Fix MjSpec introspection with visual.rgba, visual.headlight, and visual.global_.

[AaronYoung5]

Fixes issue mentioned here in #2289.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[yuvaltassa]

What about the other members of mjVisual (map, scale, quality)?

Can you please do the CLA thing?

[yuvaltassa]

Also, thank you! 😊 🙏

[AaronYoung5]

rgba/headlight have arrays (which map/scale/quality don't), so I think those types need to be explicitly defined in the generated code which calling _binding_code on rgba/headlight does.

CLA is done. Thanks!

[yuvaltassa]

Got it, makes sense.

Question: are you sure we need the underscore suffix? I.e. did you try not adding it and got an actual error? Sometimes the syntax highlighting will make you think you need it but global is actually OK.

[AaronYoung5]

Yeah, I get the following error.

$ python -c 'import mujoco; print(mujoco.MjSpec().visual.global)'
  File "<string>", line 1
    import mujoco; print(mujoco.MjSpec().visual.global)
                                                ^^^^^^
SyntaxError: invalid syntax

EDIT: after looking into it a bit, seems like all python keywords (i.e. help("keywords")) are not allowed to be used as identifiers. And PEP-8 says to append an underscore to the name if it collides.

Also, thanks for the quick responses!!

[quagla]

Great, thanks for doing this! Is there anything that prevents you to put the code inside _value_binding_code like I did for e.g. mjStatistic? I don't think the top-level function of the binding generation makes it easy to maintain the code. That would also make it easier to generalize it to arrays.

[AaronYoung5]

Thanks for taking the time!

I don't think _value_binding_code has access to the subfields, which are the array types that need to be defined explicitly, right? Like since only the ValueType is passed, it can't iteratively define the arrays. The latest commit might be more maintainable, as you suggested

[yuvaltassa]

Can you remove the trailing spaces please?

Lines 986, 988 and 989 in specs.cc

[AaronYoung5]

Done