Requirements
This assumes you have some knowledge of eclipse, and the extention point mechanism in eclipse. It also assumes you have eclipse and svn installed on your machine.
It is also assumed you have looked at the Tupelo documentation and have some basic understanding of this. Here are some starting pages to learn about Resources and Triples, Unifiers and Contexts.
Once the eclipse project is populated it might be worth it to look at edu.illinois.mmdb.extractor.* as examples on how to extend the extraction service.
Installation
Following are the instructions to extend the Extraction Service. This is written with eclipse Helios in mind as well an empty workspace.
Buckminster
First you should make sure buckminster is installed. Click on Help -> Install new software. Select "All Available Sites" and type "buck" in the search box. You will need to install Buckminster - Core, Buckminster - PDE support and Buckminster - Subversive support. Install and restart Eclipse.
Extraction Service
Click on File -> Open a component query and use https://opensource.ncsa.illinois.edu/svn/mmdb/trunk/scripts/buckminster/edu.illinois.ncsa.cyberintegrator.client.site.cquery
as input. Click on Resolve and Materialize and wait. This will get all the pieces for the Extraction Service and install them in your workspace.