Things enabled by this project

We can finally start ticking these boxes, and fix bugs as they appear (infrastructure almost usable!)

OSB/geppetto: discuss with @adrianq @tarelli

• no need to parse neuroml models twice (as nml and as lems). We get org.?ml.model.DomainType extends org.lemsml.model.Component
• plotting arbitrary expressions. Components have a scope.evaluate(symbol) method.
• rendering equations as mathml. There are methods to render expressions as mathml(tex), C, java, python (and anything else we care to implement).
• getting lists of cells/networks/channels independently of file names (i.e. na.channel.nml). There are methods to get all components of a given type (including extensions).

Elsewhere:

• automatic generation of API / xml schema directly from Lems Component Types. No more duplication of work maintaining defs and a xsd! See here
• code generation via a "proper"(room for a lot of improvement) semantic model for Lems, with a well defined pipeline (visitors) for producing intermediate representations and merging them with templates.

This issue is a broad roadmap covering equation rendering, API/XML-schema generation, and code generation; it names no files, tests, or entry points. Start by reading the project structure and the linked neuroml2model reference, then clarify which unchecked capability is in scope and what completion criteria apply.