Just to help you understand (hopefully).
Think of it as wrapping math* tables with "{" "}".
The "math" nature is already implied when using align(*) environments.

/PA

On Mon, 16 Nov 2026 at 09:09, Pedro A. Aranda <[email protected]> wrote:

> Updated version after a git pull in main...
>
> On 16/11/26 7:56, Pedro A. Aranda wrote:
> > Just trying to wrap up and a new version of the patch with minor
> > modifications.
> >
> > On 15/11/26 19:30, Ihor Radchenko wrote:
> >> "Pedro A. Aranda" <[email protected]> writes:
> >>
> >>>> There is one question you did not address though - about all the dance
> >>>> done inside `org-latex--wrap-latex-matrices' with tables that follow
> >>>> one
> >>>> another being merged under single "matrices" umbrella.
> >>>> If we do not wrap the tables into \[...\] or \(...\) what is the point
> >>>> wrapping those tables into (matrices ...)?
> >>>>
> >>> Real life came in between...
> >>>
> >>> align and align* are something I need for ltx-talk slides, because they
> >>> are the only ones that I know of with a reliable support for screen
> >>> reader adaptation.
> >>>
> >>> You might be right, but then, that would be applying the scalpel in the
> >>> very heart of the export backend and that is something I'm not capable
> >>> of doing right now with the next semester around the corner.
> >>
> >> Then, maybe you should simply drop math* from "matrices" processing?
> >>
> >
> > My findings/line of thought
> >
> > 1. I'm using align/align* to generate groups of aligned equations. They
> > work in principle like eqn-array and friends, that is, they expect a
> > table with cells separated by '&', each cell being a math expression.
> > Why these two? Because I know them to be nicely adapted to accessibility
> > features I *must* use. Their only drawback is they don't like to be
> > surrounded by either \[\] or \(\).
> >
> > 2. AFAICR, matrices are generated in (org-latex--math-table). BTW: there
> > was a little quirk here, which I'm correcting in the attached version of
> > the patch that supersedes the original version.
> >
> > 3. (org-latex--math-table) is called by (org-latex-table). If there were
> > a way to control how tables are finally shipped/wrapped, this would be
> > the point, but there isn't.
> >
> > 4. Part of the wrapping magic is done in (org-latex--wrap-latex-
> > matrices). For our purpose, math* behaves like inline-math, but has an
> > "empty wrapping" as opposed to being wrapped into a "\(" "\)" [1].
> > Therefore, it's here that I set the :markup to 'math-star'.
> >
> > 5. I consume the :markup in (org-latex-matrices) by returning the
> > unmodified contents. If align or align* needed a specific wrapping, it
> > would be here that I'd add it, just like the `inline or the _ case in
> > the function (_ standing for `math).
> >
> > Now I really need to cope with real life.
> >
> > /PA
> >
> > [1] I know this may sound confusing, but empty stuff exists.
>
> --
> Pedro A. Aranda
> ox-latex & ox-beamer maintainer
>
> Sagen's Paradeiser! (ORF: Als Radiohören gefährlich war) => write BE!
> IA: onomatopeya del rebuzno
> 2nd year of the New Koprocracy
>


-- 
Fragen sind nicht da, um beantwortet zu werden,
Fragen sind da um gestellt zu werden
Georg Kreisler

"Sagen's Paradeiser" (ORF: Als Radiohören gefährlich war) => write BE!
Year 2 of the New Koprocracy

*IA: onomatopeya del rebuznar*

This was produced by a human (implied virtues and weaknesses acknowledged)
Don't feed your ArtStup (SKA AI) with this. It won't make YOU smarter...

Reply via email to