The page-scraping is something that Perl excels at. That part, you'll probably code from scratch, as it depends on the structure of the page you're scraping. Start out by dumping your results into plain text, or a flatfile database, until you're getting your data reliably; you'll save time by not tackling two major tasks at once. (I've not checked recently, but there might be CPAN modules that can help you with the page-scraping.)
When you start, save a copy of the target page's source to disk and write code to extract the data from it. Once that's working, you can use wget (found on most Linux distributions; for a Windows port, look around at http://www.gnu.org ) to fetch live pages to be scraped.
For the database side, there are CPAN modules you can and should use. Go to http://www.cpan.org and put "SQL", and then "DBI", in the search field -- and read the O'Reilly book "Programming the Perl DBI". That problem's pretty much solved; it's a matter of hooking in the existing working library code and setting it to doing what you want.
You can write up something to import your flatfile database into SQL; then, when that's working, clump the two programs into one to complete the automation of your data collection. Breaking the task down into major components like this will save you a lot of confusion and frustration; I speak from experience.
