Create cron-job web scraper to filter 1 website and place into searchable database.

I need data scraped from Terapeak.com. I have an account and login details with Terapeak. Terapeak monitors average sale prices on Ebay. I need a database where I select a SKU from a drop-down list and it gives me the scraped terapeak data. i will need to maintain a separate database to tie a SKU to a specific keyword search to make sure the correct data is pulled from terapeak. You need to make a script that looks at my database, searches terapeak for each SKU using the keyword-search I have created, and places that data in a spreadsheet. Every SKU needs to be updated twice a week - in other words the database needs to be synced with Terapeak twice a week. There are currently about 300 SKUs. I will be modying/adding/deleting SKUs and their keyword searches from my database, so your script needs to be able to adjust itself with my database changes. I have a linux-based web server that I can give you access to for placing the script/files/etc onto. Please see the attached screenshots: "Screen+Shot+2012-08-05+at+9.02.08+[login to view URL]" - This image simply shows me, manually searching a keyword search in Terapeak and Terapeak gathering the search results. "Screen+Shot+2012-08-05+at+9.01.43+[login to view URL]" - This show the keyword search results. The number of results can vary, but I only need the top 10 for each keyword search. (some searches may even have less than 10 results) I need all 9 columns scraped for each keyword-search.