Using Beagle with web sites?

So here’s a thought about a different approach to searching a content-thick web site, or relating a variety of content types to one another, based on my messing around more with MacFUSE:

What about using something like Beagle, FUSE and BeagleFS to create a search engine of static content and related data? This might be a little heavy because you’d basically have to create a new filesystem for each different query, but it would only get that hit the first time it’s created. If that’s too much, though, maybe a PHP/Python/Ruby library to act as a query client against a Beagle database would be adequate.

At CERIAS we have a ton of static content (lots of PDFs, a huge FTP site full of source code and message archives and OS mirrors, etc.). Something like this could really help get a handle on this, both in relating disparate types of data to one another, and in finding the right content in this mass of files.

Does this make sense?

[tags]beagle, beaglefs, FUSE, metadata, searching, search engine, webdev[/tags]