I have over 10,000 pages that are using the Thunderstone. The pages have been poorly written, meaning so have meta descript. tags others don't. Some have <H1> and some don't..... I need a script that will scan each page and pull out a description for each - that corresoponds with the search being made.
