Site migration to Wordpress crawl and export site to xml

I'm doing a lot of site migrations all of which are going to Wordpress. For the most part, I can export page data and import it into Wordpress, but some for some sites, I'm either dealing with static html or missing backend login info.

Is there a tool that will crawl a site, extract the content, and export it to an XML file in the same format WordPress' import tool uses?