Tika for metadata extraction

First pass of converting a few extractors to use Tika rather than 3rd party libraries directly, or use the new style tika structure


git-svn-id: https://svn.alfresco.com/repos/alfresco-enterprise/alfresco/HEAD/root@20640 c4b6b30b-aa2e-2d43-bbcb-ca4b014f7261
This commit is contained in:
Nick Burch
2010-06-14 19:02:37 +00:00
parent 7aeff72605
commit 63b2f5983a
9 changed files with 329 additions and 226 deletions

View File

@@ -21,6 +21,7 @@ package org.alfresco.repo.content.metadata;
import java.io.IOException;
import java.io.InputStream;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashSet;
import java.util.Map;
@@ -45,7 +46,7 @@ import org.apache.poi.hsmf.MAPIMessage;
* @since 2.1
* @author Kevin Roast
*/
public class MailMetadataExtracter extends AbstractMappingMetadataExtracter
public class MailMetadataExtracter extends TikaPoweredMetadataExtracter
{
private static final String KEY_SENT_DATE = "sentDate";
private static final String KEY_ORIGINATOR = "originator";
@@ -53,11 +54,14 @@ public class MailMetadataExtracter extends AbstractMappingMetadataExtracter
private static final String KEY_ADDRESSEES = "addressees";
private static final String KEY_SUBJECT = "subjectLine";
public static String[] SUPPORTED_MIMETYPES = new String[] {MimetypeMap.MIMETYPE_OUTLOOK_MSG};
public static ArrayList<String> SUPPORTED_MIMETYPES = buildSupportedMimetypes(
new String[] {MimetypeMap.MIMETYPE_OUTLOOK_MSG},
null
);
public MailMetadataExtracter()
{
super(new HashSet<String>(Arrays.asList(SUPPORTED_MIMETYPES)));
super(SUPPORTED_MIMETYPES);
}
@Override