mime type question

[Erik Hatcher <er...@ehatchersolutions.com>]

I'm writing a parse plugin for RDF data, such as http:// 
www.rossettiarchive.org/docs/1-1847.s244.raw.rdf

The web server is currently serving it as text/plain (I've asked our  
sysadmin to adjust it to the proper application/rdf+xml type though).

How can I configure my plugin (and still have parse-text in the mix)  
to get to it?   The order of searching for the designated parse  
plugin in ParseFactory first looks for a contentType match, then a  
pathSuffix match, and then the first one with a blank pathSuffix.

I'm curious what folks recommend to do in a situation like this.   
There is the mime-types.xml that seems to only be used if the content  
is served without a content-type at all.  Should there be (or is  
there already?) some facility to override the content-type for cases  
where the web server may just be brain-dead and serve text/plain for  
non-HTML, but text, files?

Thanks,
     Erik