View source

This page describes how to setup TeamWeaver Integrated search in various flavours.

== Server Backend ==
The TeamWeaverIS backend is the core component of Integrated Search. It maintains an index, which can be built by crawling various different [[Integrated Search/Supported data sources|supported data sources]]. The backend offers services, which can be used by one of the client frontends (JSF-Frontend, Eclipse-based frontend or [[Woogle|Wiki-based frontend]]) to query and display search results.

The following steps are required for installation:
* Prepare a Java Web Application Server (usually, the latest [http://tomcat.apache.org/ Apache Tomcat] will do; see [[Integrated Search/Compatibility|supported Web Application Servers]])
** Tomcat comes as a ZIP distribution and installation is typically not more than unzipping. In the following, we will assume you chose Tomcat as your Web Application Server - however everything works very similar with other systems such as [http://jetty.mortbay.com/jetty/ Jetty].
* [[Integrated Search/Download|Download]] the latest version of the backend
* Copy the file <code>teamweaverIS-backend.war</code> to the <code>webapps</code> folder of your Tomcat
* Start the Tomcat once. It will then "deploy" the WAR-archive and automatically create a separate folder <code>teamweaverIS-backend</code> in the <code>webapps</code> folder.
** Note: Tomcat can run as an application (using <code>webapps</code>startup.bat<code>webapps</code> resp. <code>webapps</code>startup.sh<code>webapps</code> scripts) or as a service
* Stop the Tomcat again. 
* You may now proceed with the [[Integrated Search/Configuration|configuration]].
* After configuration, you can start Tomcat again. After configuration and initial crawling, you can start Tomcat again. The TeamWeaverIS server backend is configured to provide indexing and search functionility as a web service. You can use any TeamWeaverIS frontend to connect to your service. When the Tomcat runs on your local machine, the service URL should be: '''<code>http://localhost:8080/teamweaverIS-backend/services/query</code>''', unless you configured Tomcat to run on another port.

== Frontends ==
Chose any (or multiple) of these frontends to access search results from your TeamWeaverIS backend.

=== JSF Frontend ===
The JSF-Frontend is the default frontend for the server backend. It is also deployed in a Java Web Application Server (usually, but not necessarily the same as the backend) and offers a "Google-style" browser-based search interface to access data from a backend.

The following steps are required for installation:
* Prepare a Java Web Application Server (see installation steps for the backend; we recommend to use the same server as for the backend)
* [[Integrated Search/Download|Download]] the latest version of the JSF-frontend
* Copy the file <code>teamweaverIS.war</code> to the <code>webapps</code> folder of your Tomcat
* Start the Tomcat once. It will then "deploy" the WAR-archive and automatically create a separate folder <code>teamweaverIS</code> in the <code>webapps</code> folder.
* Stop the Tomcat again.
* If you run the frontend in a different Tomcat than the backend, '''or''' if your Tomcat uses another port than the default 8080, you need to edit the server URLs in the configuration file <code>\webapps\teamweaverIS\WEB-INF\queryWSClient.xml</code> 
* You can then start the Tomcat again to use the JSF frontend. You can now open the JSF-frontend URL in your browser. If you run the Tomcat on your localhost, this will typically be http://localhost:8080/teamweaverIS/. The result should look like this site: http://octopus18.fzi.de:8080/wavesIS/. 

=== (Media) Wiki Frontend ===
You may search and display results from a TeamWeaverIS server backend within a MediaWiki instance. In this scenario, TeamWeaverIS completely replaces the built-in Wiki-search - i.e. changes in the Wiki directly reflect in the TeamWeaverIS backend index. 

An implementation for MediaWiki is availalbe and fairly stable, while an implementation for the Atlassian Confluence wiki engine is currently under development. See [[Woogle]] for more details.

=== Eclipse Plugin ===
TeamWeaverIS can also be integrated into the Eclipse IDE. Both functionalities - crawling and searching - are supported within Eclipse. We currently do not offer separate/modularized distributions. However, you can neglect crawling features in Eclipse when you just want to use it as a serch UI for a server backend.

The current (unstable) update site URL is http://www.teamweaver.org/downloads/eclipse-update-site/2008-10-27-twIS-098/. Note that the Eclipse plugin might not yet be ready for production use. The plugin is tested with Eclipse 3.4 (Ganymede) and 3.3 (Europa) and will most probably run on Eclipse 3.2 as well. Note that the update site size is around 30MB - installation might thus take some time.

We also have an experimental Eclipse RCP (desktop) application.