Adding type hints and updating doc on globalconstraints

[IgnaceBleukx]

Started going through the global constraints file to update the documentation similar to what was done for the global functions.

• Adding type hints to the constructor
• Adding return types for the value/decompose functions
• Uniform docstring for the class
• Checking if an argument is None while computing the value

Still work in progress

[IgnaceBleukx]

Finished documenting the globalconstraints.py file and updated checks.
Lots of the tests were actually outdated, and do not follow the CPMpy spec... I.e., you are always allowed to use Booleans as integers. Question remains if all of these tests will still work after #780 is merged, I can see some solver interfaces not supporting this currently.

I also removed the call to flatlist in the constructors, it should not be necessary...

Final open point is regarding explicit type checking. E.g., should we check all arguments of a list are indeed the right type? It would be good to do so... but will add some boilerplate code. Although I'm not opposed to it, as all of that checking code is hidden in the constructor anyway, and we have a great friend who can help writing with these checks.

Feedback on doc changes and thoughts on typechecking are welcome : )

[tias]

Great improvement, esp from a documentation point of view.

More precise typing (e.g. Expression can also often be an int or a np.int) to be worked out in subsequent PRs, e.g. #768