Building on the idea of flat files, one might be inclined to use DBD::CSV where a flat file gets the DBI treatment. I use this on hosts where I can't install a full database, but I need simple database access lest I be forced to code it all myself.

Of course, moving from an HTML file to flat files isn't trivial if you're knee deep in data already, but the transition would easily pay for itself when the entries gets into the 1000s and you need to go to a full database solution because (suprise!) the client side is already done! Just change the dbd connect statement at the top to point to oracle, postgres, mysql, what-have-you, let out an evil cackle or two, then go to your next project =)

-Ducky