Well, just some pointers because I'm not sure I completely understand the problem. You a user to be given two downloads at the same time and then run a diff command against those two files? On the server or the client i'm not sure? You'll have alot of problems automating the client side version because forcing people to run client side code is usually frowned upon by virus protection software.
The server side diff sounds exactly like a CGI moment to me: http://www.cs.tut.fi/~jkorpela/perl/cgi.html. That will allow you to generate a web-page that shows the diff between the two. CGI allows you to run programs on your server and display their output in a webpage; that's the simple explanation.
If that was not quite what you wanted then feel free to give me a comment and i'll try and edit to answer correctly.