The Stack transform does the job well when the names of all the data sources are known. eg:
test1.csv
test2.csv
In situations where the number of input files is unknown, or changes over time, it would be handy if wild-cards/regular expressions were permitted in the input file names. eg:
test*.csv