Zilverline is what you could call a 'Reverse Search Engine'.

It indexes documents from your local disks (and UNC path style network disks), and allows you to search through them locally or if you're away from your machine, through a webserver on your machine.

Zilverline supports collections. A collection is a set of files and directories in a directory. PDF, Word, txt, java, CHM and HTML is supported, as well as zip, rar and other archive files. A collection can be indexed, and searched. The results of the search can be retrieved from local disk or remotely, if you run a webserver on your machine. Files inside zip, rar and chm files are extracted, indexed and can be cached. The cache can be mapped to sit behind your webserver as well.

You can store indexes and caches wherever you like, you could for instance store them on a DVD, as long as zilverline (and your webserver) can access them.

If you supply a url with your collection, the search results will map to the url instead of the source. This allows you to retrieve your result hits from your webserver, instead of disk.

Zilverline is internationalized (English and Dutch), and has support for skins (now three).

Zilverline is built in Java on top of Lucene and Spring. You need a Servlet Engine, such as Tomcat to run it. I'm using version 4.1.30.

Tested with Tomcat 4.1.x, Resin 3.0.7 and BEA Weblogic Server 8.1 using JDK 1.4 on Windows XP and Linux. On Linux you can get chm support from here