Add SODA integration

[skrawcz]

Is your feature request related to a problem? Please describe.
Enable checks with SODA - e.g. https://docs.soda.io/soda/connect-dask.html.

Describe the solution you'd like

I think a UX in Hamilton could look like:

@tag(scan_definition_name="test", 
 data_set_name="employee", 
 data_source_name="orders")
def some_dataset() -> pd.DataFrame:
 # pandas/spark/dask logic
 return df

# then when you execute you provide the soda adapter that will trigger after executing `some_dataset()` in memory.
soda_adapter = Soda(checks_yaml, ...)
dr = driver.Builder()...with_adapters(soda_adapter).build()
dr.execute(["some_dataset"], inputs=...)

Describe alternatives you've considered
Could go about extending @check_output or @check_output_custom like we've done with pandera.

Additional context
SODA is a contract library.