Python Bindings for base engine retreival are insufficient

[tboudreaux]

Currently the python bindings to access the base engines of either EngineViews or to access the engine stack from a Policy are insufficnent. There is a huge amount of boiler plate needed for what should be a simple operation. Further, there is a bug in getBaseEngine() on most engine views due to an incorrect return value policy.

Consider the following code

AEngine: AdaptiveEngineView
QSEEngine : MultiscalePartitioningEngineView = AEngine.getBaseEngine()

This crashes with the following error

Traceback (most recent call last):
  File "/Users/tboudreaux/Programming/4DSTAR/GridFire/validation/vv/GridFireValidationSuite.py", line 117, in <module>
    instance(args.pynucastro_compare, args.pync_engine)
    ~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/tboudreaux/Programming/4DSTAR/GridFire/validation/vv/GridFireValidationSuite.py", line 35, in __call__
    self.evolve(engine, netIn, pynucastro_compare = pynucastro_compare, policy=policy, depth=EngineNameToDepth[pync_engine.lower()])
    ~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/tboudreaux/Programming/4DSTAR/GridFire/validation/vv/testsuite.py", line 248, in evolve
    l_engine = l_engine.getBaseEngine()
RuntimeError: return_value_policy = copy, but type is non-copyable! (#define PYBIND11_DETAILED_ERROR_MESSAGES or compile in debug mode for details)

The issue here is that we are returning a reference without setting the proper return value policy in pybind.