Single source the datatype for measurement inputs/outputs

Tech Debt

In the examples, the measure function is decorated with the inputs and outputs for the measurement. One input to the decorator is the datatype of the inputs/outputs. However, the datatype is also specified as part of type hints for the measure function. So this data is duplicated and could easily get out of sync if a datatype changed. We should explore if there is a way for the decorator to infer the datatype from the type hints so that we can single source that information. We might even be able to avoid having to specify a default value by putting that in the function signature as well. Maybe we can get rid of the decorators altogether. :) Below is a simple before/after (note that syntax might not be exactly right):

Before:

@measurement_service.register_measurement
@measurement_service.configuration("string_input", nims.DataType.String, "default_value")
@measurement_service.output("string_output", nims.DataType.String)
def measure(string_input: str) -> Tuple:
    return ""

After:

@measurement_service.register_measurement
@measurement_service.configuration("string_input", "default_value")
@measurement_service.output("string_output")
def measure(string_input: str) -> Tuple[str, ]:
    return ""

AB#2267350

Start with the example measure function and its measurement_service decorators described in the issue, then inspect how type hints and decorator datatype/default arguments are currently interpreted. Done means establishing a single source for input and output datatypes and determining whether defaults or the decorators themselves can be removed.