I think a combination of omitting the first row and doing a conditional split would do just fine. This the method I use since I receive a number of flat files from legacy systems having header and trailer rows. The conditional split needs to evaluate only one column to see if it's a valid row. Maybe I'm missing something, but I don't understand the reference to manually defining the column names and use of substrings. You don't have to rename the columns and using a substring seems a heck of a lot easier than all of the manipulations as described.