Rather than manually parsing and merging the files, I would suggest looking at DBD::CSV or, at a minimum, any of the CSV parsing modules on CPAN.

As an example, you could use Text::CSV to read file 1. For each line in file 1, extract the supervisor's ID, then use DBD::CSV and SQL to locate the supervisor's name in file 2. Write the output to a new file.

Hope this gets you started...