The data is saved as a WARC file, which contains the entire HTTP request and response (compressed, of course). So it's much bigger than just a short -> long URL mapping.

did they follow the redirect and archive the page content? but why?