]

Knowing as little as I do about Indic writing, I would have a look at
some encyclopædic sources describing what they are and how they are
displayed, but maybe you can skip that, I don't know. I would also
have a look at which Unicode codepoints are available for Indic
scripts (see the third column of http://www.unicode.org/charts/ ) and
I might perhaps also have a look at src/arabic.c to see how Nadim
Shaikli solved the problem of rendering one single codepoint by one of
several glyphs depending on context — the latter might or might not
give inspiration to someone having (or wishing) to tackle the problem
of Indic scripts. Further than that, you're probably better placed
than I am to determine how to go about, considering that Mechelynck is
a Belgian family name (i.e. from a country where everything used to be
written in Latin script, though Greek, Cyrillic and Maghreb Arabic
scripts have made an apparition in the latest decades) while
Ban(n)erjee is an Indian one. ;-)

Best regards,
Tony.