I found that some of the Incremental loads have Map operation tranforms before and after the Table comparison. Did anyone know the reason why some DI developers use Map Operation transform before table comparision?
To control what is pushed down into the source database, e.g. my Query_2 contains a substr() function but I want to execute it in the ETL engine instead of pushing it into the reading SQL. If the Map is directly in front of the TC it does not make sense.
To change the opcode of normal rows to insert/update/delete and hence all input rows are read but instead of the loader issuing an insert statement, you want to e.g. delete all these rows. Again, if a Table Comparison is downstream it does not make sense.
This makes sense, but in the mappings I saw map operation is directly infront of Table comparision converting every opcode to normal and then I donβt see any transformation before that manipulating opcodes.