org.apache.jackrabbit.core.query.lucene
Class SearchIndex

java.lang.Object
  org.apache.jackrabbit.core.query.AbstractQueryHandler
      org.apache.jackrabbit.core.query.lucene.SearchIndex

All Implemented Interfaces:

QueryHandler

public class SearchIndex
extends AbstractQueryHandler

Implements a QueryHandler using Lucene.

Nested Class Summary

protected static class

SearchIndex.CombinedIndexReader Combines multiple CachingMultiReader into a MultiReader with HierarchyResolver support.

Field Summary

static int	DEFAULT_EXTRACTOR_BACK_LOG The default value for property extractorBackLog.
static int	DEFAULT_EXTRACTOR_POOL_SIZE The default value for property extractorPoolSize.
static long	DEFAULT_EXTRACTOR_TIMEOUT The default timeout in milliseconds which is granted to the text extraction process until fulltext indexing is deferred to a background thread.
static int	DEFAULT_MAX_FIELD_LENGTH the default value for property maxFieldLength.
static int	DEFAULT_MAX_MERGE_DOCS The default value for property maxMergeDocs.
static int	DEFAULT_MERGE_FACTOR the default value for property mergeFactor.
static int	DEFAULT_MIN_MERGE_DOCS The default value for property minMergeDocs.

Constructor Summary

SearchIndex()
Default constructor.

Method Summary

void	addNode(NodeState node) Adds the node to the search index.
void	close() Closes this QueryHandler and frees resources attached to this handler.
protected org.apache.lucene.document.Document	createDocument(NodeState node, NamespaceMappings nsMappings) Creates a lucene Document for a node state using the namespace mappings nsMappings.
ExcerptProvider	createExcerptProvider(org.apache.lucene.search.Query query) Creates an excerpt provider for the given query.
ExecutableQuery	createExecutableQuery(SessionImpl session, ItemManager itemMgr, String statement, String language) Creates a new query by specifying the query statement itself and the language in which the query is stated.
protected org.apache.lucene.search.SortField[]	createSortFields(QName[] orderProps, boolean[] orderSpecs) Creates the SortFields for the order properties.
protected TextExtractor	createTextExtractor() Factory method to create the TextExtractor instance.
void	deleteNode(NodeId id) Removes the node with uuid from the search index.
protected void	doInit() Initializes this QueryHandler.
QueryHits	executeQuery(QueryImpl queryImpl, org.apache.lucene.search.Query query, QName[] orderProps, boolean[] orderSpecs) Executes the query on the search index.
String	getAnalyzer() Returns the class name of the analyzer that is currently in use.
boolean	getAutoRepair()
int	getBufferSize() Returns the current value for the buffer size.
int	getCacheSize()
String	getExcerptProviderClass()
int	getExtractorBackLogSize()
int	getExtractorPoolSize()
long	getExtractorTimeout()
boolean	getForceConsistencyCheck()
protected MultiIndex	getIndex() Returns the actual index.
org.apache.lucene.index.IndexReader	getIndexReader() Returns an index reader for this search index.
int	getMaxFieldLength()
int	getMaxMergeDocs() Returns the current value for maxMergeDocs.
int	getMergeFactor() Returns the current value for the merge factor.
int	getMinMergeDocs() Returns the current value for minMergeDocs.
NamespaceMappings	getNamespaceMappings() Returns the namespace mappings for the internal representation.
String	getPath() Returns the location of the search index.
boolean	getRespectDocumentOrder()
int	getResultFetchSize()
boolean	getSupportHighlighting()
org.apache.lucene.analysis.Analyzer	getTextAnalyzer() Returns the analyzer in use for indexing.
TextExtractor	getTextExtractor() Returns the text extractor in use for indexing.
String	getTextFilterClasses() Returns the fully qualified class names of the text filter instances currently in use.
boolean	getUseCompoundFile() Returns the current value for useCompoundFile.
int	getVolatileIdleTime() Returns the current value for volatileIdleTime.
void	setAnalyzer(String analyzerClassName) Sets the analyzer in use for indexing.
void	setAutoRepair(boolean b)
void	setBufferSize(int size)
void	setCacheSize(int size)
void	setExcerptProviderClass(String className) Sets the class name for the ExcerptProvider that should be used for the rep:excerpt pseudo property in a query.
void	setExtractorBackLogSize(int backLog) The number of extractor jobs that are queued until a new job is executed with the current thread instead of using the thread pool.
void	setExtractorPoolSize(int numThreads) The number of background threads for the extractor pool.
void	setExtractorTimeout(long timeout) The timeout in milliseconds which is granted to the text extraction process until fulltext indexing is deferred to a background thread.
void	setForceConsistencyCheck(boolean b)
void	setMaxFieldLength(int length)
void	setMaxMergeDocs(int maxMergeDocs) The lucene index writer property: maxMergeDocs
void	setMergeFactor(int mergeFactor) The lucene index writer property: mergeFactor
void	setMinMergeDocs(int minMergeDocs) The lucene index writer property: minMergeDocs
void	setPath(String path) Sets the location of the search index.
void	setRespectDocumentOrder(boolean docOrder)
void	setResultFetchSize(int size) Tells the query handler how many result should be fetched initially when a query is executed.
void	setSupportHighlighting(boolean b) If set to true additional information is stored in the index to support highlighting using the rep:excerpt pseudo property.
void	setTextFilterClasses(String filterClasses) Sets the list of text extractors (and text filters) to use for extracting text content from binary properties.
void	setUseCompoundFile(boolean b) The lucene index writer property: useCompoundFile
void	setVolatileIdleTime(int volatileIdleTime) Sets the property: volatileIdleTime
void	updateNodes(NodeIdIterator remove, NodeStateIterator add) This implementation forwards the call to MultiIndex.update(java.util.Iterator, java.util.Iterator) and transforms the two iterators to the required types.

Methods inherited from class org.apache.jackrabbit.core.query.AbstractQueryHandler

getContext, init

Methods inherited from class java.lang.Object

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

Field Detail

DEFAULT_MIN_MERGE_DOCS

public static final int DEFAULT_MIN_MERGE_DOCS

The default value for property minMergeDocs.

See Also:

Constant Field Values

DEFAULT_MAX_MERGE_DOCS

public static final int DEFAULT_MAX_MERGE_DOCS

The default value for property maxMergeDocs.

See Also:

Constant Field Values

DEFAULT_MERGE_FACTOR

public static final int DEFAULT_MERGE_FACTOR

the default value for property mergeFactor.

See Also:

Constant Field Values

DEFAULT_MAX_FIELD_LENGTH

public static final int DEFAULT_MAX_FIELD_LENGTH

the default value for property maxFieldLength.

See Also:

Constant Field Values

DEFAULT_EXTRACTOR_POOL_SIZE

public static final int DEFAULT_EXTRACTOR_POOL_SIZE

The default value for property extractorPoolSize.

See Also:

Constant Field Values

DEFAULT_EXTRACTOR_BACK_LOG

public static final int DEFAULT_EXTRACTOR_BACK_LOG

The default value for property extractorBackLog.

See Also:

Constant Field Values

DEFAULT_EXTRACTOR_TIMEOUT

public static final long DEFAULT_EXTRACTOR_TIMEOUT

The default timeout in milliseconds which is granted to the text extraction process until fulltext indexing is deferred to a background thread.

See Also:

Constant Field Values

Constructor Detail

SearchIndex

public SearchIndex()

Default constructor.

Method Detail

doInit

protected void doInit()
               throws IOException

Initializes this QueryHandler. This implementation requires that a path parameter is set in the configuration. If this condition is not met, a IOException is thrown.

Specified by:

doInit in class AbstractQueryHandler

Throws:

IOException - if an error occurs while initializing this handler.

addNode

public void addNode(NodeState node)
             throws RepositoryException,
                    IOException

Adds the node to the search index.

Parameters:

node - the node to add.

Throws:

RepositoryException - if an error occurs while indexing the node.

IOException - if an error occurs while adding the node to the index.

deleteNode

public void deleteNode(NodeId id)
                throws IOException

Removes the node with uuid from the search index.

Parameters:

id - the id of the node to remove from the index.

Throws:

IOException - if an error occurs while removing the node from the index.

updateNodes

public void updateNodes(NodeIdIterator remove,
                        NodeStateIterator add)
                 throws RepositoryException,
                        IOException

This implementation forwards the call to MultiIndex.update(java.util.Iterator, java.util.Iterator) and transforms the two iterators to the required types.

Specified by:

updateNodes in interface QueryHandler

Overrides:

updateNodes in class AbstractQueryHandler

Parameters:

remove - uuids of nodes to remove.

add - NodeStates to add. Calls to next() on this iterator may return null, to indicate that a node could not be indexed successfully.

Throws:

RepositoryException - if an error occurs while indexing a node.

IOException - if an error occurs while updating the index.

createExecutableQuery

public ExecutableQuery createExecutableQuery(SessionImpl session,
                                             ItemManager itemMgr,
                                             String statement,
                                             String language)
                                      throws InvalidQueryException

Creates a new query by specifying the query statement itself and the language in which the query is stated. If the query statement is syntactically invalid, given the language specified, an InvalidQueryException is thrown. language must specify a query language string from among those returned by QueryManager.getSupportedQueryLanguages(); if it is not then an InvalidQueryException is thrown.

Parameters:

session - the session of the current user creating the query object.

itemMgr - the item manager of the current user.

statement - the query statement.

language - the syntax of the query statement.

Returns:

A Query object.

Throws:

InvalidQueryException - if statement is invalid or language is unsupported.

close

public void close()

Closes this QueryHandler and frees resources attached to this handler.

executeQuery

public QueryHits executeQuery(QueryImpl queryImpl,
                              org.apache.lucene.search.Query query,
                              QName[] orderProps,
                              boolean[] orderSpecs)
                       throws IOException

Executes the query on the search index.

Parameters:

queryImpl - the query impl.

query - the lucene query.

orderProps - name of the properties for sort order.

orderSpecs - the order specs for the sort order properties. true indicates ascending order, false indicates descending.

Returns:

the lucene Hits object.

Throws:

IOException - if an error occurs while searching the index.

createExcerptProvider

public ExcerptProvider createExcerptProvider(org.apache.lucene.search.Query query)
                                      throws IOException

Creates an excerpt provider for the given query.

Parameters:

query - the query.

Returns:

an excerpt provider for the given query.

Throws:

IOException - if the provider cannot be created.

getTextAnalyzer

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

Returns the analyzer in use for indexing.

Returns:

the analyzer in use for indexing.

getTextExtractor

public TextExtractor getTextExtractor()

Returns the text extractor in use for indexing.

Returns:

the text extractor in use for indexing.

getNamespaceMappings

public NamespaceMappings getNamespaceMappings()

Returns the namespace mappings for the internal representation.

Returns:

the namespace mappings for the internal representation.

getIndexReader

public org.apache.lucene.index.IndexReader getIndexReader()
                                                   throws IOException

Returns an index reader for this search index. The caller of this method is responsible for closing the index reader when he is finished using it.

Returns:

an index reader for this search index.

Throws:

IOException - the index reader cannot be obtained.

createSortFields

protected org.apache.lucene.search.SortField[] createSortFields(QName[] orderProps,
                                                                boolean[] orderSpecs)

Creates the SortFields for the order properties.

Parameters:

orderProps - the order properties.

orderSpecs - the order specs for the properties.

Returns:

an array of sort fields

createDocument

protected org.apache.lucene.document.Document createDocument(NodeState node,
                                                             NamespaceMappings nsMappings)
                                                      throws RepositoryException

Creates a lucene Document for a node state using the namespace mappings nsMappings.

Parameters:

node - the node state to index.

nsMappings - the namespace mappings of the search index.

Returns:

a lucene Document that contains all properties of node.

Throws:

RepositoryException - if an error occurs while indexing the node.

getIndex

protected MultiIndex getIndex()

Returns the actual index.

Returns:

the actual index.

createTextExtractor

protected TextExtractor createTextExtractor()

Factory method to create the TextExtractor instance.

Returns:

the TextExtractor instance this index should use.

setAnalyzer

public void setAnalyzer(String analyzerClassName)

Sets the analyzer in use for indexing. The given analyzer class name must satisfy the following conditions:

• the class must exist in the class path
• the class must have a public default constructor
• the class must be a Lucene Analyzer

If the above conditions are met, then a new instance of the class is set as the analyzer. Otherwise a warning is logged and the current analyzer is not changed.

This property setter method is normally invoked by the Jackrabbit configuration mechanism if the "analyzer" parameter is set in the search configuration.

Parameters:

analyzerClassName - the analyzer class name

getAnalyzer

public String getAnalyzer()

Returns the class name of the analyzer that is currently in use.

Returns:

class name of analyzer in use.

setPath

public void setPath(String path)

Sets the location of the search index.

Parameters:

path - the location of the search index.

getPath

public String getPath()

Returns the location of the search index. Returns null if not set.

Returns:

the location of the search index.

setUseCompoundFile

public void setUseCompoundFile(boolean b)

The lucene index writer property: useCompoundFile

getUseCompoundFile

public boolean getUseCompoundFile()

Returns the current value for useCompoundFile.

Returns:

the current value for useCompoundFile.

setMinMergeDocs

public void setMinMergeDocs(int minMergeDocs)

The lucene index writer property: minMergeDocs

getMinMergeDocs

public int getMinMergeDocs()

Returns the current value for minMergeDocs.

Returns:

the current value for minMergeDocs.

setVolatileIdleTime

public void setVolatileIdleTime(int volatileIdleTime)

Sets the property: volatileIdleTime

Parameters:

volatileIdleTime - idle time in seconds

getVolatileIdleTime

public int getVolatileIdleTime()

Returns the current value for volatileIdleTime.

Returns:

the current value for volatileIdleTime.

setMaxMergeDocs

public void setMaxMergeDocs(int maxMergeDocs)

The lucene index writer property: maxMergeDocs

getMaxMergeDocs

public int getMaxMergeDocs()

Returns the current value for maxMergeDocs.

Returns:

the current value for maxMergeDocs.

setMergeFactor

public void setMergeFactor(int mergeFactor)

The lucene index writer property: mergeFactor

getMergeFactor

public int getMergeFactor()

Returns the current value for the merge factor.

Returns:

the current value for the merge factor.

setBufferSize

public void setBufferSize(int size)

See Also:

VolatileIndex#setBufferSize(int)

getBufferSize

public int getBufferSize()

Returns the current value for the buffer size.

Returns:

the current value for the buffer size.

setRespectDocumentOrder

public void setRespectDocumentOrder(boolean docOrder)

getRespectDocumentOrder

public boolean getRespectDocumentOrder()

setForceConsistencyCheck

public void setForceConsistencyCheck(boolean b)

getForceConsistencyCheck

public boolean getForceConsistencyCheck()

setAutoRepair

public void setAutoRepair(boolean b)

getAutoRepair

public boolean getAutoRepair()

setCacheSize

public void setCacheSize(int size)

getCacheSize

public int getCacheSize()

setMaxFieldLength

public void setMaxFieldLength(int length)

getMaxFieldLength

public int getMaxFieldLength()

setTextFilterClasses

public void setTextFilterClasses(String filterClasses)

Sets the list of text extractors (and text filters) to use for extracting text content from binary properties. The list must be comma (or whitespace) separated, and contain fully qualified class names of the TextExtractor (and TextFilter) classes to be used. The configured classes must all have a public default constructor.

Parameters:

filterClasses - comma separated list of class names

getTextFilterClasses

public String getTextFilterClasses()

Returns the fully qualified class names of the text filter instances currently in use. The names are comma separated.

Returns:

class names of the text filters in use.

setResultFetchSize

public void setResultFetchSize(int size)

Tells the query handler how many result should be fetched initially when a query is executed.

Parameters:

size - the number of results to fetch initially.

getResultFetchSize

public int getResultFetchSize()

Returns:

the number of results the query handler will fetch initially when a query is executed.

setExtractorPoolSize

public void setExtractorPoolSize(int numThreads)

The number of background threads for the extractor pool.

Parameters:

numThreads - the number of threads.

getExtractorPoolSize

public int getExtractorPoolSize()

Returns:

the size of the thread pool which is used to run the text extractors when binary content is indexed.

setExtractorBackLogSize

public void setExtractorBackLogSize(int backLog)

The number of extractor jobs that are queued until a new job is executed with the current thread instead of using the thread pool.

Parameters:

backLog - size of the extractor job queue.

getExtractorBackLogSize

public int getExtractorBackLogSize()

Returns:

the size of the extractor queue back log.

setExtractorTimeout

public void setExtractorTimeout(long timeout)

The timeout in milliseconds which is granted to the text extraction process until fulltext indexing is deferred to a background thread.

Parameters:

timeout - the timeout in milliseconds.

getExtractorTimeout

public long getExtractorTimeout()

Returns:

the extractor timeout in milliseconds.

setSupportHighlighting

public void setSupportHighlighting(boolean b)

If set to true additional information is stored in the index to support highlighting using the rep:excerpt pseudo property.

Parameters:

b - true to enable highlighting support.

getSupportHighlighting

public boolean getSupportHighlighting()

Returns:

true if highlighting support is enabled.

setExcerptProviderClass

public void setExcerptProviderClass(String className)

Sets the class name for the ExcerptProvider that should be used for the rep:excerpt pseudo property in a query.

Parameters:

className - the name of a class that implements ExcerptProvider.

getExcerptProviderClass

public String getExcerptProviderClass()

Returns:

the class name of the excerpt provider implementation.