Allow Indirect Field Accesses in Proto Templates

[CoolSpy3]

Description
Allows use of the raw fields variable in proto templates.

Related Issues
This pull-request fixes issue #6566.

Tasks
Add the list of tasks of this PR.

• Allow the use of the fields variable in proto templates.
• Add tests for the new behavior
• Update the documentation (By my reading of the current documentation [lua] [javascript], the new behavior is already described.) Document the new tag
• Clarify proto regeneration behavior
• Update the changelog

Documentation
Procedural PROTO Nodes (Lua)
Procedural PROTO Nodes (Javascript)
("Programming Facts" and "PROTO Regeneration" sections)

[CoolSpy3]

The CI failure does appear to be reproducible on this branch and not master, so I'm marking this PR as a draft until I have time to investigate it.

[CoolSpy3]

It looks like the Car proto is being regenerated as a result of these changes, causing the controller to restart. Thus, the test supervisor thinks there's three controllers running, but it only receives completion reports from two of them. I still have to find a good workaround. The only real solution I can think of is gating the behavior behind a tag. (i.e. # tags: fieldsObject/# tags: indirectFieldAccess [open to suggestions on the name]). However, if someone else has another idea, feel free to let me know.

Side note: In general, proto robots having their controllers restarted when they're regenerated seems a little unintuitive, however, any proposed behavior I can think of has some annoying edge cases. I think the best option is just to make a note of it in the proto docs. If others feel that we should update this behavior in some way, one of us can open a new issue with the proposed behavior.

[CoolSpy3]

Note: It's hard to tell from the diff, but WbProtoModel:219 (Now 227) and WbProtoModel:281 (Now 292) are not identical to their pre-edit counterparts.

[omichel]

It looks good to me.

[omichel]

Thank you.