oracle.pg.text.lucene

Class LuceneEdgeIndex

java.lang.Object

oracle.pg.text.lucene.LuceneIndex<Edge>

oracle.pg.text.lucene.LuceneEdgeIndex

All Implemented Interfaces:

Index<Edge>, OracleIndex<Edge>

public class LuceneEdgeIndex
extends LuceneIndex<Edge>

This class extends the functions defined in LuceneIndex. It manages manual indexes defined on Edge objects

Field Summary

Fields inherited from interface oracle.pg.text.OracleIndex

EDGE_ID_PROP, EDGE_LABEL_PROP, END_NODE_ID_PROP, START_NODE_ID_PROP, TYPE_DT_BOOL, TYPE_DT_DATE, TYPE_DT_DOUBLE, TYPE_DT_FLOAT, TYPE_DT_INTEGER, TYPE_DT_SERIALIZABLE, TYPE_DT_STRING, VERTEX_ID_PROP

Constructor Summary

Constructors

Constructor and Description
LuceneEdgeIndex(String indexName, OraclePropertyGraphBase oraclePropertyGraph, Parameter<String,Object>[] indexParameters, boolean erase) Constructs a new LuceneEdgeIndex object.

Method Summary

Methods

Modifier and Type	Method and Description
CloseableIterable<Edge>	get(String key, Object value, Class dtClass, boolean acceptWildcard) Gets all edges indexed by the provided key/value.
String	getDocIdentifier() Gets the document's identifier associated to this index.
CloseableIterable<Edge>[]	getPartitioned(Object[] connections, String key, Object value, Class dtClass, boolean acceptWildcard, int startShardId) Gets an array of CloseableIterable objects that hold all the edges with an attribute matching the given K/V pair.
void	refreshIndex(long lStartSCN, long lEndSCN) Auto-refresh an index based on all insert/deletes between the specified start and end SCN.

Methods inherited from class oracle.pg.text.lucene.LuceneIndex

add, addToDeleteCache, buildSearchTerm, buildSearchTerm, clearIndex, close, commit, count, count, count, deleteExistingDirectories, dropIndex, get, get, getDirectory, getDirectoryName, getGraphName, getIndexClass, getIndexName, getIndexParameters, getOracleDirectory, getOracleIndexParameters, getOracleSearcherManager, getParameters, getPartitioned, getSearcherFactory, getTotalShards, isAutoIndex, isDelBatchEmpty, isDelBatchFull, isUpdatesBatchEmpty, isUpdatesBatchFull, put, put, query, query, remove, remove, remove, remove, removeAll, removeAll

Methods inherited from class java.lang.Object

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

Constructor Detail

LuceneEdgeIndex

public LuceneEdgeIndex(String indexName,
               OraclePropertyGraphBase oraclePropertyGraph,
               Parameter<String,Object>[] indexParameters,
               boolean erase)

Constructs a new LuceneEdgeIndex object.

Parameters:

indexName - the name of the index

oraclePropertyGraph - an OraclePropertyGraphBase object.

indexParameters - a collection of parameters for the underlying index implementation.

erase - a boolean specifying if the index should be cleared at the beginning.

Method Detail

get

public CloseableIterable<Edge> get(String key,
                          Object value,
                          Class dtClass,
                          boolean acceptWildcard)

Gets all edges indexed by the provided key/value.

Specified by:

get in interface OracleIndex<Edge>

Specified by:

get in class LuceneIndex<Edge>

Parameters:

key - the key of the indexed elements.

value - the value of the indexed elements.

dtClass - the class representing the datatype of the property value. This can be used to cast the property value to a different datatype. For example, if the value is "3.01" and dtValue is Float.Class, then this value will be treated as a 3.01f.

acceptWildcard - specifies if wild cards can be used in the value object. the index

Returns:

an iterable of elements that have a particular key/value in the index.

getDocIdentifier

public String getDocIdentifier()

Gets the document's identifier associated to this index.

Specified by:

getDocIdentifier in interface OracleIndex<Edge>

Overrides:

getDocIdentifier in class LuceneIndex<Edge>

Returns:

the document's identifier associated to this index.

getPartitioned

public CloseableIterable<Edge>[] getPartitioned(Object[] connections,
                                       String key,
                                       Object value,
                                       Class dtClass,
                                       boolean acceptWildcard,
                                       int startShardId)

Gets an array of CloseableIterable objects that hold all the edges with an attribute matching the given K/V pair. Each edge in the CloseableIterable array uses a separate connection provided to fetch a subset of the results from a corresponding Lucene sub-directory. Note that we assign an integer ID (in the range of [0, N - 1]) to all the sub-directories in a Lucene index with N shards. The subset of shards (subdirectories) queried will consist of those shards with ID value in the range of [startShardId, startShardId - 1 + size of connections array].

Specified by:

getPartitioned in interface OracleIndex<Edge>

Specified by:

getPartitioned in class LuceneIndex<Edge>

Parameters:

connections - an array of connections to the Lucene index Each edge of this array should be of type SearcherManager. See LuceneIndex.getOracleSearcherManager(int).

key - the key of a property

value - the value object of the property.

acceptWildcard - specifies if wild cards can be used in the value object.

startShardId - the ID of the starting shard.

dtClass - the class representing the datatype of the property value. This can be used to cast the property value to a different datatype. For example, if the value is "3.01" and dtValue is Float.Class, then this value will be treated as a 3.01f.

Returns:

a CloseableIterable object.

refreshIndex

public void refreshIndex(long lStartSCN,
                long lEndSCN)

Auto-refresh an index based on all insert/deletes between the specified start and end SCN.

Specified by:

refreshIndex in interface OracleIndex<Edge>

Specified by:

refreshIndex in class LuceneIndex<Edge>

Parameters:

lStartSCN - start SCN value. A NULL value will be treated as MINVALUE to the underlying Oracle Database.

lEndSCN - end SCN value. A NULL value will be treated as MAXVALUE to the underlying Oracle Database.