[squid-users] Using squid to intercept/convert .doc to .pdf?

I'm trying to come up with a way to automatically convert MS Word documents
to PDF files for display in a web browser.

* Some users MUST access documents on external sites; these docs are in MS
Word format. :-p
* We have no control/influence over the external sites.

* User platform is a Linux based thin client running Firefox/Acrobat Reader.
* The thin clients cannot be modified.

* Squid 2.6-2 is running on a server in transparent proxy mode.
* Only the thin client users are routed thru this Squid server.

Question: Could I use Squid to detect a URL whose target is an MS Word
document (eg, filename.doc), download the document, hand it off to an
external script (which will convert the document to pdf) and then pass the
resulting stream back to Squid to pass on to the browser?

Thoughts/ideas/suggestions/alterntives appreciated!

Tom