org.gcube.search.datafusion.helpers

Class IndexHelper

java.lang.Object

org.gcube.search.datafusion.helpers.IndexHelper

All Implemented Interfaces:

Serializable

public class IndexHelper
extends Object
implements Serializable

Author:

Alex Antoniadis

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
IndexHelper()

Method Summary

Methods

Modifier and Type	Method and Description
static Map<String,Long>	feedLucene(org.apache.lucene.store.Directory index, Iterator<PositionalRecordWrapper> iter, List<String> snippetFields, Set<String> fields, int count) Feeds the lucene index with (maximum count) records that are retrieved from the iter.
static org.apache.lucene.store.Directory	initializeIndex()
static org.apache.lucene.index.IndexWriter	initializeIndexWriter(org.apache.lucene.store.Directory index)
static List<gr.uoa.di.madgik.grs.record.GenericRecord>	multiget(List<RankedRecord> recs, org.apache.lucene.store.Directory index, Set<String> fieldsName) Performs a multiget on the lucene index from the given list of RankedRecords
static Map<String,Pair>	rankDocuments(org.apache.lucene.store.Directory index, String query, int count)
static Map<String,Pair>	rankDocuments(org.apache.lucene.store.Directory index, String query, int count, int hitsCount) Performs the query on the index and returns a map with key the id of the record and value a pair of score and lucene id (different from record id) Obviously, the returned result is not sorted but it requires sorting on the score

Methods inherited from class java.lang.Object

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

Constructor Detail

IndexHelper

public IndexHelper()

Method Detail

initializeIndex

public static org.apache.lucene.store.Directory initializeIndex()
                                                         throws IOException

Throws:

IOException

initializeIndexWriter

public static org.apache.lucene.index.IndexWriter initializeIndexWriter(org.apache.lucene.store.Directory index)
                                                                 throws IOException

Throws:

IOException

rankDocuments

public static Map<String,Pair> rankDocuments(org.apache.lucene.store.Directory index,
                             String query,
                             int count)
                                      throws org.apache.lucene.queryparser.classic.ParseException,
                                             IOException

Throws:

org.apache.lucene.queryparser.classic.ParseException

IOException

rankDocuments

public static Map<String,Pair> rankDocuments(org.apache.lucene.store.Directory index,
                             String query,
                             int count,
                             int hitsCount)
                                      throws org.apache.lucene.queryparser.classic.ParseException,
                                             IOException

Performs the query on the index and returns a map with key the id of the record and value a pair of score and lucene id (different from record id) Obviously, the returned result is not sorted but it requires sorting on the score

Parameters:

index -

query -

count -

hitsCount -

Returns:

map of recordIDs -> (score, luceneDocID)

Throws:

org.apache.lucene.queryparser.classic.ParseException

IOException

multiget

public static List<gr.uoa.di.madgik.grs.record.GenericRecord> multiget(List<RankedRecord> recs,
                                                       org.apache.lucene.store.Directory index,
                                                       Set<String> fieldsName)
                                                                throws Exception

Performs a multiget on the lucene index from the given list of RankedRecords

Parameters:

recs -

index -

fieldsName -

Returns:

list of GenericRecords retrieved

Throws:

Exception

feedLucene

public static Map<String,Long> feedLucene(org.apache.lucene.store.Directory index,
                          Iterator<PositionalRecordWrapper> iter,
                          List<String> snippetFields,
                          Set<String> fields,
                          int count)
                                   throws Exception

Feeds the lucene index with (maximum count) records that are retrieved from the iter. A payload field that will be used for querying later is generated either from fields that exist in the snippetFields or by the actual content of the record (objectID). Also, each fieldName that is faced in records from iter are is gathered to contstruct a union of all fields from the (different type of) records in the iter. It returns a map that has recordID as key and (the initial) record position as value. Initial record position is the position that the record had before fusion in its iterator.

Parameters:

index -

iter -

snippetFields -

fields -

count -

Returns:

map recordID -> position

Throws:

Exception