Have you considered downloading a complete dump of the
will let you download 'cur'
(the current data of all Wikipedia articles), and it's only 301
megabytes. You can then run queries against this database at your own
Erm. I misread. I thought that you wanted all content from Wikipedia. =/