Build your own Metadata extractor

ECM is capable to automatically extract metadata from files at upload time. We have extractors for most common office documents (.doc, .pdf, .ppt, .xls, ...).

But this mechanism is extensible through plugins for your own metadata management. This tutorial will show you how to extract the metadata from a simple properties file.

Warning: For this tutorial, you will need java 5, maven 2 and eclipse

Create and deploy my extractor

Create a class called org.exoplatform.tutorial MyMetadataExtractor which has to extend org.exoplatform.services.document.impl.BaseDocumentReader. As follow:

import org.exoplatform.services.document.impl.BaseDocumentReader;

public class MyMetadataExtractor extends BaseDocumentReader {

Implement the getContentAsText methods which are dedicated to the full content extraction. As follow:

/**
 * Text extraction for the full text indexing
 */
public String getContentAsText(InputStream input) throws Exception {
	// Create a Properties Object
	Properties properties = new Properties();
	// Load the properties from the input stream
	properties.load(input);
	// Create a StringBuilder Object to append the full content of my file
	StringBuilder content = new StringBuilder();
	for (Enumeration keys = properties.keys(); keys.hasMoreElements(); ) {
		String key = (String) keys.nextElement();
		// Get the value from the key
		String value = properties.getProperty(key);
		// Append the value to the content
		content.append(value).append(' ');
	}
	return content.toString();
}

/**
 * Text extraction for the full text indexing with a specific character encoding
 */
public String getContentAsText(InputStream input, String encoding) throws Exception {
	return getContentAsText(input);
}

Note: This content will be indexed by the eXo Platform in full text. That will make possible to easily retrieve the file from its content through the Search Component.

Add the configuration file to declare the component to the eXo Platform, for this create a configuration.xml file into the conf/portal directory (into the source folder) and add the following content:

<?xml version="1.0" encoding="ISO-8859-1"?>
<configuration>
  <!-- Define my Metadata Extractor -->
  <external-component-plugins>
    <target-component>org.exoplatform.services.document.DocumentReaderService</target-component>
    <component-plugin>
      <name>my.document.reader</name>
      <set-method>addDocumentReader</set-method>
      <type>org.exoplatform.tutorial.MyMetadataExtractor</type>
      <description>to read my specific stream</description>
    </component-plugin>
  </external-component-plugins>
</configuration>

Build the resulting jar, to do this you have to use maven(maven 2 must be correctly installed in your system), you can create a pom.xml in the root folder of the project with the following content:

<project>
	<modelVersion>4.0.0</modelVersion>
	<groupId>org.exoplatform.tutorial</groupId>
	<artifactId>exo.tutorial.metadata-extraction</artifactId>
	<packaging>jar</packaging>
	<version>trunk</version>
	<description>Tutorial metadata extraction</description>
	<dependencies>
		<dependency>
			<groupId>org.exoplatform.core</groupId>
			<artifactId>exo.core.component.document</artifactId>
			<version>trunk</version>
			<scope>compile</scope>
		</dependency>
		<dependency>
			<groupId>org.exoplatform.kernel</groupId>
			<artifactId>exo.kernel.commons</artifactId>
			<version>trunk</version>
			<scope>compile</scope>
		</dependency>
		<dependency>
			<groupId>org.exoplatform.kernel</groupId>
			<artifactId>exo.kernel.container</artifactId>
			<version>trunk</version>
			<scope>compile</scope>
		</dependency>
	</dependencies>
	<build>
		<resources>
		  <resource>
			<directory>src/main/java</directory>
			<includes>
			  <include>**/*.xml</include>
			</includes>
		  </resource>		
		</resources>
	</build>
</project>

When you have done this, you can launch "mvn install" command from the root directory of the project. This will create a jar file called exo.tutorial.metadata-extraction-trunk.jar into the target folder. Note: Ensure that the maven settings file use http://maven2.exoplatform.org/rest/maven2 as maven repository.

Copy the jar into the ${TOMCAT_HOME}/lib directory
Start tomcat and check for the following message "ExoContainer - org.exoplatform.tutorial.MyMetadataExtractor added to portal"

Test my extractor

To test the extractor we must:

Create a properties file called myInputFile.properties with the following content:

Author=my Author
Subject=my Subject
Description=my Description

Authenticate to the portal as root
Go to Content Management -> File Explorer
Select a drive to upload the file like Collaboration Center
Click on the Upload icon
Choose the properties file that we created below then click on the Upload icon, the file will be uploaded to the server
Click on the Save button, you will see the following window:

Click on the Edit button to see the extracted metadata, you will see the following window:

To retrieve easily your file you can select the tab Search then type "my Author" in the seach text box, you will see the following result:

Note: The source files can be found in the zip file attached to this page, see the end of this page

last modified by Patrice Lamarque on 21/08/2008 at 23:58

Metadata Extraction

Top Menu

Build your own Metadata extractor

Create and deploy my extractor

Test my extractor

Search

Navigation