Reduce the amount of data we're reading from FCstd files

[martinRenou]

The method to translate a freecad object into a jcad object reads all the properties of the freecad object.

This is a bit too much because many of these properties cannot be understood by the JupyterCAD client.

We should reduce this to reading only the properties that JupyterCAD understands, making use of the Python schemas jupytercad-core now exposes.

[trungleduc]

For properties that don't have a handler, we just set their value to None. Why do we need to filter more?

[martinRenou]

But many properties have a handler even though it's not supported by jcad ? At least that's what we see with exported files using #25

Exported files do not respect the jcad schemas so we can't open them in jupyterlite. They should respect the schemas

[trungleduc]

These properties go into the parameters key of the IJCadObject interface, which is defined as { [k: string]: any;}, I'm wondering how it can break the schema

[martinRenou]

I see, but why load those properties if the client can't understand them? That's probably early optimization but there are lots of properties we don't understand and it's useless to load.

[martinRenou]

Adding a note that those unwanted data results in doubling the size of the exported ArchDetail.jcad file.

[trungleduc]

Because JupyterCAD was designed with FCStd as a first class citizen, the idea was to eventually support them all

[martinRenou]

Sounds good! But we should maybe add the properties we support progressively in our schema