Does the input also understand tab separated files? If so, that gets around the EU character issue in seeing the right number of fields. But that would still leave the question of if the importer would recognize the EU embedded characters in the number or time fields correctly. The issue may span a couple of different stages of this process.