Scraping web-pages for to gather information from them involves several steps. Prolog may not seem to be the most obvious language for this job, but SWI-Prolog can cope quite well. Here are the steps:

1. Parse an HTML page
2. XPath.md
3. Spider.md