If your goal is to remove duplicates from a table in a RDBMS, I think you don't need Perl. Look at SELECT syntax (from MySQL documentation) and you'll find the useful DISTINCT option.

So you can populate a new table selecting distinct rows from an older one.

You can find an example of this tecnique in the documentation of DBSchema::Normalizer, from our gmax (I suggest visiting his homenode for further interesting DB-related articles).