It fallbacks to archives, yes. Basically I use different User-Agent headers, different Referer headers, it tries disabling javascript once the page has loaded and the fallback is to fetch from web archives (Wayback Machine, archive.is, Google cache).
Then the HTML is validated and parsed.