Snake case conversions across python files

[greenLeopard]

This PR updates all of the notebooks and python modules to use snake case #34 I re-ran everything I could to test that it was still working. On Jupyter notebooks, where possible, I updated the code to snake case and saved the code changes, then test ran the notebook without saving the new notebook outputs - this largely prevents the big diffs. Occasionally, I did save all the outputs, and you get the big diffs. Sorry about that.

A couple of obstacles I had when re-running code after changing to snake case:

1. examples/add_no_overlap_2d_parameters.ipynb - I get an AssertionError (assert status == cp_model.OPTIMAL) with pack() instances that have scale=10_000, so I set scale=1000 and ran the notebook to completion.
2. evaluations/tsp - I didn't rerun the benchmarks because they rely on a comparison with gurobi, and I don't have access to a gurobi license.
3. evaluations/embedding_cpsat & evaluations/optimization_api - I only made modest changes to the cp_model function calls, and didn't rerun these.
4. In examples/add_no_overlap_2d_parameters.py commented out the solve parameter use_pairwise_reasoning_in_no_overlap_2d as it wasn't found.

There are a couple of additional small fixes:

• There was one remaining place where the conjugation of proof needed to be updated: proving/proofing and proved/proofed.
• The constructor for PiecewiseConstantConstraint, the default value of parameter restrict_domain has been changed to False. (This aligns will all usages of this class.)
• In utils/piecewise_function_approximation.ipynb I named the arguments to PiecewiseConstantConstraint and PiecewiseLinearConstraint to avoid an error that had occurred on parameter ordering.

Let me know if you need any adjustments.

[d-krupke]

Looks perfect, accepted! Thank you so much and have a great Sunday :)