PYTHON: implement special methods __lt__, __le__, __gt__, __ge__, __eq__ and __neq__

Non-numerical IODE objects

The IODE databases for Comments, Equations, Identities, Lists and Tables are essentially dictionaries {name: object}.

IODE Comments and Lists are just strings. For the other databases, dedicated classes Equation, Identity, Scalar and Table has been implemented both in C++ and Python. They have their own implementation of __eq__.

It would be nice to override the special methods __eq__ and __neq__ for the databases too.
The question is, what kind of object to return ?

A very naive implementation would be to return un simple Python dict(str, bool) and use this dict in __getitem__ and __setitem__. A second possibility is to return a pandas Series. A third possibility is to implement dedicated classes Bool[Comments|...|Tables] but with which attributes and methods ?

Numerical IODE objects

For the IODE Scalars and Variables, it would be nice to have also implementation for __lt__, __le__, __gt__ and __ge__.

The case of Variables is tricky since it represents 2 dimensional data: (dimensions = names and periods).
One possibility for the Variables is a Python class BoolVariables with 3 attributes: the list of names, the sample (first_period:last_period) and a boolean 2D Numpy ndarray. Another possibility is to have a C++ version of BoolVariables and use Cython to wrap it.

@gdementen @yvdabb Any thoughts or advices would be welcome