Re: tex-numbers.el

[Paul Nelson]

Hi Arash,

Thanks, your comments were very helpful.  I implemented them (with
mild modification).

Your suggestion about incorporating the numbers into CAPF features
seems worth pursuing.  I didn't quickly spot a way to hook into
TeX-complete-symbol without rewriting it, but was able to add
annotations for M-x completion-at-point.  If you're curious, see the
updated file at https://github.com/ultronozm/tex-numbers.el.

Maybe such annotations are what you had in mind, but I find it much
more useful in practice to be able to narrow the completions using
(vertico/orderless/... and) the label number itself.  The interface
would thus be to start completing, and then type "(1.1)" to narrow the
completions to those featuring the label number (1.1).  I did hack
something like this together using consult (see around
https://github.com/ultronozm/czm-tex-ref.el/blob/91b824b62507471585331af15d3c78cc4b5c3c01/czm-tex-ref.el#L206
and attached), but don't know how to incorporate it directly into
CAPF.  A similar problem would be to provide an interface for
dynamically restricting CAPF inside \ref{...} to just lemma
environments, or just theorem environments.

> I think this is a matter of taste, but
>
>   (cond (clause true)
>         (t rest))
>
> is the same as
>
>   (if clause
>       (progn true)
>     rest)
>
> right?

Yes, I just like the balanced look of cond in minor-mode definitions.
I stole the idea from the source code of eglot.

Thanks, best,

Paul

numbers.png
Description: PNG image