improve docstrings in stages and operators

[VianneyMI]

• Copy what was done later on when interfacing Atlas Search. That is adding a description of the operator/stage
• Add link to official mongo doc

[VianneyMI]

To improve enums docstrings => https://stackoverflow.com/questions/50473951/how-can-i-attach-documentation-to-members-of-a-python-enum

[VianneyMI]

https://betterprogramming.pub/3-different-docstring-formats-for-python-d27be81e0d68

[VianneyMI]

https://numpydoc.readthedocs.io/en/latest/format.html#documenting-classes

[OlivierLegrand]

• merge 59 dans 59 bis
• finish 59 bis

[OlivierLegrand]

In pipeline.py:

• in replace_root and replace_with methods docstrings: "For more info on expressions, see Expressions." Where this should point to? There is no link atm but I think there should be one?
• in all method docstrings: "Adds a ... stage to the current pipeline": should it be the monggregate stage or its MongoDB counterpart (when it exists) ? For instance, in search_meta: should it be "Adds a search_meta stage..." or "Adds a searchMeta stage..." or "Adds a $searchMeta stage...".
• Same question in set stage. Currently, it is a mix of both:

Adds a set stage to the current pipeline. Adds new fields to documents. $set outputs documents that contain all existing fields from the inputs documents and newly added fields. Both stages are equivalent to a $project stage that explicitly specifies all existing fields in the inputs documents and adds the new fields.

[OlivierLegrand]

Made the following changes in pipeline.py:

• In the Examples section of the pipeline class doctoring, changed

Examples
--------
You can instantiate a pipeline instance as follow:

    >>> pipeline = Pipeline(collection="listingsAndReviews") # collection is the only mandatory attribute 
                                                            # [example using MongoDB AirBnB demo dataset](https://www.mongodb.com/docs/atlas/sample-data/sample-airbnb/#std-label-sample-airbnb)
and then add stages to the pipeline by calling its wrapper stages method as shown below:

to

Examples
--------
The following examples use the example using MongoDB AirBnB demo dataset.

You can instantiate a pipeline instance as follow:

    >>> pipeline = Pipeline(collection="listingsAndReviews") # collection is the only mandatory attribute 

and then add stages to the pipeline by calling its wrapper stages method as shown below:

for a better rendering.

[OlivierLegrand]

• virer expression partout (pipeline et stages)
• séparer dans la doc de pipeline les méthodes associées aux stages, et les autres. (les grouper ensemble)
• mettre privé tout ce qui est décoré par pyd.validator (n'apparaîtra plus dans la doc)
• faire une PR
• quand la PR est validée, recréer une nouvelle branche à partir de l'issue.