I agree that a database could be a good way to store this data, especially if it's going to be mined in various ways. However, if I'm given this file as-is and told to produce the reports specified by the OP, I'm not going to go to the trouble to put it in a database so that I can run SQL statements on it. I can write a script to get it directly in far less time, so a database will only become an advantage if I need to pull reports from it in several different ways. (And I already know SQL; the newbie would have to learn a second language first.)
Available for small or large Perl jobs and *nix system administration; see my home node.