what is the best approach where you read same tables from two different kind of databases, oracledb , hive etc. All the data prep steps are same except the sql that initially reads into dataframe. how to make this configurable with kedro?