org.hibernate.search.backend.impl.batch
Class DefaultBatchBackend

java.lang.Object
  org.hibernate.search.backend.impl.batch.DefaultBatchBackend

All Implemented Interfaces:

BatchBackend

public class DefaultBatchBackend
extends Object
implements BatchBackend

This is not meant to be used as a regular backend, only to apply batch changes to the index. Several threads are used to make changes to each index, so order of Work processing is not guaranteed.

Author:

Sanne Grinovero

Experimental

First BatchBackend

Constructor Summary

DefaultBatchBackend(SearchFactoryIntegrator searchFactoryImplementor, MassIndexerProgressMonitor progressMonitor)

Method Summary

void	doWorkInSync(LuceneWork work) Does one work in sync
void	enqueueAsyncWork(LuceneWork work) Enqueues one work to be processed asynchronously
void	flush(Set<Class<?>> entityTypes) Since most work is done async in the backend, we need to flush at the end to make sure we don't return control before all work was processed, and that IndexWriters are committed or closed.
void	optimize(Set<Class<?>> entityTypes) Triggers optimization of all indexes containing at least one instance of the listed targetedClasses.

Methods inherited from class java.lang.Object

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

Constructor Detail

DefaultBatchBackend

public DefaultBatchBackend(SearchFactoryIntegrator searchFactoryImplementor,
                           MassIndexerProgressMonitor progressMonitor)

Method Detail

enqueueAsyncWork

public void enqueueAsyncWork(LuceneWork work)
                      throws InterruptedException

Description copied from interface: BatchBackend

Enqueues one work to be processed asynchronously

Specified by:

enqueueAsyncWork in interface BatchBackend

Throws:

InterruptedException - if the current thread is interrupted while waiting for the work queue to have enough space.

doWorkInSync

public void doWorkInSync(LuceneWork work)

Description copied from interface: BatchBackend

Does one work in sync

Specified by:

doWorkInSync in interface BatchBackend

flush

public void flush(Set<Class<?>> entityTypes)

Description copied from interface: BatchBackend

Since most work is done async in the backend, we need to flush at the end to make sure we don't return control before all work was processed, and that IndexWriters are committed or closed.

Specified by:

flush in interface BatchBackend

optimize

public void optimize(Set<Class<?>> entityTypes)

Description copied from interface: BatchBackend

Triggers optimization of all indexes containing at least one instance of the listed targetedClasses.

Specified by:

optimize in interface BatchBackend

Parameters:

entityTypes - Used to specify which indexes need optimization.