You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@manifoldcf.apache.org by kw...@apache.org on 2014/10/08 13:27:36 UTC
svn commit: r1630073 - in
/manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler:
interfaces/IFingerprintActivity.java system/WorkerThread.java
Author: kwright
Date: Wed Oct 8 11:27:35 2014
New Revision: 1630073
URL: http://svn.apache.org/r1630073
Log:
Hook up checkDate() for processing
Modified:
manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/interfaces/IFingerprintActivity.java
manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/system/WorkerThread.java
Modified: manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/interfaces/IFingerprintActivity.java
URL: http://svn.apache.org/viewvc/manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/interfaces/IFingerprintActivity.java?rev=1630073&r1=1630072&r2=1630073&view=diff
==============================================================================
--- manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/interfaces/IFingerprintActivity.java (original)
+++ manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/interfaces/IFingerprintActivity.java Wed Oct 8 11:27:35 2014
@@ -21,6 +21,7 @@ package org.apache.manifoldcf.crawler.in
import org.apache.manifoldcf.core.interfaces.*;
import org.apache.manifoldcf.agents.interfaces.*;
import java.io.*;
+import java.util.*;
/** This interface abstracts from the activities that handle document fingerprinting and mime type acceptance.
*/
@@ -28,6 +29,14 @@ public interface IFingerprintActivity
{
public static final String _rcsid = "@(#)$Id: IFingerprintActivity.java 988245 2010-08-23 18:39:35Z kwright $";
+ /** Detect if a date is indexable or not. This method is used by participating repository connectors to pre-filter the number of
+ * unusable documents that will be passed to this output connector.
+ *@param date is the date of the document; may be null
+ *@return true if a document with that date is indexable by this connector.
+ */
+ public boolean checkDateIndexable(Date date)
+ throws ManifoldCFException, ServiceInterruption;
+
/** Detect if a mime type is indexable or not. This method is used by participating repository connectors to pre-filter the number of
* unusable documents that will be passed to this output connector.
*@param mimeType is the mime type of the document.
Modified: manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/system/WorkerThread.java
URL: http://svn.apache.org/viewvc/manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/system/WorkerThread.java?rev=1630073&r1=1630072&r2=1630073&view=diff
==============================================================================
--- manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/system/WorkerThread.java (original)
+++ manifoldcf/branches/CONNECTORS-1067/framework/pull-agent/src/main/java/org/apache/manifoldcf/crawler/system/WorkerThread.java Wed Oct 8 11:27:35 2014
@@ -1907,6 +1907,19 @@ public class WorkerThread extends Thread
abortSet.add(localIdentifier);
}
+ /** Detect if a date is indexable or not. This method is used by participating repository connectors to pre-filter the number of
+ * unusable documents that will be passed to this output connector.
+ *@param date is the date of the document; may be null
+ *@return true if a document with that date is indexable by this connector.
+ */
+ public boolean checkDateIndexable(Date date)
+ throws ManifoldCFException, ServiceInterruption
+ {
+ return ingester.checkDateIndexable(
+ pipelineSpecification,date,
+ ingestLogger);
+ }
+
/** Check whether a mime type is indexable by the currently specified output connector.
*@param mimeType is the mime type to check, not including any character set specification.
*@return true if the mime type is indexable.