You didn't say whether you slurp the data file into an array, or read it one line at a time. The latter will be faster for large files.

How do you parse the fixed-width records? The unpack function is recommended for speed.

Not a speed issue, but Text::CSV will give accurate CSV records from a list.

After Compline,
Zaxo