Re: Better emoji support

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Cc: kevin.legouguec@gmail.com,  emacs-devel@gnu.org
> Date: Mon, 20 Sep 2021 15:50:49 +0200
> 
>     Eli> Btw, we could perhaps tweak the composition code to produce emoji
>     Eli> representation for the likes of U+231A followed by VS-16, even though
>     Eli> U+231A etc. aren't in the emoji script.  Remind me about this after
>     Eli> the basic emoji composition code is installed, and let's see if it
>     Eli> could be improved for those sequences.
> 
> ? U+231A has Emoji_Presentation=Yes.

Well, bad example.  U+21AA is better.

> In any case, donʼt you need at least one codepoint to be using the
> emoji font to do that? (Noto Color Emoji doesnʼt have a glyph for
> VS-16, so that canʼt be used)

My idea is to trick the font_range function, which verifies that the
sequence can be composed using a single font, to use an Emoji font
when it sees something followed by a variation selector, instead of
the font for the first character in the sequence.  For that, the rules
for Emoji sequences in composition-function-table should be anchored
on the VS-n codepoints (which I think is a good idea regardless).

> As another aside: lisp/language/japanese.el sets up some
> composition-function-table entries for the Variation Selectors
> already, so eg
> 
> U+231A U+FE0F
> 
> already gets composed.

Yeah, that's the old kludge, we may wish to revisit it at some point.