Oh well awesome site anyway :)
Posted under General
Since it doesn't look like one exists (although you might want to Dmail albert to see for sure), your best bet is probably just to use regular expressions to scrape the HTML of the aliases page.
Delete all newlines from the HTML (or set your regex parser to allow . to match newlines), then run something like this (very untested):
<tr.*<td><a.*>(.*)</a>.*<td><a.*>(.*)</a>.*</tr>
Each regex match will have the alias source and destination in capturing groups 1 and 2, so if you're doing this in a decent text editor, you can do this as a search and replace with your replace text as "\1,\2" to generate something approximating a CSV.
http://danbooru.donmai.us/tag_alias/index.xml works. You can paginate through it using page=2, page=3, etc. Look up the tag ids through tag/index.xml.