Carrot2 Web Search Clustering Engine
------------------------------------

This archive contains a web-based application for clustering search results
based on the clustering algorithms available in the Carrot2 Framework
(http://www.carrot2.org).


Requirements
------------

The application requires:

    * Java Runtime Environment (JRE) version 1.4.2 or newer:
      http://java.sun.com/javase/downloads/index.jsp

    * Tomcat servlet container:
      http://tomcat.apache.org/download-55.cgi


Installation and running
------------------------

read Installation nodes first.

1. Copy the carrot2-demo-webapp.war file to Tomcat's webapps/ directory.

2. Start Tomcat.

3. Access: http://localhost:8080/carrot2-demo-webapp/


Installation notes
------------------

1. Be aware of query encoding issues. See the FAQ at:

http://project.carrot2.org/faq.html#corrupted-characters

2. For Google API input component, you need to define 'googleapi.keypool'
system property pointing at a folder where GoogleAPI keys can be found. Each
key file in that folder should have a ".key" extension and consist of a single
GoogleAPI key (one per file).

You can pass a JVM system property to Tomcat by defining CATALINA_OPTS:

CATALINA_OPTS=-Dgoogleapi.keypool=[absolute-path]


3. Carrot2 uses Log4j logging extensively. Log4j is by default included in the
Web application (WEB-INF/lib folder) and defines a default logging configuration
that places logging statements in Catalina's /log folder. If you plan to deploy
more than one application that uses Log4j, you may encounter class loading problems.
This is an advanced topic -- search mailing list archives for Log4j, Apache Commons Logging
and Tomcat issues.