Add OptimizationPyCMA Wrapper Package

[ChrisRackauckas]

Old versions of BlackboxOptimizationBenchmarking.jl show that this was a rather good algorithm https://github.com/jonathanBieler/BlackBoxOptimizationBenchmarking.jl/tree/v0.1.0 and thus it would be good to have this in the benchmarking suite and available for users of Optimization.jl.

It can be wrapped similarly to other blackbox optimizers like OptimizationCMAEvolutionStrategy https://github.com/SciML/Optimization.jl/blob/v4.3.0/lib/OptimizationCMAEvolutionStrategy/src/OptimizationCMAEvolutionStrategy.jl because it doesn't need any derivative information, so the wrapper does not need to be so difficult. Some starter code can be pulled from the old BlackboxOptimizationBenchmarking code https://github.com/jonathanBieler/BlackBoxOptimizationBenchmarking.jl/blob/v0.1.0/scripts/optimizers_interface.jl#L54-L67, though that should be updated to instead use the newer PythonCall https://github.com/JuliaPy/PythonCall.jl setup for the Python wrapping.