I'm currently working on doing this right now. My only quirk is figuring out the best fields types to use for different columns.

Basically you just select the data you want from the database, decide which column goes in which lucene field, and you should be off and running.

I was able to modify the file system indexer to do a rough hack to index a database. It works fairly well and, although I have not yet benchmarked it, is very fast.

In order to accomplish this you will have to write some custom code to first retrieve the data that you want to index from the database and then feed it to Lucene for indexing.


That will allow you to create a full-text index with the data from the database.


Once you have that piece done you have to write some Java code using Lucene's API to search the created index and retrieve results from there.