Thanks...I took a look at mechanize and recording before posting but hoped for something a little more direct. Automating the browser is an interesting idea but crawling is recursive while I would think automation assumes a determinate number of end states?