How are the .dat files send through Data Cubes, related to each other for each quote?

When we are sending the .dat file through Datacube, we receive multiple .dat and .ctl files, based on the transaction and transaction lines and the arraysets. Now we want to convert this files into .csv files. The problem is for all the files, there is no common link between them in the .csv file. For example, if there is an arrayset, how do we know for which transaction are these arraysets for, since we do not have any any transaction linking attribute in that arrayset file.