The Wumpus Information Retrieval System – How to use Wumpus?

Author: Stefan Buettcher (stefan@buettcher.org)
Last change: 2006-04-26

After you have downloaded the Wumpus software package and extracted all files from the archive ("tar xzf wumpus.tgz"), change to the wumpus/ directory and type in "make". This will build Wumpus and create the system executables in wumpus/bin/.

After the system has been built, you can start Wumpus by typing
bin/wumpus --config=wumpus.cfg
The additional "--config" parameter is necessary to tell wumpus what configuration file should be used ("wumpus.cfg"). Among other things, the configuration file defines the directory in which Wumpus will create its index structures. If you use the configuration file that came with Wumpus, you will find the index in wumpus/database/.

After you start the system, Wumpus will create a couple of files in the index directory and then tell you that it is ready to process your queries: 
@0-Index loaded.
You can then add files to the index and submit queries to the system. Every response to a query ends with a line starting with "@" that indicates whether the query was successful or not and also tells you how long it took Wumpus to process your query. 
Hi Wumpus, how are you doing?
@1-Syntax error. (0 ms)
When you are using Wumpus and are modifying the index (adding or removing files), please make sure to properly end your session by typing Ctrl+D (closing the input stream). If you terminate Wumpus by pressing Ctrl+C, then this will leave the index in an inconsistent state. When you start Wumpus the next time, it will "recover" from this inconsistency by deleting all information in the index. Therefore, it is important that you always use Ctrl+D to end a Wumpus session.

The next section, Index updates and simple queries, shows how you can add files to Wumpus' index and explains the basic query types.