Package org.apache.lucene.search

Class TopDocs

java.lang.Object

org.apache.lucene.search.TopDocs

Direct Known Subclasses:

TopFieldDocs

public class TopDocs
extends Object

Represents hits returned by IndexSearcher.search(Query,int).

Field Summary

Fields

Modifier and Type	Field	Description
ScoreDoc[]	scoreDocs	The top hits for the query.
TotalHits	totalHits	The total number of hits for the query.

Constructor Summary

Constructors

Constructor	Description
TopDocs(TotalHits totalHits, ScoreDoc[] scoreDocs)	Constructs a TopDocs.

Method Summary

Modifier and Type	Method	Description
static TopDocs	merge(int start, int topN, TopDocs[] shardHits)	Same as merge(int, TopDocs[]) but also ignores the top start top docs.
static TopDocs	merge(int start, int topN, TopDocs[] shardHits, Comparator<ScoreDoc> tieBreaker)	Same as above, but accepts the passed in tie breaker
static TopDocs	merge(int topN, TopDocs[] shardHits)	Returns a new TopDocs, containing topN results across the provided TopDocs, sorting by score.
static TopFieldDocs	merge(Sort sort, int start, int topN, TopFieldDocs[] shardHits)	Same as merge(Sort, int, TopFieldDocs[]) but also ignores the top start top docs.
static TopFieldDocs	merge(Sort sort, int start, int topN, TopFieldDocs[] shardHits, Comparator<ScoreDoc> tieBreaker)	Pass in a custom tie breaker for ordering results
static TopFieldDocs	merge(Sort sort, int topN, TopFieldDocs[] shardHits)	Returns a new TopFieldDocs, containing topN results across the provided TopFieldDocs, sorting by the specified Sort.
static TopDocs	rrf(int topN, int k, TopDocs[] hits)	Reciprocal Rank Fusion method.

Methods inherited from class java.lang.Object

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

Field Details

totalHits

public TotalHits totalHits

The total number of hits for the query.

scoreDocs

public ScoreDoc[] scoreDocs

The top hits for the query.

Constructor Details

TopDocs

public TopDocs(TotalHits totalHits,
 ScoreDoc[] scoreDocs)

Constructs a TopDocs.

Method Details

merge

public static TopDocs merge(int topN,
 TopDocs[] shardHits)

Returns a new TopDocs, containing topN results across the provided TopDocs, sorting by score. Each TopDocs instance must be sorted.

See Also:

• merge(int, int, TopDocs[])

WARNING: This API is experimental and might change in incompatible ways in the next release.

merge

public static TopDocs merge(int start,
 int topN,
 TopDocs[] shardHits)

Same as merge(int, TopDocs[]) but also ignores the top start top docs. This is typically useful for pagination.

docIDs are expected to be in consistent pattern i.e. either all ScoreDocs have their shardIndex set, or all have them as -1 (signifying that all hits belong to same searcher)

WARNING: This API is experimental and might change in incompatible ways in the next release.

merge

public static TopDocs merge(int start,
 int topN,
 TopDocs[] shardHits,
 Comparator<ScoreDoc> tieBreaker)

Same as above, but accepts the passed in tie breaker

docIDs are expected to be in consistent pattern i.e. either all ScoreDocs have their shardIndex set, or all have them as -1 (signifying that all hits belong to same searcher)

WARNING: This API is experimental and might change in incompatible ways in the next release.

merge

public static TopFieldDocs merge(Sort sort,
 int topN,
 TopFieldDocs[] shardHits)

Returns a new TopFieldDocs, containing topN results across the provided TopFieldDocs, sorting by the specified Sort. Each of the TopDocs must have been sorted by the same Sort, and sort field values must have been filled.

See Also:

• merge(Sort, int, int, TopFieldDocs[])

WARNING: This API is experimental and might change in incompatible ways in the next release.

merge

public static TopFieldDocs merge(Sort sort,
 int start,
 int topN,
 TopFieldDocs[] shardHits)

Same as merge(Sort, int, TopFieldDocs[]) but also ignores the top start top docs. This is typically useful for pagination.

docIDs are expected to be in consistent pattern i.e. either all ScoreDocs have their shardIndex set, or all have them as -1 (signifying that all hits belong to same searcher)

WARNING: This API is experimental and might change in incompatible ways in the next release.

merge

public static TopFieldDocs merge(Sort sort,
 int start,
 int topN,
 TopFieldDocs[] shardHits,
 Comparator<ScoreDoc> tieBreaker)

Pass in a custom tie breaker for ordering results

WARNING: This API is experimental and might change in incompatible ways in the next release.

rrf

public static TopDocs rrf(int topN,
 int k,
 TopDocs[] hits)

Reciprocal Rank Fusion method.

This method combines different search results into a single ranked list by combining their ranks. This is especially well suited when combining hits computed via different methods, whose score distributions are hardly comparable.

Parameters:

topN - the top N results to be returned

k - a constant determines how much influence documents in individual rankings have on the final result. A higher value gives lower rank documents more influence. k should be greater than or equal to 1.

hits - a list of TopDocs to apply RRF on

Returns:

a TopDocs contains the top N ranked results.