DSXchange

:P Thanks DSguru2B and sb_akarmarkar. In another way, we can use aggregator and count row by Key colum in a HashFile. Then, in a transformer, split the file with count(key column) = 1 => OK count(key column) > 1 => doublon File But I have another questions : Aggregator : Hash File, high performance ...

Hi, Could you help me to find simplest solutions to my problems : :cry: 1. I've to to save doublon in a file (if I use HashFile, I 'll lost the doublons) ? For example : Input : Key1, Data 11, Data 12 Key2, Data 21, Data 22 Key1, Data 31, Data 32 Output1 : Key1, Data 11, Data 12 Key2, Data 21, Data ...