org.hibernate.search.backend.impl.lucene

Class AbstractWorkspaceImpl

java.lang.Object

org.hibernate.search.backend.impl.lucene.AbstractWorkspaceImpl

All Implemented Interfaces:

Workspace

Direct Known Subclasses:

ExclusiveIndexWorkspaceImpl, NRTWorkspaceImpl, SharedIndexWorkspaceImpl

public abstract class AbstractWorkspaceImpl
extends Object
implements Workspace

Lucene workspace for an IndexManager

Author:

Emmanuel Bernard, Hardy Ferentschik, Sanne Grinovero

Field Summary

Fields

Modifier and Type	Field and Description
protected org.hibernate.search.backend.impl.lucene.IndexWriterHolder	writerHolder

Constructor Summary

Constructors

Constructor and Description
AbstractWorkspaceImpl(DirectoryBasedIndexManager indexManager, WorkerBuildContext context, Properties cfg)

Method Summary

Modifier and Type	Method and Description
void	afterTransactionApplied(boolean someFailureHappened, boolean streaming) Invoked after all changes of a transaction are applied.
boolean	areSingleTermDeletesSafe() Returns true if one and only one entity type is stored in the targeted index.
void	closeIndexWriter()
void	flush() Makes sure eventually pending changes are made visible to IndexReaders.
org.apache.lucene.analysis.Analyzer	getAnalyzer(String name)
DocumentBuilderIndexedEntity	getDocumentBuilder(Class<?> entity)
Set<Class<?>>	getEntitiesInIndexManager()
String	getIndexName() Returns the name of the index this workspace is being used for.
org.apache.lucene.index.IndexWriter	getIndexWriter() Gets the IndexWriter, opening one if needed.
org.apache.lucene.index.IndexWriter	getIndexWriter(ErrorContextBuilder errorContextBuilder)
IndexWriterDelegate	getIndexWriterDelegate()
IndexWriterDelegate	getIndexWriterDelegate(ErrorContextBuilder errorContextBuilder)
protected void	incrementModificationCounter()
boolean	isDeleteByTermEnforced() Returns true if either the configuration guarantees that one can use delete by term on all indexes or if we ensure that entity types stored in the index return positive to IdUniquenessResolver and that the document id and JPA id are the same property.
void	optimizerPhase() If optimization has not been forced give a chance to configured OptimizerStrategy to optimize the index.
void	performOptimization(org.apache.lucene.index.IndexWriter writer) Used by OptimizeLuceneWork to start an optimization process of the index.
void	shutDownNow()

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.hibernate.search.store.Workspace

getCommitPolicy, notifyWorkApplied

Field Detail

writerHolder

protected final org.hibernate.search.backend.impl.lucene.IndexWriterHolder writerHolder

Constructor Detail

AbstractWorkspaceImpl

public AbstractWorkspaceImpl(DirectoryBasedIndexManager indexManager,
                             WorkerBuildContext context,
                             Properties cfg)

Method Detail

getDocumentBuilder

public DocumentBuilderIndexedEntity getDocumentBuilder(Class<?> entity)

Specified by:

getDocumentBuilder in interface Workspace

getAnalyzer

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

Specified by:

getAnalyzer in interface Workspace

optimizerPhase

public void optimizerPhase()

Description copied from interface: Workspace

If optimization has not been forced give a chance to configured OptimizerStrategy to optimize the index.

Specified by:

optimizerPhase in interface Workspace

performOptimization

public void performOptimization(org.apache.lucene.index.IndexWriter writer)

Description copied from interface: Workspace

Used by OptimizeLuceneWork to start an optimization process of the index.

Specified by:

performOptimization in interface Workspace

Parameters:

writer - the IndexWriter to use for optimization

See Also:

OptimizeLuceneWork, SearchIntegrator.optimize(), SearchIntegrator.optimize(Class)

incrementModificationCounter

protected void incrementModificationCounter()

getEntitiesInIndexManager

public Set<Class<?>> getEntitiesInIndexManager()

Specified by:

getEntitiesInIndexManager in interface Workspace

Returns:

The set of entity types being indexed in the underlying IndexManager backing this Workspace.

afterTransactionApplied

public void afterTransactionApplied(boolean someFailureHappened,
                                    boolean streaming)

Description copied from interface: Workspace

Invoked after all changes of a transaction are applied. Must be invoked strictly once after every Workspace.getIndexWriter() in a finally block as implementations might rely on counters to release the IndexWriter.

Specified by:

afterTransactionApplied in interface Workspace

Parameters:

someFailureHappened - usually false, set to true if errors where caught while using the IndexWriter

streaming - if no immediate visibility of the change is required (hint for performance)

shutDownNow

public void shutDownNow()

closeIndexWriter

public void closeIndexWriter()

getIndexWriter

public org.apache.lucene.index.IndexWriter getIndexWriter()

Description copied from interface: Workspace

Gets the IndexWriter, opening one if needed.

Specified by:

getIndexWriter in interface Workspace

Returns:

a new IndexWriter or an already open one, or null if an error happened.

getIndexWriter

public org.apache.lucene.index.IndexWriter getIndexWriter(ErrorContextBuilder errorContextBuilder)

areSingleTermDeletesSafe

public boolean areSingleTermDeletesSafe()

Description copied from interface: Workspace

Returns true if one and only one entity type is stored in the targeted index. This allows to use delete by the identifier term which is much faster. This method should really be named isSingleClassInIndex but that's a public contract.

Specified by:

areSingleTermDeletesSafe in interface Workspace

Returns:

true if one and only one entity type is stored in the targeted index.

isDeleteByTermEnforced

public boolean isDeleteByTermEnforced()

Description copied from interface: Workspace

Returns true if either the configuration guarantees that one can use delete by term on all indexes or if we ensure that entity types stored in the index return positive to IdUniquenessResolver and that the document id and JPA id are the same property. This allows to use delete by identifier term in a safe way and is much more efficient. If unsure, we will delete taking the class field into account to avoid unwanted document deletions. The method should really be named isDeleteByTermEnforcedOrSafe but that's a public contract.

Specified by:

isDeleteByTermEnforced in interface Workspace

Returns:

true if either the configuration guarantees that one can use delete by term on all indexes

flush

public void flush()

Description copied from interface: Workspace

Makes sure eventually pending changes are made visible to IndexReaders.

Specified by:

flush in interface Workspace

getIndexName

public String getIndexName()

Description copied from interface: Workspace

Returns the name of the index this workspace is being used for.

Specified by:

getIndexName in interface Workspace

Returns:

the name of the index

getIndexWriterDelegate

public IndexWriterDelegate getIndexWriterDelegate(ErrorContextBuilder errorContextBuilder)

getIndexWriterDelegate

public IndexWriterDelegate getIndexWriterDelegate()