What is the best way to get the current text of an article from the Toolserver these days - the API, WikiProxy, or something else?  I need to examine thousands of pages on a daily basis and want to streamline the process as much as possible. Thanks, Jason