A cron job which harvests the HTML files occasionally would be the ideal solution. Therefore this would eliminate the need for a CMS and naturally, cut the costs overall as well as user intervention. Then simply have a 3NF database to return the relevant results based on the keywords specified by the end user.