Simplifed/static DoenetML/SpaTeXt for use by CheckIt

[StevenClontz]

Issue to continue a conversation I had with @siefkenj and @dqnykamp today...

I am interested in deprecating https://spatext.clontz.org/0.3/ and adopting DoenetML or a DoenetML variant as the language used by the next version of https://checkit.clontz.org/. Failing that, I would like to reimplement the next version of SpaTeXt using similar tooling as DoenetML, so we can better support each other's projects, and make it as easy as possible for SpaTeXt to support DoenetML content.

Here are what I think are the hard specifications:

• [Near-]instant rendering of markup for authors in either a VS Code environment or in the browser.
• Conversion of markup to portable LaTeX and static HTML snippets.
• Support for including randomized data provided as a JSON blob, including both insertion of strings and support for control flow (if and foreach).

Here are my like-to-haves:

• A simplifed language with no more than a dozen elements. If authors stick to this set of elements, they get lots of promises of portability, accessibility, etc. If these elements are insufficient, then they make the concession to use advanced elements with the understanding that their content will be less portable. (If SpaTeXt remains a separate project, then this would be acheived through a plugin infrastructure; e.g. if you want an interactive diagram in your SpaTeXt document, then you'd need to use a DoenetML or Geogebra plugin.)

[StevenClontz]

Part of this conversation: @siefkenj mentioned it'd be a good thought experiement to consider what would the generalization of a PreTeXt document (or at least something that covers 80-90% of such documents). This is kind of what I have been working on with SpaTeXt, so I'll outline my vision here (using elements that I chose for the draft SpaTeXt 0.3 implementation).

• A root document <spatext/>. This could have several modes: a "book", an "article", etc.
  • A <title/>.
  • Several <division/>s. Depending on the mode of the document, these might be chapters, sections, etc.
    • A <title/>
    • An optional <intro/>
    • Either <content/> or several sub<division/>s
    • An optional <outtro/>
• Within <content/>:
  • Paragraphs <p/>
  • <list/>s of <item/>s of either <p/>s or a sub<list/>.
  • Knowls <knowl/> (unless this is the content of a knowl). Different modes support things like theorems, definitions, activities, figures, tables, etc.
• Within <knowl/>:
  • A <title/>
  • An optional <intro/>
  • Either <content/> or several sub<knowl/>s
  • An optional <outtro/>
• Within <p/>:
  • Text nodes
  • Math <m/> (also displaymath with <me/> or <m mode="display"/>)
  • Emphasis <em/>
  • URLs
  • Other stuff I'm forgetting right now, but here we're pretty close to existing PreTeXt support

I'm curious what obvious stuff I'm missing here. (For one: images.)