Preserve NoneType types in UUID fields

[joshmatthews-ch]

I am using this library to read data from Postgres into a Pandas dataframe and have noticed that when the datatype in the source table is uuid, null values in those columns appear in the resulting dataframe as the "None".

To me this seems unexpected behaviour as it is caused by the Python decision to have str(None) == "None".

Should this be changed so that NoneType types are not cast to strings?