Oracle Universal Content Management Sample Content Service for WCI

Date: August, 2009

Introduction
Application Overview
Sample Files
Deployment
Configuration
Additional References

Introduction

Software Prerequisites
Knowledge Prerequisites

Software Prerequisites

This sample content service requires the following:

Oracle Universal Content Management 10gR3
Oracle WebCenter Interaction 10gR3

If you do not have the above software available, please visit the Oracle WebCenter Suite Home Page for WebCenter Interaction and the Oracle Universal Content Management Home Page for UCM.

Knowledge Prerequisites

This sample content service will be deployed and configured on a installation of UCM as it will need access to the APIs of that installation. It is assumed that the reader is familiar with the location of this installation.

Application Overview

WCI Content Services provide access to document based content from many different sources all through the UI of the Knowledge Directory. The content of the source being crawled will be linked to and then all full-text and metadata will be indexed within the Search Server of WCI. The content service can synchronize folder structures and security of the actual source being crawled.

This content service is designed to crawl and index content from a UCM repository. The content service can be configured to crawl and index the content along with any of its metadata. When the content is crawled, the security of the content can be mirrored at the same time as long as both WCI and UCM are using the same authentication source. Lastly, if the folder module has been installed as part of the UCM installation, then folder mirroring can be configured as well. Installation and Configuration instructions can be found below.

Sample Files

The sample content service is packaged as a Java Web Archive (WAR) file that contains all of the files necessary to deploy the application. The following list of files may be used as a reference when deploying, configuring, modifying, and otherwise using the sample.

Portlet Package Files

Directory	Description
/documentation	Installation & Configuration documentation for sample
/serverpackages	PTE file used for importing of WCI portal objects
/src	All Java classes used for content service functionality
/webapp	Web Application Archive needed for deployment of the content service

Deployment

Deployment of the sample content service will follow all of the standard procedures for WCI content service deployment. Please follow all of the steps below:

1. Deploy the ucmcrawler.war from webapp folder application to a valid Java application server
2. Using the Migration Wizard, import the UCM Crawler.pte serverpackage
3. Change the Remote Server object's URL in the UCM Crawler folder to point to the server and port of your application server
4. Create a Content Source object, using the UCM Crawler Web Service. Specify the UCM server, port, and name used to crawl
5. Create a Content Crawler object, using the Content Source object specified above. Specify the folder to crawl and the crawl depth
6. Create a Job object, and run the crawl

DEPLOYMENT NOTES:

The UCM server will not allow connections unless the server name or IP address is added to the SocketHostAddressSecurityFilter field in the config.cfg file, usually located in the /oracle/ucm/server directory. The port is usually 4444, and that is specified by the IntradocServerPort in the same file.
There is a known issue with the folder (tree) picker: a folder absolutely needs to be selected. If you get the error that a folder needs to be selected, click the parent folder, and the previously selected folder should now show as selected. This will occur when the crawler page is refreshed, or when the crawler is edited again.

IMPORTING SECURITY NOTES:

The Domain name of the UCM crawler is "UCM". In the Global ACL Sync Map utility, map the UCM Domain to your Authentication Source Prefix in the Prefix: Domain Map setting.
The UCM Crawler imports both Account Names, and roles for Security Groups. Map these to your existing groups using the Portal: External Group Map setting.
Note that while UCM accounts act as a filter before checking the roles in a security group, the portal will allow users to view items in the folder as long as they belong to any mapped group from either a role or an account.

UCM PROPERTY NOTES:

The fields returned for UCM are listed in ucmFieldNames.properties in WEB-INF/classes. Any fields listed will be added to the document metadata. The fields are separated into text fields and date fields, so we know which fields should be converted in the portal date format. Only fields that have been added to the Global document property map will be added. The properties UCM author and UCM File size have been added as examples.