Request: new optional `required: list` parameter on data fetcher functions

[mxkpp]

Currently, the method RasPlanHdf.mesh_faces_timeseries_output and similar silently do not include RAS variables if those variables are not available in the HDF file at the time of read:

rashdf/src/rashdf/plan.py

Lines 976 to 977 in 6b8f7ff

	if self.get(var_path) is None:
	continue

For example, optional output parameters that a RAS modeler may have chosen not to include when running the plan.

If a user wants to assert that certain variables are available, they have to first receive the xr.Dataset returned by the method and then assert that it has what they need. It would be convenient to be able to pass in a list of required variables, and raise a custom exception if any required variables are missing.

For example a new optional required parameter used like this:

hdf = rashdf.RasPlanHdf("foobar.hdf")
faces_timeseries: xr.Dataset = hdf.mesh_faces_timeseries_output(
    mesh_name="FlowArea",
    required=[TimeSeriesOutputVar.FACE_VELOCITY, TimeSeriesOutputVar.FACE_WATER_SURFACE],
)