Well, tr will remove the " characters, like below...
tr is a simple minded critter, but it is faster than substitute.

If you created this with Text::CSV, then I would use that to parse it back in. Normally, I don't think that you have to specify the XS version, if its there, then it gets used - to the best of my knowledge.

I'm still on a marathon DB project, but my machine has to think for 4-5 hours about what I've done so far. But on this project, | is used as the CSV delimiter instead of "," and that often works out very well. Sometimes I also see || and that is ok too (in the split, you can specify /\|\|/ as the splitting regex. In the DB that I'm working with | is explicitly not allowed as a valid data field value and hence it can be used as a simple delimiter in the CSV format and the Text::CSV module is not needed. But mileage varies...

#!/usr/bin/perl -w
use strict;

while (<DATA>)
{
   tr/"//d;
   print;
}

# prints:
# some,always,quoted stuff
# more,stuff,like that

__DATA__
"some","always","quoted stuff"
"more","stuff","like that"
[download]