For Acrhive.org don't screen scape to get files and meta xml

I'm proposing here that we use a simple direct URL to the _files.xml and _meta.xml rather than the current scraping of the screen looking for the "All Files" HTML link. Although without an API neither is perfect I feel like this way is simpler and perhaps "slightly" less likely to change.

If you have a video (or audio for that matter) with video code (title) of "myvideo", then your normal download URL for files is

http://www.archive.org/download/myvideo/somefile

I also noticed that the myvideo_files.xml and myvideo_meta.xml can be found at the following URLS

http://www.archive.org/download/myvideo/myvideo_files.xml
http://www.archive.org/download/myvideo/myvideo_meta.xml

So as long as you have the video code it is a trivial matter to construct the above URLs. That is what the attached patch does.

Now I will mention that when you go to the above links archive.org does a redirect. However, the drupal_http_request (called from emfield_request_xml is designed to handle simple redirects). In my (admittedly brief) testing it appears to work fine.