Revert "Improve `min_score` implementation."
This reverts commit 56ea65b0f8
.
This commit is contained in:
parent
cb3122baab
commit
f69c350960
|
@ -0,0 +1,68 @@
|
|||
/*
|
||||
* Licensed to Elasticsearch under one or more contributor
|
||||
* license agreements. See the NOTICE file distributed with
|
||||
* this work for additional information regarding copyright
|
||||
* ownership. Elasticsearch licenses this file to you under
|
||||
* the Apache License, Version 2.0 (the "License"); you may
|
||||
* not use this file except in compliance with the License.
|
||||
* You may obtain a copy of the License at
|
||||
*
|
||||
* http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* Unless required by applicable law or agreed to in writing,
|
||||
* software distributed under the License is distributed on an
|
||||
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
|
||||
* KIND, either express or implied. See the License for the
|
||||
* specific language governing permissions and limitations
|
||||
* under the License.
|
||||
*/
|
||||
|
||||
package org.elasticsearch.common.lucene;
|
||||
|
||||
import org.apache.lucene.index.LeafReaderContext;
|
||||
import org.apache.lucene.search.*;
|
||||
|
||||
import java.io.IOException;
|
||||
|
||||
/**
|
||||
*
|
||||
*/
|
||||
public class MinimumScoreCollector extends SimpleCollector {
|
||||
|
||||
private final Collector collector;
|
||||
private final float minimumScore;
|
||||
|
||||
private Scorer scorer;
|
||||
private LeafCollector leafCollector;
|
||||
|
||||
public MinimumScoreCollector(Collector collector, float minimumScore) {
|
||||
this.collector = collector;
|
||||
this.minimumScore = minimumScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public void setScorer(Scorer scorer) throws IOException {
|
||||
if (!(scorer instanceof ScoreCachingWrappingScorer)) {
|
||||
scorer = new ScoreCachingWrappingScorer(scorer);
|
||||
}
|
||||
this.scorer = scorer;
|
||||
leafCollector.setScorer(scorer);
|
||||
}
|
||||
|
||||
@Override
|
||||
public void collect(int doc) throws IOException {
|
||||
if (scorer.score() >= minimumScore) {
|
||||
leafCollector.collect(doc);
|
||||
}
|
||||
}
|
||||
|
||||
@Override
|
||||
public void doSetNextReader(LeafReaderContext context) throws IOException {
|
||||
leafCollector = collector.getLeafCollector(context);
|
||||
}
|
||||
|
||||
@Override
|
||||
public boolean needsScores() {
|
||||
return true;
|
||||
}
|
||||
}
|
|
@ -1,285 +0,0 @@
|
|||
/*
|
||||
* Licensed to Elasticsearch under one or more contributor
|
||||
* license agreements. See the NOTICE file distributed with
|
||||
* this work for additional information regarding copyright
|
||||
* ownership. Elasticsearch licenses this file to you under
|
||||
* the Apache License, Version 2.0 (the "License"); you may
|
||||
* not use this file except in compliance with the License.
|
||||
* You may obtain a copy of the License at
|
||||
*
|
||||
* http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* Unless required by applicable law or agreed to in writing,
|
||||
* software distributed under the License is distributed on an
|
||||
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
|
||||
* KIND, either express or implied. See the License for the
|
||||
* specific language governing permissions and limitations
|
||||
* under the License.
|
||||
*/
|
||||
|
||||
package org.elasticsearch.common.lucene.search;
|
||||
|
||||
import org.apache.lucene.index.IndexReader;
|
||||
import org.apache.lucene.index.LeafReaderContext;
|
||||
import org.apache.lucene.index.Term;
|
||||
import org.apache.lucene.search.BulkScorer;
|
||||
import org.apache.lucene.search.DocIdSetIterator;
|
||||
import org.apache.lucene.search.Explanation;
|
||||
import org.apache.lucene.search.IndexSearcher;
|
||||
import org.apache.lucene.search.LeafCollector;
|
||||
import org.apache.lucene.search.Query;
|
||||
import org.apache.lucene.search.ScoreCachingWrappingScorer;
|
||||
import org.apache.lucene.search.Scorer;
|
||||
import org.apache.lucene.search.TwoPhaseIterator;
|
||||
import org.apache.lucene.search.Weight;
|
||||
import org.apache.lucene.util.Bits;
|
||||
|
||||
import java.io.IOException;
|
||||
import java.util.Objects;
|
||||
import java.util.Set;
|
||||
|
||||
/**
|
||||
* A {@link Query} wrapper that only emits as hits documents whose score is
|
||||
* above a given threshold. This query only really makes sense for queries
|
||||
* whose score is computed manually, like eg. function score queries.
|
||||
*/
|
||||
public final class MinScoreQuery extends Query {
|
||||
|
||||
private final Query query;
|
||||
private final float minScore;
|
||||
private final IndexSearcher searcher;
|
||||
|
||||
/** Sole constructor. */
|
||||
public MinScoreQuery(Query query, float minScore) {
|
||||
this(query, minScore, null);
|
||||
}
|
||||
|
||||
MinScoreQuery(Query query, float minScore, IndexSearcher searcher) {
|
||||
this.query = query;
|
||||
this.minScore = minScore;
|
||||
this.searcher = searcher;
|
||||
}
|
||||
|
||||
/** Return the wrapped query. */
|
||||
public Query getQuery() {
|
||||
return query;
|
||||
}
|
||||
|
||||
/** Return the minimum score. */
|
||||
public float getMinScore() {
|
||||
return minScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public String toString(String field) {
|
||||
return getClass().getSimpleName() + "(" + query.toString(field) + ", minScore=" + minScore + ")";
|
||||
}
|
||||
|
||||
@Override
|
||||
public boolean equals(Object obj) {
|
||||
if (super.equals(obj) == false) {
|
||||
return false;
|
||||
}
|
||||
MinScoreQuery that = (MinScoreQuery) obj;
|
||||
return minScore == that.minScore
|
||||
&& searcher == that.searcher
|
||||
&& query.equals(that.query);
|
||||
}
|
||||
|
||||
@Override
|
||||
public int hashCode() {
|
||||
return 31 * super.hashCode() + Objects.hash(query, minScore, searcher);
|
||||
}
|
||||
|
||||
@Override
|
||||
public Query rewrite(IndexReader reader) throws IOException {
|
||||
if (getBoost() != 1f) {
|
||||
return super.rewrite(reader);
|
||||
}
|
||||
Query rewritten = query.rewrite(reader);
|
||||
if (rewritten != query) {
|
||||
return new MinScoreQuery(rewritten, minScore);
|
||||
}
|
||||
return super.rewrite(reader);
|
||||
}
|
||||
|
||||
@Override
|
||||
public Weight createWeight(IndexSearcher searcher, boolean needsScores) throws IOException {
|
||||
final Weight weight = searcher.createWeight(query, true);
|
||||
// We specialize the query for the provided index searcher because it
|
||||
// can't really be cached as the documents that match depend on the
|
||||
// Similarity implementation and the top-level reader.
|
||||
final Query key = new MinScoreQuery(query, minScore, searcher);
|
||||
return new Weight(key) {
|
||||
|
||||
@Override
|
||||
public Scorer scorer(LeafReaderContext context) throws IOException {
|
||||
Scorer scorer = weight.scorer(context);
|
||||
if (scorer == null) {
|
||||
return null;
|
||||
}
|
||||
return new MinScoreScorer(this, scorer, minScore);
|
||||
}
|
||||
|
||||
@Override
|
||||
public BulkScorer bulkScorer(LeafReaderContext context) throws IOException {
|
||||
BulkScorer bulkScorer = weight.bulkScorer(context);
|
||||
if (bulkScorer == null) {
|
||||
return null;
|
||||
}
|
||||
return new MinScoreBulkScorer(bulkScorer, minScore);
|
||||
}
|
||||
|
||||
@Override
|
||||
public void normalize(float norm, float boost) {
|
||||
weight.normalize(norm, boost);
|
||||
}
|
||||
|
||||
@Override
|
||||
public float getValueForNormalization() throws IOException {
|
||||
return weight.getValueForNormalization();
|
||||
}
|
||||
|
||||
@Override
|
||||
public void extractTerms(Set<Term> terms) {
|
||||
weight.extractTerms(terms);
|
||||
}
|
||||
|
||||
@Override
|
||||
public Explanation explain(LeafReaderContext context, int doc) throws IOException {
|
||||
Explanation expl = weight.explain(context, doc);
|
||||
if (expl.isMatch() == false || expl.getValue() >= minScore) {
|
||||
return expl;
|
||||
} else {
|
||||
return Explanation.noMatch("Min score is less than the configured min score=" + minScore, expl);
|
||||
}
|
||||
}
|
||||
};
|
||||
}
|
||||
|
||||
private static class MinScoreScorer extends Scorer {
|
||||
|
||||
private final Scorer scorer;
|
||||
private final float minScore;
|
||||
private float score;
|
||||
|
||||
protected MinScoreScorer(Weight weight, Scorer scorer, float minScore) {
|
||||
super(weight);
|
||||
this.scorer = scorer;
|
||||
this.minScore = minScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public float score() throws IOException {
|
||||
return score;
|
||||
}
|
||||
|
||||
@Override
|
||||
public int freq() throws IOException {
|
||||
return scorer.freq();
|
||||
}
|
||||
|
||||
@Override
|
||||
public int docID() {
|
||||
return scorer.docID();
|
||||
}
|
||||
|
||||
@Override
|
||||
public int nextDoc() throws IOException {
|
||||
return doNext(scorer.nextDoc());
|
||||
}
|
||||
|
||||
@Override
|
||||
public int advance(int target) throws IOException {
|
||||
return doNext(scorer.advance(target));
|
||||
}
|
||||
|
||||
private int doNext(int doc) throws IOException {
|
||||
for (; doc != NO_MORE_DOCS; doc = scorer.nextDoc()) {
|
||||
final float score = scorer.score();
|
||||
if (score >= minScore) {
|
||||
this.score = score;
|
||||
return doc;
|
||||
}
|
||||
}
|
||||
return NO_MORE_DOCS;
|
||||
}
|
||||
|
||||
@Override
|
||||
public TwoPhaseIterator asTwoPhaseIterator() {
|
||||
final TwoPhaseIterator twoPhase = scorer.asTwoPhaseIterator();
|
||||
final DocIdSetIterator approximation = twoPhase == null
|
||||
? scorer
|
||||
: twoPhase.approximation();
|
||||
return new TwoPhaseIterator(approximation) {
|
||||
@Override
|
||||
public boolean matches() throws IOException {
|
||||
if (twoPhase != null && twoPhase.matches() == false) {
|
||||
return false;
|
||||
}
|
||||
score = scorer.score();
|
||||
return score >= minScore;
|
||||
}
|
||||
};
|
||||
}
|
||||
|
||||
@Override
|
||||
public long cost() {
|
||||
return scorer.cost();
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
private static class MinScoreBulkScorer extends BulkScorer {
|
||||
|
||||
private final BulkScorer bulkScorer;
|
||||
private final float minScore;
|
||||
|
||||
public MinScoreBulkScorer(BulkScorer bulkScorer, float minScore) {
|
||||
this.bulkScorer = bulkScorer;
|
||||
this.minScore = minScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public int score(LeafCollector collector, Bits acceptDocs, int min, int max) throws IOException {
|
||||
return bulkScorer.score(new MinScoreLeafCollector(collector, minScore), acceptDocs, min, max);
|
||||
}
|
||||
|
||||
@Override
|
||||
public long cost() {
|
||||
return bulkScorer.cost();
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
private static class MinScoreLeafCollector implements LeafCollector {
|
||||
|
||||
private final LeafCollector collector;
|
||||
private final float minScore;
|
||||
private Scorer scorer;
|
||||
|
||||
public MinScoreLeafCollector(LeafCollector collector, float minScore) {
|
||||
this.collector = collector;
|
||||
this.minScore = minScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public void setScorer(Scorer scorer) throws IOException {
|
||||
// we will need scores at least once, maybe more due to the wrapped
|
||||
// collector, so we wrap with a ScoreCachingWrappingScorer
|
||||
if (scorer instanceof ScoreCachingWrappingScorer == false) {
|
||||
scorer = new ScoreCachingWrappingScorer(scorer);
|
||||
}
|
||||
this.scorer = scorer;
|
||||
collector.setScorer(scorer);
|
||||
}
|
||||
|
||||
@Override
|
||||
public void collect(int doc) throws IOException {
|
||||
if (scorer.score() >= minScore) {
|
||||
collector.collect(doc);
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
}
|
|
@ -30,12 +30,21 @@ abstract class CustomBoostFactorScorer extends Scorer {
|
|||
final float maxBoost;
|
||||
final CombineFunction scoreCombiner;
|
||||
|
||||
CustomBoostFactorScorer(Weight w, Scorer scorer, float maxBoost, CombineFunction scoreCombiner)
|
||||
Float minScore;
|
||||
NextDoc nextDoc;
|
||||
|
||||
CustomBoostFactorScorer(Weight w, Scorer scorer, float maxBoost, CombineFunction scoreCombiner, Float minScore)
|
||||
throws IOException {
|
||||
super(w);
|
||||
if (minScore == null) {
|
||||
nextDoc = new AnyNextDoc();
|
||||
} else {
|
||||
nextDoc = new MinScoreNextDoc();
|
||||
}
|
||||
this.scorer = scorer;
|
||||
this.maxBoost = maxBoost;
|
||||
this.scoreCombiner = scoreCombiner;
|
||||
this.minScore = minScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
|
@ -45,16 +54,20 @@ abstract class CustomBoostFactorScorer extends Scorer {
|
|||
|
||||
@Override
|
||||
public int advance(int target) throws IOException {
|
||||
return scorer.advance(target);
|
||||
return nextDoc.advance(target);
|
||||
}
|
||||
|
||||
@Override
|
||||
public int nextDoc() throws IOException {
|
||||
return scorer.nextDoc();
|
||||
return nextDoc.nextDoc();
|
||||
}
|
||||
|
||||
public abstract float innerScore() throws IOException;
|
||||
|
||||
@Override
|
||||
public abstract float score() throws IOException;
|
||||
public float score() throws IOException {
|
||||
return nextDoc.score();
|
||||
}
|
||||
|
||||
@Override
|
||||
public int freq() throws IOException {
|
||||
|
@ -66,4 +79,64 @@ abstract class CustomBoostFactorScorer extends Scorer {
|
|||
return scorer.cost();
|
||||
}
|
||||
|
||||
public interface NextDoc {
|
||||
public int advance(int target) throws IOException;
|
||||
|
||||
public int nextDoc() throws IOException;
|
||||
|
||||
public float score() throws IOException;
|
||||
}
|
||||
|
||||
public class MinScoreNextDoc implements NextDoc {
|
||||
float currentScore = Float.MAX_VALUE * -1.0f;
|
||||
|
||||
@Override
|
||||
public int nextDoc() throws IOException {
|
||||
int doc;
|
||||
do {
|
||||
doc = scorer.nextDoc();
|
||||
if (doc == NO_MORE_DOCS) {
|
||||
return doc;
|
||||
}
|
||||
currentScore = innerScore();
|
||||
} while (currentScore < minScore);
|
||||
return doc;
|
||||
}
|
||||
|
||||
@Override
|
||||
public float score() throws IOException {
|
||||
return currentScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public int advance(int target) throws IOException {
|
||||
int doc = scorer.advance(target);
|
||||
if (doc == NO_MORE_DOCS) {
|
||||
return doc;
|
||||
}
|
||||
currentScore = innerScore();
|
||||
if (currentScore < minScore) {
|
||||
return scorer.nextDoc();
|
||||
}
|
||||
return doc;
|
||||
}
|
||||
}
|
||||
|
||||
public class AnyNextDoc implements NextDoc {
|
||||
|
||||
@Override
|
||||
public int nextDoc() throws IOException {
|
||||
return scorer.nextDoc();
|
||||
}
|
||||
|
||||
@Override
|
||||
public float score() throws IOException {
|
||||
return innerScore();
|
||||
}
|
||||
|
||||
@Override
|
||||
public int advance(int target) throws IOException {
|
||||
return scorer.advance(target);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
@ -100,15 +100,17 @@ public class FiltersFunctionScoreQuery extends Query {
|
|||
final FilterFunction[] filterFunctions;
|
||||
final ScoreMode scoreMode;
|
||||
final float maxBoost;
|
||||
private final Float minScore;
|
||||
|
||||
final protected CombineFunction combineFunction;
|
||||
|
||||
public FiltersFunctionScoreQuery(Query subQuery, ScoreMode scoreMode, FilterFunction[] filterFunctions, float maxBoost, CombineFunction combineFunction) {
|
||||
public FiltersFunctionScoreQuery(Query subQuery, ScoreMode scoreMode, FilterFunction[] filterFunctions, float maxBoost, Float minScore, CombineFunction combineFunction) {
|
||||
this.subQuery = subQuery;
|
||||
this.scoreMode = scoreMode;
|
||||
this.filterFunctions = filterFunctions;
|
||||
this.maxBoost = maxBoost;
|
||||
this.combineFunction = combineFunction;
|
||||
this.minScore = minScore;
|
||||
}
|
||||
|
||||
public Query getSubQuery() {
|
||||
|
@ -193,7 +195,7 @@ public class FiltersFunctionScoreQuery extends Query {
|
|||
Scorer filterScorer = filterWeights[i].scorer(context);
|
||||
docSets[i] = Lucene.asSequentialAccessBits(context.reader().maxDoc(), filterScorer);
|
||||
}
|
||||
return new FiltersFunctionFactorScorer(this, subQueryScorer, scoreMode, filterFunctions, maxBoost, functions, docSets, combineFunction, needsScores);
|
||||
return new FiltersFunctionFactorScorer(this, subQueryScorer, scoreMode, filterFunctions, maxBoost, functions, docSets, combineFunction, minScore, needsScores);
|
||||
}
|
||||
|
||||
@Override
|
||||
|
@ -242,8 +244,8 @@ public class FiltersFunctionScoreQuery extends Query {
|
|||
private final boolean needsScores;
|
||||
|
||||
private FiltersFunctionFactorScorer(CustomBoostFactorWeight w, Scorer scorer, ScoreMode scoreMode, FilterFunction[] filterFunctions,
|
||||
float maxBoost, LeafScoreFunction[] functions, Bits[] docSets, CombineFunction scoreCombiner, boolean needsScores) throws IOException {
|
||||
super(w, scorer, maxBoost, scoreCombiner);
|
||||
float maxBoost, LeafScoreFunction[] functions, Bits[] docSets, CombineFunction scoreCombiner, Float minScore, boolean needsScores) throws IOException {
|
||||
super(w, scorer, maxBoost, scoreCombiner, minScore);
|
||||
this.scoreMode = scoreMode;
|
||||
this.filterFunctions = filterFunctions;
|
||||
this.functions = functions;
|
||||
|
@ -252,7 +254,7 @@ public class FiltersFunctionScoreQuery extends Query {
|
|||
}
|
||||
|
||||
@Override
|
||||
public float score() throws IOException {
|
||||
public float innerScore() throws IOException {
|
||||
int docId = scorer.docID();
|
||||
// Even if the weight is created with needsScores=false, it might
|
||||
// be costly to call score(), so we explicitly check if scores
|
||||
|
@ -349,12 +351,12 @@ public class FiltersFunctionScoreQuery extends Query {
|
|||
}
|
||||
FiltersFunctionScoreQuery other = (FiltersFunctionScoreQuery) o;
|
||||
return Objects.equals(this.subQuery, other.subQuery) && this.maxBoost == other.maxBoost &&
|
||||
Objects.equals(this.combineFunction, other.combineFunction) &&
|
||||
Objects.equals(this.combineFunction, other.combineFunction) && Objects.equals(this.minScore, other.minScore) &&
|
||||
Arrays.equals(this.filterFunctions, other.filterFunctions);
|
||||
}
|
||||
|
||||
@Override
|
||||
public int hashCode() {
|
||||
return Objects.hash(super.hashCode(), subQuery, maxBoost, combineFunction, filterFunctions);
|
||||
return Objects.hash(super.hashCode(), subQuery, maxBoost, combineFunction, minScore, filterFunctions);
|
||||
}
|
||||
}
|
||||
|
|
|
@ -40,11 +40,13 @@ public class FunctionScoreQuery extends Query {
|
|||
final ScoreFunction function;
|
||||
final float maxBoost;
|
||||
final CombineFunction combineFunction;
|
||||
private Float minScore;
|
||||
|
||||
public FunctionScoreQuery(Query subQuery, ScoreFunction function, CombineFunction combineFunction, float maxBoost) {
|
||||
public FunctionScoreQuery(Query subQuery, ScoreFunction function, Float minScore, CombineFunction combineFunction, float maxBoost) {
|
||||
this.subQuery = subQuery;
|
||||
this.function = function;
|
||||
this.combineFunction = combineFunction;
|
||||
this.minScore = minScore;
|
||||
this.maxBoost = maxBoost;
|
||||
}
|
||||
|
||||
|
@ -131,7 +133,7 @@ public class FunctionScoreQuery extends Query {
|
|||
if (function != null) {
|
||||
leafFunction = function.getLeafScoreFunction(context);
|
||||
}
|
||||
return new FunctionFactorScorer(this, subQueryScorer, leafFunction, maxBoost, combineFunction, needsScores);
|
||||
return new FunctionFactorScorer(this, subQueryScorer, leafFunction, maxBoost, combineFunction, minScore, needsScores);
|
||||
}
|
||||
|
||||
@Override
|
||||
|
@ -154,15 +156,15 @@ public class FunctionScoreQuery extends Query {
|
|||
private final LeafScoreFunction function;
|
||||
private final boolean needsScores;
|
||||
|
||||
private FunctionFactorScorer(CustomBoostFactorWeight w, Scorer scorer, LeafScoreFunction function, float maxBoost, CombineFunction scoreCombiner, boolean needsScores)
|
||||
private FunctionFactorScorer(CustomBoostFactorWeight w, Scorer scorer, LeafScoreFunction function, float maxBoost, CombineFunction scoreCombiner, Float minScore, boolean needsScores)
|
||||
throws IOException {
|
||||
super(w, scorer, maxBoost, scoreCombiner);
|
||||
super(w, scorer, maxBoost, scoreCombiner, minScore);
|
||||
this.function = function;
|
||||
this.needsScores = needsScores;
|
||||
}
|
||||
|
||||
@Override
|
||||
public float score() throws IOException {
|
||||
public float innerScore() throws IOException {
|
||||
// Even if the weight is created with needsScores=false, it might
|
||||
// be costly to call score(), so we explicitly check if scores
|
||||
// are needed
|
||||
|
@ -195,11 +197,11 @@ public class FunctionScoreQuery extends Query {
|
|||
FunctionScoreQuery other = (FunctionScoreQuery) o;
|
||||
return Objects.equals(this.subQuery, other.subQuery) && Objects.equals(this.function, other.function)
|
||||
&& Objects.equals(this.combineFunction, other.combineFunction)
|
||||
&& this.maxBoost == other.maxBoost;
|
||||
&& Objects.equals(this.minScore, other.minScore) && this.maxBoost == other.maxBoost;
|
||||
}
|
||||
|
||||
@Override
|
||||
public int hashCode() {
|
||||
return Objects.hash(super.hashCode(), subQuery.hashCode(), function, combineFunction, maxBoost);
|
||||
return Objects.hash(super.hashCode(), subQuery.hashCode(), function, combineFunction, minScore, maxBoost);
|
||||
}
|
||||
}
|
||||
|
|
|
@ -24,7 +24,6 @@ import org.apache.lucene.search.Query;
|
|||
import org.elasticsearch.common.io.stream.StreamInput;
|
||||
import org.elasticsearch.common.io.stream.StreamOutput;
|
||||
import org.elasticsearch.common.io.stream.Writeable;
|
||||
import org.elasticsearch.common.lucene.search.MinScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.CombineFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.FiltersFunctionScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.FunctionScoreQuery;
|
||||
|
@ -313,17 +312,10 @@ public class FunctionScoreQueryBuilder extends AbstractQueryBuilder<FunctionScor
|
|||
combineFunction = DEFAULT_BOOST_MODE;
|
||||
}
|
||||
}
|
||||
query = new FunctionScoreQuery(query, function, combineFunction, maxBoost);
|
||||
} else {
|
||||
// in all other cases we create a FiltersFunctionScoreQuery
|
||||
query = new FiltersFunctionScoreQuery(query, scoreMode, filterFunctions, maxBoost, boostMode == null ? DEFAULT_BOOST_MODE : boostMode);
|
||||
return new FunctionScoreQuery(query, function, minScore, combineFunction, maxBoost);
|
||||
}
|
||||
|
||||
if (minScore != null) {
|
||||
query = new MinScoreQuery(query, minScore);
|
||||
}
|
||||
|
||||
return query;
|
||||
// in all other cases we create a FiltersFunctionScoreQuery
|
||||
return new FiltersFunctionScoreQuery(query, scoreMode, filterFunctions, maxBoost, minScore, boostMode == null ? DEFAULT_BOOST_MODE : boostMode);
|
||||
}
|
||||
|
||||
/**
|
||||
|
|
|
@ -478,6 +478,11 @@ public class PercolateContext extends SearchContext {
|
|||
throw new UnsupportedOperationException();
|
||||
}
|
||||
|
||||
@Override
|
||||
public Float minimumScore() {
|
||||
return null;
|
||||
}
|
||||
|
||||
@Override
|
||||
public SearchContext sort(Sort sort) {
|
||||
this.sort = sort;
|
||||
|
|
|
@ -27,7 +27,6 @@ import org.elasticsearch.cache.recycler.PageCacheRecycler;
|
|||
import org.elasticsearch.common.Nullable;
|
||||
import org.elasticsearch.common.ParseFieldMatcher;
|
||||
import org.elasticsearch.common.lease.Releasables;
|
||||
import org.elasticsearch.common.lucene.search.MinScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.Queries;
|
||||
import org.elasticsearch.common.lucene.search.function.FunctionScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.WeightFactorFunction;
|
||||
|
@ -216,9 +215,6 @@ public class DefaultSearchContext extends SearchContext {
|
|||
parsedQuery(new ParsedQuery(filtered, parsedQuery()));
|
||||
}
|
||||
}
|
||||
if (minimumScore != null) {
|
||||
this.query = new MinScoreQuery(query, minimumScore);
|
||||
}
|
||||
try {
|
||||
this.query = searcher().rewrite(this.query);
|
||||
} catch (IOException e) {
|
||||
|
@ -495,6 +491,11 @@ public class DefaultSearchContext extends SearchContext {
|
|||
return this;
|
||||
}
|
||||
|
||||
@Override
|
||||
public Float minimumScore() {
|
||||
return this.minimumScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public SearchContext sort(Sort sort) {
|
||||
this.sort = sort;
|
||||
|
|
|
@ -317,6 +317,11 @@ public abstract class FilteredSearchContext extends SearchContext {
|
|||
return in.minimumScore(minimumScore);
|
||||
}
|
||||
|
||||
@Override
|
||||
public Float minimumScore() {
|
||||
return in.minimumScore();
|
||||
}
|
||||
|
||||
@Override
|
||||
public SearchContext sort(Sort sort) {
|
||||
return in.sort(sort);
|
||||
|
|
|
@ -230,6 +230,8 @@ public abstract class SearchContext extends DelegatingHasContextAndHeaders imple
|
|||
|
||||
public abstract SearchContext minimumScore(float minimumScore);
|
||||
|
||||
public abstract Float minimumScore();
|
||||
|
||||
public abstract SearchContext sort(Sort sort);
|
||||
|
||||
public abstract Sort sort();
|
||||
|
|
|
@ -44,6 +44,7 @@ import org.apache.lucene.search.Weight;
|
|||
import org.elasticsearch.action.search.SearchType;
|
||||
import org.elasticsearch.common.inject.Inject;
|
||||
import org.elasticsearch.common.lucene.Lucene;
|
||||
import org.elasticsearch.common.lucene.MinimumScoreCollector;
|
||||
import org.elasticsearch.common.lucene.search.FilteredCollector;
|
||||
import org.elasticsearch.search.SearchParseElement;
|
||||
import org.elasticsearch.search.SearchPhase;
|
||||
|
@ -276,6 +277,11 @@ public class QueryPhase implements SearchPhase {
|
|||
allCollectors.addAll(searchContext.queryCollectors().values());
|
||||
collector = MultiCollector.wrap(allCollectors);
|
||||
|
||||
// apply the minimum score after multi collector so we filter aggs as well
|
||||
if (searchContext.minimumScore() != null) {
|
||||
collector = new MinimumScoreCollector(collector, searchContext.minimumScore());
|
||||
}
|
||||
|
||||
if (collector.getClass() == TotalHitCountCollector.class) {
|
||||
// Optimize counts in simple cases to return in constant time
|
||||
// instead of using a collector
|
||||
|
|
|
@ -1,195 +0,0 @@
|
|||
/*
|
||||
* Licensed to Elasticsearch under one or more contributor
|
||||
* license agreements. See the NOTICE file distributed with
|
||||
* this work for additional information regarding copyright
|
||||
* ownership. Elasticsearch licenses this file to you under
|
||||
* the Apache License, Version 2.0 (the "License"); you may
|
||||
* not use this file except in compliance with the License.
|
||||
* You may obtain a copy of the License at
|
||||
*
|
||||
* http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* Unless required by applicable law or agreed to in writing,
|
||||
* software distributed under the License is distributed on an
|
||||
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
|
||||
* KIND, either express or implied. See the License for the
|
||||
* specific language governing permissions and limitations
|
||||
* under the License.
|
||||
*/
|
||||
|
||||
package org.elasticsearch.common.lucene.search;
|
||||
|
||||
import org.apache.lucene.analysis.Analyzer;
|
||||
import org.apache.lucene.analysis.MockAnalyzer;
|
||||
import org.apache.lucene.analysis.MockTokenizer;
|
||||
import org.apache.lucene.document.Document;
|
||||
import org.apache.lucene.document.Field.Store;
|
||||
import org.apache.lucene.document.TextField;
|
||||
import org.apache.lucene.index.IndexReader;
|
||||
import org.apache.lucene.index.MultiReader;
|
||||
import org.apache.lucene.index.RandomIndexWriter;
|
||||
import org.apache.lucene.index.Term;
|
||||
import org.apache.lucene.search.BooleanClause.Occur;
|
||||
import org.apache.lucene.search.*;
|
||||
import org.apache.lucene.store.Directory;
|
||||
import org.apache.lucene.util.IOUtils;
|
||||
import org.apache.lucene.util.TestUtil;
|
||||
import org.elasticsearch.test.ESTestCase;
|
||||
import org.junit.AfterClass;
|
||||
import org.junit.BeforeClass;
|
||||
|
||||
import java.io.IOException;
|
||||
|
||||
public class MinScoreQueryTests extends ESTestCase {
|
||||
|
||||
private static String[] terms;
|
||||
private static Directory dir;
|
||||
private static IndexReader r;
|
||||
private static IndexSearcher s;
|
||||
|
||||
@BeforeClass
|
||||
public static void before() throws IOException {
|
||||
dir = newDirectory();
|
||||
terms = new String[TestUtil.nextInt(random(), 2, 15)];
|
||||
for (int i = 0; i < terms.length; ++i) {
|
||||
terms[i] = TestUtil.randomSimpleString(random());
|
||||
}
|
||||
final int numDocs = TestUtil.nextInt(random(), 1, 200);
|
||||
Analyzer analyzer = new MockAnalyzer(random(), MockTokenizer.WHITESPACE, false);
|
||||
RandomIndexWriter w = new RandomIndexWriter(random(), dir, newIndexWriterConfig(analyzer));
|
||||
for (int i = 0; i < numDocs; ++i) {
|
||||
StringBuilder value = new StringBuilder();
|
||||
final int numTerms = random().nextInt(10);
|
||||
for (int j = 0; j < numTerms; ++j) {
|
||||
if (j > 0) {
|
||||
value.append(' ');
|
||||
}
|
||||
// simulate zipf distribution
|
||||
String term = terms[TestUtil.nextInt(random(), 0, TestUtil.nextInt(random(), 0, terms.length - 1))];
|
||||
value.append(term);
|
||||
}
|
||||
Document doc = new Document();
|
||||
doc.add(new TextField("field", value.toString(), Store.NO));
|
||||
w.addDocument(doc);
|
||||
}
|
||||
r = w.getReader();
|
||||
s = newSearcher(r);
|
||||
w.close();
|
||||
}
|
||||
|
||||
@AfterClass
|
||||
public static void after() throws IOException {
|
||||
IOUtils.close(r, dir);
|
||||
terms = null;
|
||||
r = null;
|
||||
s = null;
|
||||
dir = null;
|
||||
}
|
||||
|
||||
private static Term randomTerm() {
|
||||
return new Term("field", terms[random().nextInt(terms.length)]);
|
||||
}
|
||||
|
||||
public void testEquals() throws IOException {
|
||||
QueryUtils.checkEqual(new MinScoreQuery(new MatchAllDocsQuery(), 0.5f), new MinScoreQuery(new MatchAllDocsQuery(), 0.5f));
|
||||
QueryUtils.checkUnequal(new MinScoreQuery(new MatchAllDocsQuery(), 0.5f), new MinScoreQuery(new MatchAllDocsQuery(), 0.3f));
|
||||
QueryUtils.checkUnequal(new MinScoreQuery(new MatchAllDocsQuery(), 0.5f), new MinScoreQuery(new MatchNoDocsQuery(), 0.5f));
|
||||
|
||||
IndexSearcher s1 = new IndexSearcher(new MultiReader());
|
||||
IndexSearcher s2 = new IndexSearcher(new MultiReader());
|
||||
QueryUtils.checkEqual(new MinScoreQuery(new MatchAllDocsQuery(), 0.5f, s1), new MinScoreQuery(new MatchAllDocsQuery(), 0.5f, s1));
|
||||
QueryUtils.checkUnequal(new MinScoreQuery(new MatchAllDocsQuery(), 0.5f, s1), new MinScoreQuery(new MatchAllDocsQuery(), 0.5f, s2));
|
||||
}
|
||||
|
||||
/** pick a min score which is in the range of scores produced by the query */
|
||||
private float randomMinScore(Query query) throws IOException {
|
||||
TopDocs topDocs = s.search(query, 2);
|
||||
float base = 0;
|
||||
switch (topDocs.totalHits) {
|
||||
case 0:
|
||||
break;
|
||||
case 1:
|
||||
base = topDocs.scoreDocs[0].score;
|
||||
break;
|
||||
default:
|
||||
base = (topDocs.scoreDocs[0].score + topDocs.scoreDocs[1].score) / 2;
|
||||
break;
|
||||
}
|
||||
float delta = random().nextFloat() - 0.5f;
|
||||
return base + delta;
|
||||
}
|
||||
|
||||
private void assertMinScoreEquivalence(Query query, Query minScoreQuery, float minScore) throws IOException {
|
||||
final TopDocs topDocs = s.search(query, s.getIndexReader().maxDoc());
|
||||
final TopDocs minScoreTopDocs = s.search(minScoreQuery, s.getIndexReader().maxDoc());
|
||||
|
||||
int j = 0;
|
||||
for (int i = 0; i < topDocs.totalHits; ++i) {
|
||||
if (topDocs.scoreDocs[i].score >= minScore) {
|
||||
assertEquals(topDocs.scoreDocs[i].doc, minScoreTopDocs.scoreDocs[j].doc);
|
||||
assertEquals(topDocs.scoreDocs[i].score, minScoreTopDocs.scoreDocs[j].score, 1e-5f);
|
||||
j++;
|
||||
}
|
||||
}
|
||||
assertEquals(minScoreTopDocs.totalHits, j);
|
||||
}
|
||||
|
||||
public void testBasics() throws Exception {
|
||||
final int iters = 5;
|
||||
for (int iter = 0; iter < iters; ++iter) {
|
||||
Term term = randomTerm();
|
||||
Query query = new TermQuery(term);
|
||||
float minScore = randomMinScore(query);
|
||||
Query minScoreQuery = new MinScoreQuery(query, minScore);
|
||||
assertMinScoreEquivalence(query, minScoreQuery, minScore);
|
||||
}
|
||||
}
|
||||
|
||||
public void testFilteredApproxQuery() throws Exception {
|
||||
// same as testBasics but with a query that exposes approximations
|
||||
final int iters = 5;
|
||||
for (int iter = 0; iter < iters; ++iter) {
|
||||
Term term = randomTerm();
|
||||
Query query = new TermQuery(term);
|
||||
float minScore = randomMinScore(query);
|
||||
Query minScoreQuery = new MinScoreQuery(new RandomApproximationQuery(query, random()), minScore);
|
||||
assertMinScoreEquivalence(query, minScoreQuery, minScore);
|
||||
}
|
||||
}
|
||||
|
||||
public void testNestedInConjunction() throws Exception {
|
||||
// To test scorers as well, not only bulk scorers
|
||||
Term t1 = randomTerm();
|
||||
Term t2 = randomTerm();
|
||||
Query tq1 = new TermQuery(t1);
|
||||
Query tq2 = new TermQuery(t2);
|
||||
float minScore = randomMinScore(tq1);
|
||||
BooleanQuery bq1 = new BooleanQuery.Builder()
|
||||
.add(new MinScoreQuery(tq1, minScore), Occur.MUST)
|
||||
.add(tq2, Occur.FILTER)
|
||||
.build();
|
||||
BooleanQuery bq2 = new BooleanQuery.Builder()
|
||||
.add(tq1, Occur.MUST)
|
||||
.add(tq2, Occur.FILTER)
|
||||
.build();
|
||||
assertMinScoreEquivalence(bq2, bq1, minScore);
|
||||
}
|
||||
|
||||
public void testNestedInConjunctionWithApprox() throws Exception {
|
||||
// same, but with approximations
|
||||
Term t1 = randomTerm();
|
||||
Term t2 = randomTerm();
|
||||
Query tq1 = new TermQuery(t1);
|
||||
Query tq2 = new TermQuery(t2);
|
||||
float minScore = randomMinScore(tq1);
|
||||
BooleanQuery bq1 = new BooleanQuery.Builder()
|
||||
.add(new MinScoreQuery(new RandomApproximationQuery(tq1, random()), minScore), Occur.MUST)
|
||||
.add(tq2, Occur.FILTER)
|
||||
.build();
|
||||
BooleanQuery bq2 = new BooleanQuery.Builder()
|
||||
.add(tq1, Occur.MUST)
|
||||
.add(tq2, Occur.FILTER)
|
||||
.build();
|
||||
assertMinScoreEquivalence(bq2, bq1, minScore);
|
||||
}
|
||||
}
|
|
@ -26,8 +26,11 @@ import org.apache.lucene.search.MatchAllDocsQuery;
|
|||
import org.apache.lucene.search.Query;
|
||||
import org.apache.lucene.search.TermQuery;
|
||||
import org.elasticsearch.common.ParsingException;
|
||||
import org.elasticsearch.common.lucene.search.MinScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.*;
|
||||
import org.elasticsearch.common.lucene.search.function.CombineFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.FieldValueFactorFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.FiltersFunctionScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.FunctionScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.WeightFactorFunction;
|
||||
import org.elasticsearch.common.xcontent.XContentType;
|
||||
import org.elasticsearch.index.query.AbstractQueryBuilder;
|
||||
import org.elasticsearch.index.query.AbstractQueryTestCase;
|
||||
|
@ -179,12 +182,6 @@ public class FunctionScoreQueryBuilderTests extends AbstractQueryTestCase<Functi
|
|||
|
||||
@Override
|
||||
protected void doAssertLuceneQuery(FunctionScoreQueryBuilder queryBuilder, Query query, QueryShardContext context) throws IOException {
|
||||
if (queryBuilder.getMinScore() != null) {
|
||||
assertThat(query, instanceOf(MinScoreQuery.class));
|
||||
MinScoreQuery msq = (MinScoreQuery) query;
|
||||
assertEquals(queryBuilder.getMinScore(), msq.getMinScore(), 0f);
|
||||
query = msq.getQuery();
|
||||
}
|
||||
assertThat(query, either(instanceOf(FunctionScoreQuery.class)).or(instanceOf(FiltersFunctionScoreQuery.class)));
|
||||
}
|
||||
|
||||
|
|
|
@ -24,6 +24,7 @@ import org.apache.lucene.document.Document;
|
|||
import org.apache.lucene.document.Field;
|
||||
import org.apache.lucene.document.TextField;
|
||||
import org.apache.lucene.index.DirectoryReader;
|
||||
import org.apache.lucene.index.IndexReader;
|
||||
import org.apache.lucene.index.IndexWriter;
|
||||
import org.apache.lucene.index.LeafReaderContext;
|
||||
import org.apache.lucene.index.SortedNumericDocValues;
|
||||
|
@ -31,7 +32,6 @@ import org.apache.lucene.index.Term;
|
|||
import org.apache.lucene.search.Explanation;
|
||||
import org.apache.lucene.search.IndexSearcher;
|
||||
import org.apache.lucene.search.MatchAllDocsQuery;
|
||||
import org.apache.lucene.search.Query;
|
||||
import org.apache.lucene.search.TermQuery;
|
||||
import org.apache.lucene.search.TopDocs;
|
||||
import org.apache.lucene.search.Weight;
|
||||
|
@ -39,8 +39,14 @@ import org.apache.lucene.store.Directory;
|
|||
import org.apache.lucene.util.Accountable;
|
||||
import org.apache.lucene.util.BytesRef;
|
||||
import org.elasticsearch.common.Nullable;
|
||||
import org.elasticsearch.common.lucene.search.MinScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.*;
|
||||
import org.elasticsearch.common.lucene.search.function.CombineFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.FieldValueFactorFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.FiltersFunctionScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.FunctionScoreQuery;
|
||||
import org.elasticsearch.common.lucene.search.function.LeafScoreFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.RandomScoreFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.ScoreFunction;
|
||||
import org.elasticsearch.common.lucene.search.function.WeightFactorFunction;
|
||||
import org.elasticsearch.index.Index;
|
||||
import org.elasticsearch.index.fielddata.AtomicFieldData;
|
||||
import org.elasticsearch.index.fielddata.AtomicNumericFieldData;
|
||||
|
@ -312,7 +318,7 @@ public class FunctionScoreTests extends ESTestCase {
|
|||
}
|
||||
|
||||
public Explanation getFunctionScoreExplanation(IndexSearcher searcher, ScoreFunction scoreFunction) throws IOException {
|
||||
Query functionScoreQuery = new MinScoreQuery(new FunctionScoreQuery(new TermQuery(TERM), scoreFunction, CombineFunction.AVG, 100), 0f);
|
||||
FunctionScoreQuery functionScoreQuery = new FunctionScoreQuery(new TermQuery(TERM), scoreFunction, 0.0f, CombineFunction.AVG, 100);
|
||||
Weight weight = searcher.createNormalizedWeight(functionScoreQuery, true);
|
||||
Explanation explanation = weight.explain(searcher.getIndexReader().leaves().get(0), 0);
|
||||
return explanation.getDetails()[1];
|
||||
|
@ -376,22 +382,22 @@ public class FunctionScoreTests extends ESTestCase {
|
|||
}
|
||||
|
||||
public Explanation getFiltersFunctionScoreExplanation(IndexSearcher searcher, ScoreFunction... scoreFunctions) throws IOException {
|
||||
Query filtersFunctionScoreQuery = getFiltersFunctionScoreQuery(FiltersFunctionScoreQuery.ScoreMode.AVG, CombineFunction.AVG, scoreFunctions);
|
||||
FiltersFunctionScoreQuery filtersFunctionScoreQuery = getFiltersFunctionScoreQuery(FiltersFunctionScoreQuery.ScoreMode.AVG, CombineFunction.AVG, scoreFunctions);
|
||||
return getExplanation(searcher, filtersFunctionScoreQuery).getDetails()[1];
|
||||
}
|
||||
|
||||
protected Explanation getExplanation(IndexSearcher searcher, Query filtersFunctionScoreQuery) throws IOException {
|
||||
protected Explanation getExplanation(IndexSearcher searcher, FiltersFunctionScoreQuery filtersFunctionScoreQuery) throws IOException {
|
||||
Weight weight = searcher.createNormalizedWeight(filtersFunctionScoreQuery, true);
|
||||
return weight.explain(searcher.getIndexReader().leaves().get(0), 0);
|
||||
}
|
||||
|
||||
public Query getFiltersFunctionScoreQuery(FiltersFunctionScoreQuery.ScoreMode scoreMode, CombineFunction combineFunction, ScoreFunction... scoreFunctions) {
|
||||
public FiltersFunctionScoreQuery getFiltersFunctionScoreQuery(FiltersFunctionScoreQuery.ScoreMode scoreMode, CombineFunction combineFunction, ScoreFunction... scoreFunctions) {
|
||||
FiltersFunctionScoreQuery.FilterFunction[] filterFunctions = new FiltersFunctionScoreQuery.FilterFunction[scoreFunctions.length];
|
||||
for (int i = 0; i < scoreFunctions.length; i++) {
|
||||
filterFunctions[i] = new FiltersFunctionScoreQuery.FilterFunction(
|
||||
new TermQuery(TERM), scoreFunctions[i]);
|
||||
}
|
||||
return new MinScoreQuery(new FiltersFunctionScoreQuery(new TermQuery(TERM), scoreMode, filterFunctions, Float.MAX_VALUE, combineFunction), Float.MAX_VALUE * -1);
|
||||
return new FiltersFunctionScoreQuery(new TermQuery(TERM), scoreMode, filterFunctions, Float.MAX_VALUE, Float.MAX_VALUE * -1, combineFunction);
|
||||
}
|
||||
|
||||
public void checkFiltersFunctionScoreExplanation(Explanation randomExplanation, String functionExpl, int whichFunction) {
|
||||
|
@ -465,7 +471,7 @@ public class FunctionScoreTests extends ESTestCase {
|
|||
weightFunctionStubs[i] = new WeightFactorFunction(weights[i], scoreFunctionStubs[i]);
|
||||
}
|
||||
|
||||
Query filtersFunctionScoreQueryWithWeights = getFiltersFunctionScoreQuery(
|
||||
FiltersFunctionScoreQuery filtersFunctionScoreQueryWithWeights = getFiltersFunctionScoreQuery(
|
||||
FiltersFunctionScoreQuery.ScoreMode.MULTIPLY
|
||||
, CombineFunction.REPLACE
|
||||
, weightFunctionStubs
|
||||
|
@ -548,10 +554,10 @@ public class FunctionScoreTests extends ESTestCase {
|
|||
assertThat(explainedScore / scoreWithWeight, is(1f));
|
||||
}
|
||||
|
||||
public void checkWeightOnlyCreatesBoostFunction() throws IOException {
|
||||
Query filtersFunctionScoreQueryWithWeights = new MinScoreQuery(new FunctionScoreQuery(new MatchAllDocsQuery(), new WeightFactorFunction(2), CombineFunction.MULTIPLY, 100), 0f);
|
||||
public void testWeightOnlyCreatesBoostFunction() throws IOException {
|
||||
FunctionScoreQuery filtersFunctionScoreQueryWithWeights = new FunctionScoreQuery(new MatchAllDocsQuery(), new WeightFactorFunction(2), 0.0f, CombineFunction.MULTIPLY, 100);
|
||||
TopDocs topDocsWithWeights = searcher.search(filtersFunctionScoreQueryWithWeights, 1);
|
||||
float score = topDocsWithWeights.scoreDocs[0].score;
|
||||
assertThat(score, equalTo(2.0f));
|
||||
}
|
||||
}
|
||||
}
|
|
@ -143,6 +143,10 @@ public class QueryPhaseTests extends ESTestCase {
|
|||
}
|
||||
|
||||
public void testMinScoreDisablesCountOptimization() throws Exception {
|
||||
TestSearchContext context = new TestSearchContext();
|
||||
context.parsedQuery(new ParsedQuery(new MatchAllDocsQuery()));
|
||||
context.setSize(0);
|
||||
|
||||
final AtomicBoolean collected = new AtomicBoolean();
|
||||
IndexSearcher contextSearcher = new IndexSearcher(new MultiReader()) {
|
||||
protected void search(List<LeafReaderContext> leaves, Weight weight, Collector collector) throws IOException {
|
||||
|
@ -151,19 +155,11 @@ public class QueryPhaseTests extends ESTestCase {
|
|||
}
|
||||
};
|
||||
|
||||
TestSearchContext context = new TestSearchContext();
|
||||
context.parsedQuery(new ParsedQuery(new MatchAllDocsQuery()));
|
||||
context.setSize(0);
|
||||
context.preProcess();
|
||||
QueryPhase.execute(context, contextSearcher);
|
||||
assertEquals(0, context.queryResult().topDocs().totalHits);
|
||||
assertFalse(collected.get());
|
||||
|
||||
context = new TestSearchContext();
|
||||
context.parsedQuery(new ParsedQuery(new MatchAllDocsQuery()));
|
||||
context.setSize(0);
|
||||
context.minimumScore(1);
|
||||
context.preProcess();
|
||||
QueryPhase.execute(context, contextSearcher);
|
||||
assertEquals(0, context.queryResult().topDocs().totalHits);
|
||||
assertTrue(collected.get());
|
||||
|
|
|
@ -31,7 +31,6 @@ import org.elasticsearch.common.HasContextAndHeaders;
|
|||
import org.elasticsearch.common.HasHeaders;
|
||||
import org.elasticsearch.common.ParseFieldMatcher;
|
||||
import org.elasticsearch.common.collect.ImmutableOpenMap;
|
||||
import org.elasticsearch.common.lucene.search.MinScoreQuery;
|
||||
import org.elasticsearch.common.util.BigArrays;
|
||||
import org.elasticsearch.index.IndexService;
|
||||
import org.elasticsearch.index.analysis.AnalysisService;
|
||||
|
@ -130,9 +129,6 @@ public class TestSearchContext extends SearchContext {
|
|||
|
||||
@Override
|
||||
public void preProcess() {
|
||||
if (minScore != null) {
|
||||
this.query = new MinScoreQuery(query, minScore);
|
||||
}
|
||||
}
|
||||
|
||||
@Override
|
||||
|
@ -379,6 +375,11 @@ public class TestSearchContext extends SearchContext {
|
|||
return this;
|
||||
}
|
||||
|
||||
@Override
|
||||
public Float minimumScore() {
|
||||
return minScore;
|
||||
}
|
||||
|
||||
@Override
|
||||
public SearchContext sort(Sort sort) {
|
||||
return null;
|
||||
|
|
Loading…
Reference in New Issue