Re: zsh/complist colours improperly handle multibyte characters

[Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>]

On Oct 21,  9:33am, Peter Stephenson wrote:
}
} Yes, unfortunately character-by-character analysis is built into
} completion at quite a low level.

I think this particular issue may be a lot more tractible than you
believe.  clnicezputs() already has the MULTIBYTE_SUPPORT branch
that seems also to be needed by clprintfmt() -- I could probably
make an attempt to translate, but someone who has actually worked
with e.g. ZMB_nicewidth() and re-composing of bytes into multibyte
characters could probably do it faster.

The other part seems a plain pattern compilation issue.  Neither of
these gets involved with comparing strings on the line to strings
in the candidates.  This is all on the tail end where we already
have the matches and we're displaying them.