in reply to multiple file reading

Try doing the following:
Read each table into it's own hash, and at the same time, put an entry in the 'global' hash that contains all unique keys
Cycle the keys for the 'global' hash, and use the 'exists' function to check for existence in hash 'a' or hash 'b'.
Output as required, and all done.
This does look sneakily like a homework question, so I'm not about to post any code, but hopefully this pointer helps.

Cheers,

Malk.