Can a flat-file pipeline output back to the database so that a second xml pipeline can sort and debatch the data or can I make a single pipeline do this?