com.atlassian.bonnie.index
Class OnlineMultiThreadedIndexer

java.lang.Object
  com.atlassian.bonnie.index.BaseMultiThreadedIndexer
      com.atlassian.bonnie.index.OnlineMultiThreadedIndexer

All Implemented Interfaces:

BatchOpIndexer, SingleObjectIndexer

public class OnlineMultiThreadedIndexer
extends BaseMultiThreadedIndexer
implements SingleObjectIndexer

Performs batch indexing in multiple threads.

This class is designed for:

• an online environment where changes are being made to the index as batch reindexing occurs
• indexing large files

Each thread writes to its own temporary index and when all threads finish, these are merged into the final index. This has the desirable effect where a few large files do not bottleneck the entire operation.

Nested Class Summary

Nested classes/interfaces inherited from interface com.atlassian.bonnie.index.BatchOpIndexer

BatchOpIndexer.DocumentWritingScheme

Field Summary

protected edu.emory.mathcs.backport.java.util.concurrent.BlockingQueue	reindexAddedQueue
protected edu.emory.mathcs.backport.java.util.concurrent.BlockingQueue	reindexDeletedQueue

Fields inherited from class com.atlassian.bonnie.index.BaseMultiThreadedIndexer

log, luceneConnection

Constructor Summary

OnlineMultiThreadedIndexer()

Method Summary

protected void	allThreadsComplete(BatchOpIndexer.DocumentWritingScheme scheme, boolean truncate, com.atlassian.core.util.ProgressWrapper progress) Perform clean-up operations such as closing writers, optimizing, merging with original index, etc.
protected void	doAdd(java.lang.Object o, org.apache.lucene.index.IndexWriter writer) Performs the actual adding of the object to the index.
protected void	doDelete(java.lang.Object o, org.apache.lucene.index.IndexReader reader) Perform the actual deleting.
org.apache.lucene.analysis.Analyzer	getAnalyzer() Get analyzer used for indexing.
protected BatchOpIndexer.DocumentWritingScheme	getDocumentWritingScheme(boolean reindex) Creates a new instance of a DocumentWritingScheme.
protected ObjectToDocumentConverter	getObjectToDocumentConverter()
void	index(java.lang.Object o) Add an object to the index.
protected void	optimize(LuceneConnection conn) Optimize the index.
void	setLuceneConnection(LuceneConnection luceneConnection)
void	setObjectToDocumentConverter(ObjectToDocumentConverter objectToDocumentConverter)
protected void	setReindexingStatus(boolean status)
void	unindex(java.lang.Object o) Remove an object from the index.

Methods inherited from class com.atlassian.bonnie.index.BaseMultiThreadedIndexer

calculateNumberOfThreads, getQueueProcessingRunnable, isReindexing, reindex, setReindexing, truncateIndex

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

reindexAddedQueue

protected edu.emory.mathcs.backport.java.util.concurrent.BlockingQueue reindexAddedQueue

reindexDeletedQueue

protected edu.emory.mathcs.backport.java.util.concurrent.BlockingQueue reindexDeletedQueue

Constructor Detail

OnlineMultiThreadedIndexer

public OnlineMultiThreadedIndexer()

Method Detail

allThreadsComplete

protected void allThreadsComplete(BatchOpIndexer.DocumentWritingScheme scheme,
                                  boolean truncate,
                                  com.atlassian.core.util.ProgressWrapper progress)

Description copied from class: BaseMultiThreadedIndexer

Perform clean-up operations such as closing writers, optimizing, merging with original index, etc.

Specified by:

allThreadsComplete in class BaseMultiThreadedIndexer

Parameters:

scheme - scheme

truncate - should the original index be truncated

index

public void index(java.lang.Object o)

Description copied from interface: SingleObjectIndexer

Add an object to the index. Calls SingleObjectIndexer.unindex(java.lang.Object) before the object is added.

Specified by:

index in interface SingleObjectIndexer

doAdd

protected void doAdd(java.lang.Object o,
                     org.apache.lucene.index.IndexWriter writer)
              throws java.io.IOException

Performs the actual adding of the object to the index.

Parameters:

o -

writer -

Throws:

java.io.IOException

unindex

public void unindex(java.lang.Object o)

Description copied from interface: SingleObjectIndexer

Remove an object from the index.

Specified by:

unindex in interface SingleObjectIndexer

doDelete

protected void doDelete(java.lang.Object o,
                        org.apache.lucene.index.IndexReader reader)
                 throws java.io.IOException

Perform the actual deleting.

Parameters:

o - object to delete

reader - reader

Throws:

java.io.IOException

optimize

protected final void optimize(LuceneConnection conn)

Optimize the index.

Parameters:

conn -

getDocumentWritingScheme

protected BatchOpIndexer.DocumentWritingScheme getDocumentWritingScheme(boolean reindex)

Creates a new instance of a DocumentWritingScheme.

Parameters:

reindex - is this a reindex

setLuceneConnection

public void setLuceneConnection(LuceneConnection luceneConnection)

Specified by:

setLuceneConnection in interface SingleObjectIndexer

getObjectToDocumentConverter

protected ObjectToDocumentConverter getObjectToDocumentConverter()

setObjectToDocumentConverter

public void setObjectToDocumentConverter(ObjectToDocumentConverter objectToDocumentConverter)

Specified by:

setObjectToDocumentConverter in interface SingleObjectIndexer

getAnalyzer

public org.apache.lucene.analysis.Analyzer getAnalyzer()

Description copied from interface: BatchOpIndexer

Get analyzer used for indexing.

Specified by:

getAnalyzer in interface BatchOpIndexer

setReindexingStatus

protected void setReindexingStatus(boolean status)
                            throws java.io.IOException

Throws:

java.io.IOException