search

slatex / LaTeXML-Plugin-sTeX

A LaTeXML Plugin for Semantic LaTeX (sTeX)
LaTeX Project Public License v1.3c
2 stars 3 forks source link

expected behavior of vardef? #132

Open Jazzpirate opened 4 years ago

Jazzpirate commented 4 years ago

https://gl.mathhub.info/smglom/algebra/blob/master/source/quotient-group.de.tex#L3 causes malformed:omdoc:symbol <omdoc:symbol> isn't allowed in <omdoc:definition>. The current binding for \vardef is ######## temporal hack by MiKo; @DG,please remove after implementation. Let('\vardef', '\symdef');

so should \vardef (ultimately) be allowed in a definition? If so, should it yield something other than a omdoc:symbol or does the schema need to be adapted?

kohlhase commented 4 years ago

Hmm, this is really one of the unresolved OMDoc/sTeX language design issues, that needs to taken care of now to make progress. I will use this place of develop my thoughts.

Background: I noticed in sTeX/SMGloM that I needed something like local symbols with complex notations in definitions. \vardef is just the user interface, since mathematicians (I believe) think of them as variables. And an implementation as \symdef just fit the bill.

The situation in quotient-group is typical, you need a structure S=<a,b,c>, and you want to say things about the components in the definitions later. Which would be simple, if they were all individuals, but some of them are functions, so I need notations on them. Hence \vardec.

I wonder if what I really want here is a structure. In MMT I would (non-modularly write down) 

a: t
b: t
c: t -> t

hmmm, I am confused, I need a whiteboard and a discussion partner.