Well, that's the problem right there. There have been numerous comments about "Matt's Scripts" here at the monastery, including statements like:

• When I think of visible Perl code with poor security, I think of the original Matt Wright script archive code, before the NMS project. (from Re^4: A Vision for Easy Web Application Deployment for Perl)
• Take (Matt Wright's book) on your list and be very careful NOT to read it. (from Re: Perl CGI and books)
• I too made the mistake of buying Matt Wright's book when I was first dipping my toes into Perl. (from Re: Re: Perl CGI and books)
• I just have reiterate what other people have said about Matt Wright's book. Don't even open it until you've read all of the other ones. Only then will you be in a position to see just how bad it is. (from Re: Perl CGI and books)

The last three quotes come from the same thread, which you might want to look at because it also provides a number of very useful resources (although it's now 6 years old).

Note that the NMS project mentioned above can be found here: http://nms-cgi.sf.net/ (UPDATE: actually, a better link would be: http://sourceforge.net/projects/nms-cgi/, which shows clear evidence of fairly recent maintenance by our esteemed davorg, among others.)

Bottom line: with regard to the code in the OP, the best course of action will be to ditch it and start over, using some better prototype as a starting point.

(UPDATE: I think it's worth pointing out that the name "NMS" derives from "Not Matt's Scripts".)