It is a single batch type program. Efficiency is important.
1) The program must read through all domains ending with ".co.za" .
2) For each main (index or landing page) it must search for a suppled keyword or phrase. It must only search the first page and when it finds a match it can stop searching
3) If found it must write the domain name to a SQL database.
One must have the ability to change the domain "end" and the ability to alter the keyword or phrase. Effectively I am trying to extract all used car dealers in a country.