You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by si...@apache.org on 2010/08/31 21:50:13 UTC
svn commit: r991310 - in /lucene/dev/trunk/lucene: ./
src/java/org/apache/lucene/search/
src/java/org/apache/lucene/search/function/
src/java/org/apache/lucene/search/spans/ src/test/org/apache/lucene/search/
Author: simonw
Date: Tue Aug 31 19:50:12 2010
New Revision: 991310
URL: http://svn.apache.org/viewvc?rev=991310&view=rev
Log:
LUCENE-2590: Enable access to the freq information in a Query's sub-scorers
Added:
lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java (with props)
Modified:
lucene/dev/trunk/lucene/CHANGES.txt
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer2.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ConstantScoreQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ExactPhraseScorer.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/FilteredQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MatchAllDocsQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MultiPhraseQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseScorer.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ScoreCachingWrappingScorer.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/Scorer.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/TermScorer.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/CustomScoreQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/ValueSourceQuery.java
lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/spans/SpanScorer.java
lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestBooleanScorer.java
Modified: lucene/dev/trunk/lucene/CHANGES.txt
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/CHANGES.txt?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/CHANGES.txt (original)
+++ lucene/dev/trunk/lucene/CHANGES.txt Tue Aug 31 19:50:12 2010
@@ -208,6 +208,11 @@ New features
* LUCENE-2559: Added SegmentReader.reopen methods (John Wang via Mike
McCandless)
+* LUCENE-2590: Added Scorer.visitSubScorers, and Scorer.freq. Along
+ with a custom Collector these experimental methods make it possible
+ to gather the hit-count per sub-clause and per document while a
+ search is running. (Simon Willnauer, Mike McCandless)
+
Optimizations
* LUCENE-2410: ~20% speedup on exact (slop=0) PhraseQuery matching.
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanQuery.java Tue Aug 31 19:50:12 2010
@@ -320,7 +320,7 @@ public class BooleanQuery extends Query
// Check if we can return a BooleanScorer
if (!scoreDocsInOrder && topScorer && required.size() == 0 && prohibited.size() < 32) {
- return new BooleanScorer(similarity, minNrShouldMatch, optional, prohibited, maxCoord);
+ return new BooleanScorer(this, similarity, minNrShouldMatch, optional, prohibited, maxCoord);
}
if (required.size() == 0 && optional.size() == 0) {
@@ -334,7 +334,7 @@ public class BooleanQuery extends Query
}
// Return a BooleanScorer2
- return new BooleanScorer2(similarity, minNrShouldMatch, required, prohibited, optional, maxCoord);
+ return new BooleanScorer2(this, similarity, minNrShouldMatch, required, prohibited, optional, maxCoord);
}
@Override
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer.java Tue Aug 31 19:50:12 2010
@@ -21,6 +21,7 @@ import java.io.IOException;
import java.util.List;
import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.search.BooleanClause.Occur;
/* Description from Doug Cutting (excerpted from
* LUCENE-1483):
@@ -115,6 +116,7 @@ final class BooleanScorer extends Scorer
float score;
int doc = NO_MORE_DOCS;
+ int freq;
public BucketScorer() { super(null); }
@@ -125,6 +127,9 @@ final class BooleanScorer extends Scorer
public int docID() { return doc; }
@Override
+ public float freq() { return freq; }
+
+ @Override
public int nextDoc() throws IOException { return NO_MORE_DOCS; }
@Override
@@ -159,7 +164,8 @@ final class BooleanScorer extends Scorer
static final class SubScorer {
public Scorer scorer;
- public boolean required = false;
+ // TODO: re-enable this if BQ ever sends us required clauses
+ //public boolean required = false;
public boolean prohibited = false;
public Collector collector;
public SubScorer next;
@@ -167,8 +173,12 @@ final class BooleanScorer extends Scorer
public SubScorer(Scorer scorer, boolean required, boolean prohibited,
Collector collector, SubScorer next)
throws IOException {
+ if (required) {
+ throw new IllegalArgumentException("this scorer cannot handle required=true");
+ }
this.scorer = scorer;
- this.required = required;
+ // TODO: re-enable this if BQ ever sends us required clauses
+ //this.required = required;
this.prohibited = prohibited;
this.collector = collector;
this.next = next;
@@ -178,17 +188,18 @@ final class BooleanScorer extends Scorer
private SubScorer scorers = null;
private BucketTable bucketTable = new BucketTable();
private final float[] coordFactors;
- private int requiredMask = 0;
+ // TODO: re-enable this if BQ ever sends us required clauses
+ //private int requiredMask = 0;
private int prohibitedMask = 0;
private int nextMask = 1;
private final int minNrShouldMatch;
private int end;
private Bucket current;
private int doc = -1;
-
- BooleanScorer(Similarity similarity, int minNrShouldMatch,
+
+ BooleanScorer(Weight weight, Similarity similarity, int minNrShouldMatch,
List<Scorer> optionalScorers, List<Scorer> prohibitedScorers, int maxCoord) throws IOException {
- super(similarity);
+ super(similarity, weight);
this.minNrShouldMatch = minNrShouldMatch;
if (optionalScorers != null && optionalScorers.size() > 0) {
@@ -231,8 +242,11 @@ final class BooleanScorer extends Scorer
while (current != null) { // more queued
// check prohibited & required
- if ((current.bits & prohibitedMask) == 0 &&
- (current.bits & requiredMask) == requiredMask) {
+ if ((current.bits & prohibitedMask) == 0) {
+
+ // TODO: re-enable this if BQ ever sends us required
+ // clauses
+ //&& (current.bits & requiredMask) == requiredMask) {
if (current.doc >= max){
tmp = current;
@@ -245,6 +259,7 @@ final class BooleanScorer extends Scorer
if (current.coord >= minNrShouldMatch) {
bs.score = current.score * coordFactors[current.coord];
bs.doc = current.doc;
+ bs.freq = current.coord;
collector.collect(current.doc);
}
}
@@ -294,8 +309,9 @@ final class BooleanScorer extends Scorer
// check prohibited & required, and minNrShouldMatch
if ((current.bits & prohibitedMask) == 0 &&
- (current.bits & requiredMask) == requiredMask &&
current.coord >= minNrShouldMatch) {
+ // TODO: re-enable this if BQ ever sends us required clauses
+ // (current.bits & requiredMask) == requiredMask &&
return doc = current.doc;
}
}
@@ -339,5 +355,28 @@ final class BooleanScorer extends Scorer
buffer.append(")");
return buffer.toString();
}
+
+ @Override
+ protected void visitSubScorers(Query parent, Occur relationship, ScorerVisitor<Query, Query, Scorer> visitor) {
+ super.visitSubScorers(parent, relationship, visitor);
+ final Query q = weight.getQuery();
+ SubScorer sub = scorers;
+ while(sub != null) {
+ // TODO: re-enable this if BQ ever sends us required
+ //clauses
+ //if (sub.required) {
+ //relationship = Occur.MUST;
+ if (!sub.prohibited) {
+ relationship = Occur.SHOULD;
+ } else {
+ // TODO: maybe it's pointless to do this, but, it is
+ // possible the doc may still be collected, eg foo
+ // OR (bar -fee)
+ relationship = Occur.MUST_NOT;
+ }
+ sub.scorer.visitSubScorers(q, relationship, visitor);
+ sub = sub.next;
+ }
+ }
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer2.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer2.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer2.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/BooleanScorer2.java Tue Aug 31 19:50:12 2010
@@ -21,6 +21,8 @@ import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
+import org.apache.lucene.search.BooleanClause.Occur;
+
/* See the description in BooleanScorer.java, comparing
* BooleanScorer & BooleanScorer2 */
@@ -59,7 +61,7 @@ class BooleanScorer2 extends Scorer {
/** The number of optionalScorers that need to match (if there are any) */
private final int minNrShouldMatch;
-
+
private int doc = -1;
/**
@@ -80,9 +82,9 @@ class BooleanScorer2 extends Scorer {
* @param optional
* the list of optional scorers.
*/
- public BooleanScorer2(Similarity similarity, int minNrShouldMatch,
+ public BooleanScorer2(Weight weight, Similarity similarity, int minNrShouldMatch,
List<Scorer> required, List<Scorer> prohibited, List<Scorer> optional, int maxCoord) throws IOException {
- super(similarity);
+ super(similarity, weight);
if (minNrShouldMatch < 0) {
throw new IllegalArgumentException("Minimum number of optional scorers should not be negative");
}
@@ -302,9 +304,27 @@ class BooleanScorer2 extends Scorer {
}
@Override
+ public float freq() {
+ return coordinator.nrMatchers;
+ }
+
+ @Override
public int advance(int target) throws IOException {
return doc = countingSumScorer.advance(target);
}
-}
-
+ @Override
+ protected void visitSubScorers(Query parent, Occur relationship, ScorerVisitor<Query, Query, Scorer> visitor) {
+ super.visitSubScorers(parent, relationship, visitor);
+ final Query q = weight.getQuery();
+ for (Scorer s : optionalScorers) {
+ s.visitSubScorers(q, Occur.SHOULD, visitor);
+ }
+ for (Scorer s : prohibitedScorers) {
+ s.visitSubScorers(q, Occur.MUST_NOT, visitor);
+ }
+ for (Scorer s : requiredScorers) {
+ s.visitSubScorers(q, Occur.MUST, visitor);
+ }
+ }
+}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ConstantScoreQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ConstantScoreQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ConstantScoreQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ConstantScoreQuery.java Tue Aug 31 19:50:12 2010
@@ -123,7 +123,7 @@ public class ConstantScoreQuery extends
int doc = -1;
public ConstantScorer(Similarity similarity, IndexReader reader, Weight w) throws IOException {
- super(similarity);
+ super(similarity,w);
theScore = w.getValue();
DocIdSet docIdSet = filter.getDocIdSet(reader);
if (docIdSet == null) {
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ExactPhraseScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ExactPhraseScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ExactPhraseScorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ExactPhraseScorer.java Tue Aug 31 19:50:12 2010
@@ -21,9 +21,9 @@ import java.io.IOException;
import java.util.Arrays;
import org.apache.lucene.index.*;
+import org.apache.lucene.search.BooleanClause.Occur;
final class ExactPhraseScorer extends Scorer {
- private final Weight weight;
private final byte[] norms;
private final float value;
@@ -63,8 +63,7 @@ final class ExactPhraseScorer extends Sc
ExactPhraseScorer(Weight weight, PhraseQuery.PostingsAndFreq[] postings,
Similarity similarity, byte[] norms) throws IOException {
- super(similarity);
- this.weight = weight;
+ super(similarity, weight);
this.norms = norms;
this.value = weight.getValue();
@@ -193,8 +192,8 @@ final class ExactPhraseScorer extends Sc
return "ExactPhraseScorer(" + weight + ")";
}
- // used by MultiPhraseQuery
- float currentFreq() {
+ @Override
+ public float freq() {
return freq;
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/FilteredQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/FilteredQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/FilteredQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/FilteredQuery.java Tue Aug 31 19:50:12 2010
@@ -126,7 +126,7 @@ extends Query {
return null;
}
- return new Scorer(similarity) {
+ return new Scorer(similarity, this) {
private int doc = -1;
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MatchAllDocsQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MatchAllDocsQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MatchAllDocsQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MatchAllDocsQuery.java Tue Aug 31 19:50:12 2010
@@ -54,7 +54,7 @@ public class MatchAllDocsQuery extends Q
MatchAllScorer(IndexReader reader, Similarity similarity, Weight w,
byte[] norms) throws IOException {
- super(similarity);
+ super(similarity,w);
delDocs = MultiFields.getDeletedDocs(reader);
score = w.getValue();
maxDoc = reader.maxDoc();
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MultiPhraseQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MultiPhraseQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MultiPhraseQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/MultiPhraseQuery.java Tue Aug 31 19:50:12 2010
@@ -271,11 +271,7 @@ public class MultiPhraseQuery extends Qu
int d = scorer.advance(doc);
float phraseFreq;
if (d == doc) {
- if (slop == 0) {
- phraseFreq = ((ExactPhraseScorer) scorer).currentFreq();
- } else {
- phraseFreq = ((SloppyPhraseScorer) scorer).currentFreq();
- }
+ phraseFreq = scorer.freq();
} else {
phraseFreq = 0.0f;
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseQuery.java Tue Aug 31 19:50:12 2010
@@ -275,11 +275,7 @@ public class PhraseQuery extends Query {
int d = scorer.advance(doc);
float phraseFreq;
if (d == doc) {
- if (slop == 0) {
- phraseFreq = ((ExactPhraseScorer) scorer).currentFreq();
- } else {
- phraseFreq = ((SloppyPhraseScorer) scorer).currentFreq();
- }
+ phraseFreq = scorer.freq();
} else {
phraseFreq = 0.0f;
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseScorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/PhraseScorer.java Tue Aug 31 19:50:12 2010
@@ -19,6 +19,8 @@ package org.apache.lucene.search;
import java.io.IOException;
+import org.apache.lucene.search.BooleanClause.Occur;
+
/** Expert: Scoring functionality for phrase queries.
* <br>A document is considered matching if it contains the phrase-query terms
* at "valid" positions. What "valid positions" are
@@ -30,7 +32,6 @@ import java.io.IOException;
* means a match.
*/
abstract class PhraseScorer extends Scorer {
- private Weight weight;
protected byte[] norms;
protected float value;
@@ -43,9 +44,8 @@ abstract class PhraseScorer extends Scor
PhraseScorer(Weight weight, PhraseQuery.PostingsAndFreq[] postings,
Similarity similarity, byte[] norms) {
- super(similarity);
+ super(similarity, weight);
this.norms = norms;
- this.weight = weight;
this.value = weight.getValue();
// convert tps to a list of phrase positions.
@@ -129,8 +129,11 @@ abstract class PhraseScorer extends Scor
/**
* phrase frequency in current doc as computed by phraseFreq().
*/
- public final float currentFreq() { return freq; }
-
+ @Override
+ public final float freq() {
+ return freq;
+ }
+
/**
* For a document containing all the phrase query terms, compute the
* frequency of the phrase in that document.
@@ -179,5 +182,5 @@ abstract class PhraseScorer extends Scor
@Override
public String toString() { return "scorer(" + weight + ")"; }
-
+
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ScoreCachingWrappingScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ScoreCachingWrappingScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ScoreCachingWrappingScorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/ScoreCachingWrappingScorer.java Tue Aug 31 19:50:12 2010
@@ -32,7 +32,7 @@ import java.io.IOException;
*/
public class ScoreCachingWrappingScorer extends Scorer {
- private Scorer scorer;
+ private final Scorer scorer;
private int curDoc = -1;
private float curScore;
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/Scorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/Scorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/Scorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/Scorer.java Tue Aug 31 19:50:12 2010
@@ -19,6 +19,8 @@ package org.apache.lucene.search;
import java.io.IOException;
+import org.apache.lucene.search.BooleanClause.Occur;
+
/**
* Expert: Common scoring functionality for different types of queries.
*
@@ -39,12 +41,23 @@ import java.io.IOException;
*/
public abstract class Scorer extends DocIdSetIterator {
private final Similarity similarity;
+ protected final Weight weight;
/** Constructs a Scorer.
* @param similarity The <code>Similarity</code> implementation used by this scorer.
*/
protected Scorer(Similarity similarity) {
+ this(similarity, null);
+ }
+
+ /**
+ * Constructs a Scorer
+ * @param similarity The <code>Similarity</code> implementation used by this scorer.
+ * @param weight The scorers <code>Weight</code>
+ */
+ protected Scorer(Similarity similarity, Weight weight) {
this.similarity = similarity;
+ this.weight = weight;
}
/** Returns the Similarity implementation used by this scorer. */
@@ -94,4 +107,92 @@ public abstract class Scorer extends Doc
*/
public abstract float score() throws IOException;
+ /** Returns number of matches for the current document.
+ * This returns a float (not int) because
+ * SloppyPhraseScorer discounts its freq according to how
+ * "sloppy" the match was.
+ *
+ * @lucene.experimental */
+ public float freq() throws IOException {
+ throw new UnsupportedOperationException(this + " does not implement freq()");
+ }
+
+ /**
+ * A callback to gather information from a scorer and its sub-scorers. Each
+ * the top-level scorer as well as each of its sub-scorers are passed to
+ * either one of the visit methods depending on their boolean relationship in
+ * the query.
+ * @lucene.experimental
+ */
+ public static abstract class ScorerVisitor<P extends Query, C extends Query, S extends Scorer> {
+ /**
+ * Invoked for all optional scorer
+ *
+ * @param parent the parent query of the child query or <code>null</code> if the child is a top-level query
+ * @param child the query of the currently visited scorer
+ * @param scorer the current scorer
+ */
+ public void visitOptional(P parent, C child, S scorer) {}
+
+ /**
+ * Invoked for all required scorer
+ *
+ * @param parent the parent query of the child query or <code>null</code> if the child is a top-level query
+ * @param child the query of the currently visited scorer
+ * @param scorer the current scorer
+ */
+ public void visitRequired(P parent, C child, S scorer) {}
+
+ /**
+ * Invoked for all prohibited scorer
+ *
+ * @param parent the parent query of the child query or <code>null</code> if the child is a top-level query
+ * @param child the query of the currently visited scorer
+ * @param scorer the current scorer
+ */
+ public void visitProhibited(P parent, C child, S scorer) {}
+ }
+
+ /**
+ * Expert: call this to gather details for all sub-scorers for this query.
+ * This can be used, in conjunction with a custom {@link Collector} to gather
+ * details about how each sub-query matched the current hit.
+ *
+ * @param visitor a callback executed for each sub-scorer
+ * @lucene.experimental
+ */
+ public void visitScorers(ScorerVisitor<Query, Query, Scorer> visitor) {
+ visitSubScorers(null, Occur.MUST/*must id default*/, visitor);
+ }
+
+ /**
+ * {@link Scorer} subclasses should implement this method if the subclass
+ * itself contains multiple scorers to support gathering details for
+ * sub-scorers via {@link ScorerVisitor}
+ * <p>
+ * Note: this method will throw {@link UnsupportedOperationException} if no
+ * associated {@link Weight} instance is provided to
+ * {@link #Scorer(Similarity, Weight)}
+ * </p>
+ *
+ * @lucene.experimental
+ */
+ protected void visitSubScorers(Query parent, Occur relationship,
+ ScorerVisitor<Query, Query, Scorer> visitor) {
+ if (weight == null)
+ throw new UnsupportedOperationException();
+
+ final Query q = weight.getQuery();
+ switch (relationship) {
+ case MUST:
+ visitor.visitRequired(parent, q, this);
+ break;
+ case MUST_NOT:
+ visitor.visitProhibited(parent, q, this);
+ break;
+ case SHOULD:
+ visitor.visitOptional(parent, q, this);
+ break;
+ }
+ }
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/TermScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/TermScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/TermScorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/TermScorer.java Tue Aug 31 19:50:12 2010
@@ -20,11 +20,11 @@ package org.apache.lucene.search;
import java.io.IOException;
import org.apache.lucene.index.DocsEnum;
+import org.apache.lucene.search.BooleanClause.Occur;
/** Expert: A <code>Scorer</code> for documents matching a <code>Term</code>.
*/
final class TermScorer extends Scorer {
- private Weight weight;
private DocsEnum docsEnum;
private byte[] norms;
private float weightValue;
@@ -54,9 +54,8 @@ final class TermScorer extends Scorer {
* The field norms of the document fields for the <code>Term</code>.
*/
TermScorer(Weight weight, DocsEnum td, Similarity similarity, byte[] norms) {
- super(similarity);
+ super(similarity, weight);
- this.weight = weight;
this.docsEnum = td;
this.norms = norms;
this.weightValue = weight.getValue();
@@ -103,6 +102,11 @@ final class TermScorer extends Scorer {
return doc;
}
+ @Override
+ public float freq() {
+ return freq;
+ }
+
/**
* Advances to the next document matching the query. <br>
* The iterator over the matching documents is buffered using
@@ -172,4 +176,5 @@ final class TermScorer extends Scorer {
/** Returns a string representation of this <code>TermScorer</code>. */
@Override
public String toString() { return "scorer(" + weight + ")"; }
+
}
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/CustomScoreQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/CustomScoreQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/CustomScoreQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/CustomScoreQuery.java Tue Aug 31 19:50:12 2010
@@ -300,7 +300,7 @@ public class CustomScoreQuery extends Qu
// constructor
private CustomScorer(Similarity similarity, IndexReader reader, CustomWeight w,
Scorer subQueryScorer, Scorer[] valSrcScorers) throws IOException {
- super(similarity);
+ super(similarity,w);
this.qWeight = w.getValue();
this.subQueryScorer = subQueryScorer;
this.valSrcScorers = valSrcScorers;
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/ValueSourceQuery.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/ValueSourceQuery.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/ValueSourceQuery.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/function/ValueSourceQuery.java Tue Aug 31 19:50:12 2010
@@ -134,7 +134,7 @@ public class ValueSourceQuery extends Qu
// constructor
private ValueSourceScorer(Similarity similarity, IndexReader reader, ValueSourceWeight w) throws IOException {
- super(similarity);
+ super(similarity,w);
qWeight = w.getValue();
// this is when/where the values are first created.
vals = valSrc.getValues(reader);
Modified: lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/spans/SpanScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/spans/SpanScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/spans/SpanScorer.java (original)
+++ lucene/dev/trunk/lucene/src/java/org/apache/lucene/search/spans/SpanScorer.java Tue Aug 31 19:50:12 2010
@@ -29,7 +29,6 @@ import org.apache.lucene.search.Similari
*/
public class SpanScorer extends Scorer {
protected Spans spans;
- protected Weight weight;
protected byte[] norms;
protected float value;
@@ -40,10 +39,9 @@ public class SpanScorer extends Scorer {
protected SpanScorer(Spans spans, Weight weight, Similarity similarity, byte[] norms)
throws IOException {
- super(similarity);
+ super(similarity, weight);
this.spans = spans;
this.norms = norms;
- this.weight = weight;
this.value = weight.getValue();
if (this.spans.next()) {
doc = -1;
@@ -97,6 +95,11 @@ public class SpanScorer extends Scorer {
float raw = getSimilarity().tf(freq) * value; // raw score
return norms == null? raw : raw * getSimilarity().decodeNormValue(norms[doc]); // normalize
}
+
+ @Override
+ public float freq() throws IOException {
+ return freq;
+ }
/** This method is no longer an official member of {@link Scorer},
* but it is needed by SpanWeight to build an explanation. */
Modified: lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestBooleanScorer.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestBooleanScorer.java?rev=991310&r1=991309&r2=991310&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestBooleanScorer.java (original)
+++ lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestBooleanScorer.java Tue Aug 31 19:50:12 2010
@@ -90,7 +90,7 @@ public class TestBooleanScorer extends L
}
}};
- BooleanScorer bs = new BooleanScorer(sim, 1, Arrays.asList(scorers), null, scorers.length);
+ BooleanScorer bs = new BooleanScorer(null, sim, 1, Arrays.asList(scorers), null, scorers.length);
assertEquals("should have received 3000", 3000, bs.nextDoc());
assertEquals("should have received NO_MORE_DOCS", DocIdSetIterator.NO_MORE_DOCS, bs.nextDoc());
Added: lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java?rev=991310&view=auto
==============================================================================
--- lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java (added)
+++ lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java Tue Aug 31 19:50:12 2010
@@ -0,0 +1,226 @@
+package org.apache.lucene.search;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF 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.
+ */
+
+import org.apache.lucene.document.*;
+import org.apache.lucene.index.*;
+import org.apache.lucene.util.*;
+import org.apache.lucene.search.BooleanClause.Occur;
+import org.apache.lucene.search.Scorer.ScorerVisitor;
+import org.apache.lucene.store.*;
+
+import java.util.*;
+import java.io.*;
+
+import org.junit.Test;
+import org.junit.AfterClass;
+import org.junit.BeforeClass;
+
+import static org.junit.Assert.*;
+
+public class TestSubScorerFreqs extends LuceneTestCaseJ4 {
+
+ private static Directory dir;
+ private static IndexSearcher s;
+
+ @BeforeClass
+ public static void makeIndex() throws Exception {
+ dir = new RAMDirectory();
+ RandomIndexWriter w = new RandomIndexWriter(
+ newStaticRandom(TestSubScorerFreqs.class), dir);
+ // make sure we have more than one segment occationally
+ for (int i = 0; i < 31 * RANDOM_MULTIPLIER; i++) {
+ Document doc = new Document();
+ doc.add(new Field("f", "a b c d b c d c d d", Field.Store.NO,
+ Field.Index.ANALYZED));
+ w.addDocument(doc);
+
+ doc = new Document();
+ doc.add(new Field("f", "a b c d", Field.Store.NO, Field.Index.ANALYZED));
+ w.addDocument(doc);
+ }
+
+ s = new IndexSearcher(w.getReader());
+ w.close();
+ }
+
+ @AfterClass
+ public static void finish() throws Exception {
+ s.getIndexReader().close();
+ s.close();
+ dir.close();
+ }
+
+ private static class CountingCollector extends Collector {
+ private final Collector other;
+ private int docBase;
+
+ public final Map<Integer, Map<Query, Float>> docCounts = new HashMap<Integer, Map<Query, Float>>();
+
+ private final Map<Query, Scorer> subScorers = new HashMap<Query, Scorer>();
+ private final ScorerVisitor<Query, Query, Scorer> visitor = new MockScorerVisitor();
+ private final EnumSet<Occur> collect;
+
+ private class MockScorerVisitor extends ScorerVisitor<Query, Query, Scorer> {
+
+ @Override
+ public void visitOptional(Query parent, Query child, Scorer scorer) {
+ if (collect.contains(Occur.SHOULD))
+ subScorers.put(child, scorer);
+ }
+
+ @Override
+ public void visitProhibited(Query parent, Query child, Scorer scorer) {
+ if (collect.contains(Occur.MUST_NOT))
+ subScorers.put(child, scorer);
+ }
+
+ @Override
+ public void visitRequired(Query parent, Query child, Scorer scorer) {
+ if (collect.contains(Occur.MUST))
+ subScorers.put(child, scorer);
+ }
+
+ }
+
+ public CountingCollector(Collector other) {
+ this(other, EnumSet.allOf(Occur.class));
+ }
+
+ public CountingCollector(Collector other, EnumSet<Occur> collect) {
+ this.other = other;
+ this.collect = collect;
+ }
+
+ @Override
+ public void setScorer(Scorer scorer) throws IOException {
+ other.setScorer(scorer);
+ scorer.visitScorers(visitor);
+ }
+
+ @Override
+ public void collect(int doc) throws IOException {
+ final Map<Query, Float> freqs = new HashMap<Query, Float>();
+ for (Map.Entry<Query, Scorer> ent : subScorers.entrySet()) {
+ Scorer value = ent.getValue();
+ int matchId = value.docID();
+ freqs.put(ent.getKey(), matchId == doc ? value.freq() : 0.0f);
+ }
+ docCounts.put(doc + docBase, freqs);
+ other.collect(doc);
+ }
+
+ @Override
+ public void setNextReader(IndexReader reader, int docBase)
+ throws IOException {
+ this.docBase = docBase;
+ other.setNextReader(reader, docBase);
+ }
+
+ @Override
+ public boolean acceptsDocsOutOfOrder() {
+ return other.acceptsDocsOutOfOrder();
+ }
+ }
+
+ private static final float FLOAT_TOLERANCE = 0.00001F;
+
+ @Test
+ public void testTermQuery() throws Exception {
+ TermQuery q = new TermQuery(new Term("f", "d"));
+ CountingCollector c = new CountingCollector(TopScoreDocCollector.create(10,
+ true));
+ s.search(q, null, c);
+ final int maxDocs = s.maxDoc();
+ assertEquals(maxDocs, c.docCounts.size());
+ for (int i = 0; i < maxDocs; i++) {
+ Map<Query, Float> doc0 = c.docCounts.get(i);
+ assertEquals(1, doc0.size());
+ assertEquals(4.0F, doc0.get(q), FLOAT_TOLERANCE);
+
+ Map<Query, Float> doc1 = c.docCounts.get(++i);
+ assertEquals(1, doc1.size());
+ assertEquals(1.0F, doc1.get(q), FLOAT_TOLERANCE);
+ }
+ }
+
+ @SuppressWarnings("unchecked")
+ @Test
+ public void testBooleanQuery() throws Exception {
+ TermQuery aQuery = new TermQuery(new Term("f", "a"));
+ TermQuery dQuery = new TermQuery(new Term("f", "d"));
+ TermQuery cQuery = new TermQuery(new Term("f", "c"));
+ TermQuery yQuery = new TermQuery(new Term("f", "y"));
+
+ BooleanQuery query = new BooleanQuery();
+ BooleanQuery inner = new BooleanQuery();
+
+ inner.add(cQuery, Occur.SHOULD);
+ inner.add(yQuery, Occur.MUST_NOT);
+ query.add(inner, Occur.MUST);
+ query.add(aQuery, Occur.MUST);
+ query.add(dQuery, Occur.MUST);
+ EnumSet<Occur>[] occurList = new EnumSet[] {EnumSet.of(Occur.MUST), EnumSet.of(Occur.MUST, Occur.SHOULD)};
+ for (EnumSet<Occur> occur : occurList) {
+ CountingCollector c = new CountingCollector(TopScoreDocCollector.create(
+ 10, true), occur);
+ s.search(query, null, c);
+ final int maxDocs = s.maxDoc();
+ assertEquals(maxDocs, c.docCounts.size());
+ boolean includeOptional = occur.contains(Occur.SHOULD);
+ for (int i = 0; i < maxDocs; i++) {
+ Map<Query, Float> doc0 = c.docCounts.get(i);
+ assertEquals(includeOptional ? 5 : 4, doc0.size());
+ assertEquals(1.0F, doc0.get(aQuery), FLOAT_TOLERANCE);
+ assertEquals(4.0F, doc0.get(dQuery), FLOAT_TOLERANCE);
+ if (includeOptional)
+ assertEquals(3.0F, doc0.get(cQuery), FLOAT_TOLERANCE);
+
+ Map<Query, Float> doc1 = c.docCounts.get(++i);
+ assertEquals(includeOptional ? 5 : 4, doc1.size());
+ assertEquals(1.0F, doc1.get(aQuery), FLOAT_TOLERANCE);
+ assertEquals(1.0F, doc1.get(dQuery), FLOAT_TOLERANCE);
+ if (includeOptional)
+ assertEquals(1.0F, doc1.get(cQuery), FLOAT_TOLERANCE);
+
+ }
+ }
+ }
+
+ @Test
+ public void testPhraseQuery() throws Exception {
+ PhraseQuery q = new PhraseQuery();
+ q.add(new Term("f", "b"));
+ q.add(new Term("f", "c"));
+ CountingCollector c = new CountingCollector(TopScoreDocCollector.create(10,
+ true));
+ s.search(q, null, c);
+ final int maxDocs = s.maxDoc();
+ assertEquals(maxDocs, c.docCounts.size());
+ for (int i = 0; i < maxDocs; i++) {
+ Map<Query, Float> doc0 = c.docCounts.get(i);
+ assertEquals(1, doc0.size());
+ assertEquals(2.0F, doc0.get(q), FLOAT_TOLERANCE);
+
+ Map<Query, Float> doc1 = c.docCounts.get(++i);
+ assertEquals(1, doc1.size());
+ assertEquals(1.0F, doc1.get(q), FLOAT_TOLERANCE);
+ }
+
+ }
+}
Propchange: lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java
------------------------------------------------------------------------------
svn:eol-style = native
Propchange: lucene/dev/trunk/lucene/src/test/org/apache/lucene/search/TestSubScorerFreqs.java
------------------------------------------------------------------------------
svn:keywords = Date Author Id Revision HeadURL