There's this example schema that illustrates DFDL being used to transform a pair of lists (which are hidden in a hidden group) into a list of pairs (which is the visible infoset).

I'm curious if this illustrates the nature of the transformations you describe that zTPF users are doing or if there is more to it. 

The output is not the same "shape" as the parsed input data, but it is linearly related in size. 

-mike