Can screen-scraper extract information from PDF files?

Sort of, yes. See this blog posting.