BTW, before someone mentions using Text::CSV_XS, I have a version that uses the module but it takes over 4 times longer to process. The real files I will be processing will be huge and I need all the speed I can squeeze out of this.

Oh really? I'd like to see that ;)