From 169fac073f7a4b45ae10a9162da7a459e04a40e3 Mon Sep 17 00:00:00 2001 From: Doug Cutting Date: Mon, 13 Jan 2003 23:50:34 +0000 Subject: [PATCH] Revised internal search APIs. Changes include: a. Queries are no longer modified during a search. This makes it possible, e.g., to reuse the same query instance with multiple indexes from multiple threads. b. Term-expanding queries (e.g. PrefixQuery, WildcardQuery, etc.) now work correctly with MultiSearcher, fixing bugs 12619 and 12667. c. Boosting BooleanQuery's now works, and is supported by the query parser (problem reported by Lee Mallabone). Thus a query like "(+foo +bar)^2 +baz" is now supported and equivalent to "(+foo^2 +bar^2) +baz". d. New method: Query.rewrite(IndexReader). This permits a query to re-write itself as an alternate, more primitive query. Most of the term-expanding query classes (PrefixQuery, WildcardQuery, etc.) are now implemented using this method. e. New method: Searchable.explain(Query q, int doc). This returns an Explanation instance that describes how a particular document is scored against a query. An explanation can be displayed as either plain text, with the toString() method, or as HTML, with the toHtml() method. Note that computing an explanation is as expensive as executing the query over the entire index. This is intended to be used in developing Similarity implementations, and, for good performance, should not be displayed with every hit. f. Scorer and Weight are public, not package protected. It now possible for someone to write a Scorer implementation that is not in the org.apache.lucene.search package. This is still fairly advanced programming, and I don't expect anyone to do this anytime soon, but at least now it is possible. Caution: These are extensive changes and they have not yet been tested extensively. Bug reports are appreciated. git-svn-id: https://svn.apache.org/repos/asf/lucene/java/trunk@149922 13f79535-47bb-0310-9956-ffa450edef68 --- CHANGES.txt | 39 +++ default.properties | 2 +- src/java/org/apache/lucene/index/Term.java | 4 +- .../apache/lucene/queryParser/QueryParser.jj | 12 +- .../apache/lucene/search/BooleanClause.java | 16 + .../apache/lucene/search/BooleanQuery.java | 139 +++++--- .../apache/lucene/search/BooleanScorer.java | 44 ++- .../lucene/search/ExactPhraseScorer.java | 6 +- .../org/apache/lucene/search/Explanation.java | 145 ++++++++ .../org/apache/lucene/search/FuzzyQuery.java | 23 +- src/java/org/apache/lucene/search/Hits.java | 4 +- .../apache/lucene/search/IndexSearcher.java | 19 +- .../apache/lucene/search/MultiSearcher.java | 17 + .../apache/lucene/search/MultiTermQuery.java | 86 ++--- .../lucene/search/PhrasePrefixQuery.java | 310 +++++++++--------- .../org/apache/lucene/search/PhraseQuery.java | 148 ++++++--- .../apache/lucene/search/PhraseScorer.java | 49 ++- .../org/apache/lucene/search/PrefixQuery.java | 76 ++--- src/java/org/apache/lucene/search/Query.java | 127 ++++--- .../org/apache/lucene/search/RangeQuery.java | 146 +++------ .../lucene/search/RemoteSearchable.java | 8 + src/java/org/apache/lucene/search/Scorer.java | 15 +- .../org/apache/lucene/search/Searchable.java | 8 + .../lucene/search/SloppyPhraseScorer.java | 6 +- .../org/apache/lucene/search/TermQuery.java | 104 ++++-- .../org/apache/lucene/search/TermScorer.java | 58 +++- src/java/org/apache/lucene/search/Weight.java | 88 +++++ .../apache/lucene/search/WildcardQuery.java | 17 +- .../lucene/queryParser/TestQueryParser.java | 2 + 29 files changed, 1157 insertions(+), 561 deletions(-) create mode 100644 src/java/org/apache/lucene/search/Explanation.java create mode 100644 src/java/org/apache/lucene/search/Weight.java diff --git a/CHANGES.txt b/CHANGES.txt index 5cde4123de2..19ef764734d 100644 --- a/CHANGES.txt +++ b/CHANGES.txt @@ -104,6 +104,45 @@ $Id$ 21. Added getFields(String) and getValues(String) methods. (Rasik Pandey via otis) + 22. Revised internal search APIs. Changes include: + + a. Queries are no longer modified during a search. This makes + it possible, e.g., to reuse the same query instance with + multiple indexes from multiple threads. + + b. Term-expanding queries (e.g. PrefixQuery, WildcardQuery, + etc.) now work correctly with MultiSearcher, fixing bugs 12619 + and 12667. + + c. Boosting BooleanQuery's now works, and is supported by the + query parser (problem reported by Lee Mallabone). Thus a query + like "(+foo +bar)^2 +baz" is now supported and equivalent to + "(+foo^2 +bar^2) +baz". + + d. New method: Query.rewrite(IndexReader). This permits a + query to re-write itself as an alternate, more primitive query. + Most of the term-expanding query classes (PrefixQuery, + WildcardQuery, etc.) are now implemented using this method. + + e. New method: Searchable.explain(Query q, int doc). This + returns an Explanation instance that describes how a particular + document is scored against a query. An explanation can be + displayed as either plain text, with the toString() method, or + as HTML, with the toHtml() method. Note that computing an + explanation is as expensive as executing the query over the + entire index. This is intended to be used in developing + Similarity implementations, and, for good performance, should + not be displayed with every hit. + + f. Scorer and Weight are public, not package protected. It now + possible for someone to write a Scorer implementation that is + not in the org.apache.lucene.search package. This is still + fairly advanced programming, and I don't expect anyone to do + this anytime soon, but at least now it is possible. + + Caution: These are extensive changes and they have not yet been + tested extensively. Bug reports are appreciated. + Contributed by Rasik Pandey on 2002-10-09 1.2 RC6 diff --git a/default.properties b/default.properties index 7b08e2547a2..bac388bc8e3 100644 --- a/default.properties +++ b/default.properties @@ -10,7 +10,7 @@ Name=Lucene version=1.3-dev1 year=2000-2002 final.name=${name}-${version} -debug=off +debug=on project.name = site docs.src = ./xdocs diff --git a/src/java/org/apache/lucene/index/Term.java b/src/java/org/apache/lucene/index/Term.java index 684be1146ae..9c155efeb3e 100644 --- a/src/java/org/apache/lucene/index/Term.java +++ b/src/java/org/apache/lucene/index/Term.java @@ -116,9 +116,7 @@ public final class Term implements java.io.Serializable { text = txt; } - public final String toString() { - return "Term<" + field + ":" + text + ">"; - } + public final String toString() { return field + ":" + text; } private void readObject(java.io.ObjectInputStream in) throws java.io.IOException, ClassNotFoundException diff --git a/src/java/org/apache/lucene/queryParser/QueryParser.jj b/src/java/org/apache/lucene/queryParser/QueryParser.jj index 098a0c8a2dc..f1d644839a0 100644 --- a/src/java/org/apache/lucene/queryParser/QueryParser.jj +++ b/src/java/org/apache/lucene/queryParser/QueryParser.jj @@ -433,7 +433,7 @@ Query Query(String field) : Query Clause(String field) : { Query q; - Token fieldToken=null; + Token fieldToken=null, boost=null; } { [ @@ -443,9 +443,17 @@ Query Clause(String field) : { ( q=Term(field) - | q=Query(field) + | q=Query(field) ( boost=)? + ) { + if (boost != null) { + float f = (float)1.0; + try { + f = Float.valueOf(boost.image).floatValue(); + q.setBoost(f); + } catch (Exception ignored) { } + } return q; } } diff --git a/src/java/org/apache/lucene/search/BooleanClause.java b/src/java/org/apache/lucene/search/BooleanClause.java index e8337ed4a32..a33d1f82468 100644 --- a/src/java/org/apache/lucene/search/BooleanClause.java +++ b/src/java/org/apache/lucene/search/BooleanClause.java @@ -72,4 +72,20 @@ public class BooleanClause implements java.io.Serializable { required = r; prohibited = p; } + + /** Returns true iff o is equal to this. */ + public boolean equals(Object o) { + if (!(o instanceof BooleanClause)) + return false; + BooleanClause other = (BooleanClause)o; + return this.query.equals(other.query) + && (this.required == other.required) + && (this.prohibited == other.prohibited); + } + + /** Returns a hash code value for this object.*/ + public int hashCode() { + return query.hashCode() ^ (this.required?1:0) ^ (this.prohibited?2:0); + } + } diff --git a/src/java/org/apache/lucene/search/BooleanQuery.java b/src/java/org/apache/lucene/search/BooleanQuery.java index 5d226263577..ffc16798eeb 100644 --- a/src/java/org/apache/lucene/search/BooleanQuery.java +++ b/src/java/org/apache/lucene/search/BooleanQuery.java @@ -88,60 +88,95 @@ public class BooleanQuery extends Query { clauses.addElement(clause); } - void prepare(IndexReader reader) { - for (int i = 0 ; i < clauses.size(); i++) { - BooleanClause c = (BooleanClause)clauses.elementAt(i); - c.query.prepare(reader); - } + /** Returns the set of clauses in this query. */ + public BooleanClause[] getClauses() { + return (BooleanClause[])clauses.toArray(new BooleanClause[0]); } - float sumOfSquaredWeights(Searcher searcher) - throws IOException { - float sum = 0.0f; + private class BooleanWeight implements Weight { + private Searcher searcher; + private float norm; + private Vector weights = new Vector(); - for (int i = 0 ; i < clauses.size(); i++) { - BooleanClause c = (BooleanClause)clauses.elementAt(i); - if (!c.prohibited) - sum += c.query.sumOfSquaredWeights(searcher); // sum sub-query weights + public BooleanWeight(Searcher searcher) { + this.searcher = searcher; + for (int i = 0 ; i < clauses.size(); i++) { + BooleanClause c = (BooleanClause)clauses.elementAt(i); + weights.add(c.query.createWeight(searcher)); + } + } + + public Query getQuery() { return BooleanQuery.this; } + public float getValue() { return getBoost(); } + + public float sumOfSquaredWeights() throws IOException { + float sum = 0.0f; + for (int i = 0 ; i < weights.size(); i++) { + BooleanClause c = (BooleanClause)clauses.elementAt(i); + Weight w = (Weight)weights.elementAt(i); + if (!c.prohibited) + sum += w.sumOfSquaredWeights(); // sum sub weights + } + + sum *= getBoost() * getBoost(); // boost each sub-weight + + return sum ; + } + + + public void normalize(float norm) { + norm *= getBoost(); // incorporate boost + for (int i = 0 ; i < weights.size(); i++) { + BooleanClause c = (BooleanClause)clauses.elementAt(i); + Weight w = (Weight)weights.elementAt(i); + if (!c.prohibited) + w.normalize(norm); + } + } + + public Scorer scorer(IndexReader reader) throws IOException { + if (weights.size() == 1) { // optimize 1-clause queries + BooleanClause c = (BooleanClause)clauses.elementAt(0); + Weight w = (Weight)weights.elementAt(0); + if (!c.prohibited) // just return clause scorer + return w.scorer(reader); + } + + BooleanScorer result = new BooleanScorer(searcher.getSimilarity()); + + for (int i = 0 ; i < weights.size(); i++) { + BooleanClause c = (BooleanClause)clauses.elementAt(0); + Weight w = (Weight)weights.elementAt(i); + Scorer subScorer = w.scorer(reader); + if (subScorer != null) + result.add(subScorer, c.required, c.prohibited); + else if (c.required) + return null; + } + + return result; + } + + public Explanation explain() throws IOException { + Explanation result = new Explanation(); + result.setDescription("boost(" + getQuery() + ")"); + result.setValue(getBoost()); + return result; } - return sum; } - void normalize(float norm) { - for (int i = 0 ; i < clauses.size(); i++) { - BooleanClause c = (BooleanClause)clauses.elementAt(i); - if (!c.prohibited) - c.query.normalize(norm); - } - } - - Scorer scorer(IndexReader reader, Similarity similarity) - throws IOException { - - if (clauses.size() == 1) { // optimize 1-term queries - BooleanClause c = (BooleanClause)clauses.elementAt(0); - if (!c.prohibited) // just return term scorer - return c.query.scorer(reader, similarity); - } - - BooleanScorer result = new BooleanScorer(similarity); - - for (int i = 0 ; i < clauses.size(); i++) { - BooleanClause c = (BooleanClause)clauses.elementAt(i); - Scorer subScorer = c.query.scorer(reader, similarity); - if (subScorer != null) - result.add(subScorer, c.required, c.prohibited); - else if (c.required) - return null; - } - - return result; + protected Weight createWeight(Searcher searcher) { + return new BooleanWeight(searcher); } /** Prints a user-readable version of this query. */ public String toString(String field) { StringBuffer buffer = new StringBuffer(); + if (getBoost() > 1.0) { + buffer.append("("); + } + for (int i = 0 ; i < clauses.size(); i++) { BooleanClause c = (BooleanClause)clauses.elementAt(i); if (c.prohibited) @@ -160,7 +195,27 @@ public class BooleanQuery extends Query { if (i != clauses.size()-1) buffer.append(" "); } + + if (getBoost() > 1.0) { + buffer.append(")^"); + buffer.append(getBoost()); + } + return buffer.toString(); } + /** Returns true iff o is equal to this. */ + public boolean equals(Object o) { + if (!(o instanceof BooleanQuery)) + return false; + BooleanQuery other = (BooleanQuery)o; + return (this.getBoost() == other.getBoost()) + && this.clauses.equals(other.clauses); + } + + /** Returns a hash code value for this object.*/ + public int hashCode() { + return Float.floatToIntBits(getBoost()) ^ clauses.hashCode(); + } + } diff --git a/src/java/org/apache/lucene/search/BooleanScorer.java b/src/java/org/apache/lucene/search/BooleanScorer.java index 7cf416be82a..cd696b68f8d 100644 --- a/src/java/org/apache/lucene/search/BooleanScorer.java +++ b/src/java/org/apache/lucene/search/BooleanScorer.java @@ -117,10 +117,11 @@ final class BooleanScorer extends Scorer { private final void computeCoordFactors() throws IOException { coordFactors = new float[maxCoord]; for (int i = 0; i < maxCoord; i++) - coordFactors[i] = getSimilarity().coord(i, maxCoord); + coordFactors[i] = getSimilarity().coord(i, maxCoord-1); } - final void score(HitCollector results, int maxDoc) throws IOException { + public final void score(HitCollector results, int maxDoc) + throws IOException { if (coordFactors == null) computeCoordFactors(); @@ -205,4 +206,43 @@ final class BooleanScorer extends Scorer { } } } + + public Explanation explain(int doc) throws IOException { + Explanation sumExpl = new Explanation(); + sumExpl.setDescription("sum of:"); + int coord = 0; + float sum = 0.0f; + for (SubScorer s = scorers; s != null; s = s.next) { + Explanation e = s.scorer.explain(doc); + if (e.getValue() > 0) { + if (!s.prohibited) { + sumExpl.addDetail(e); + sum += e.getValue(); + coord++; + } else { + return new Explanation(0.0f, "match prohibited"); + } + } else if (s.required) { + return new Explanation(0.0f, "match required"); + } + } + sumExpl.setValue(sum); + + if (coord == 1) // only one clause matched + sumExpl = sumExpl.getDetails()[0]; // eliminate wrapper + + float coordFactor = getSimilarity().coord(coord, maxCoord-1); + if (coordFactor == 1.0f) // coord is no-op + return sumExpl; // eliminate wrapper + else { + Explanation result = new Explanation(); + result.setDescription("product of:"); + result.addDetail(sumExpl); + result.addDetail(new Explanation(coordFactor, + "coord("+coord+"/"+(maxCoord-1)+")")); + result.setValue(sum*coordFactor); + return result; + } + } + } diff --git a/src/java/org/apache/lucene/search/ExactPhraseScorer.java b/src/java/org/apache/lucene/search/ExactPhraseScorer.java index c33c5c59ba3..32f941d4c70 100644 --- a/src/java/org/apache/lucene/search/ExactPhraseScorer.java +++ b/src/java/org/apache/lucene/search/ExactPhraseScorer.java @@ -61,9 +61,9 @@ import org.apache.lucene.index.*; final class ExactPhraseScorer extends PhraseScorer { - ExactPhraseScorer(TermPositions[] tps, Similarity similarity, - byte[] norms, float weight) throws IOException { - super(tps, similarity, norms, weight); + ExactPhraseScorer(Weight weight, TermPositions[] tps, Similarity similarity, + byte[] norms) throws IOException { + super(weight, tps, similarity, norms); } protected final float phraseFreq() throws IOException { diff --git a/src/java/org/apache/lucene/search/Explanation.java b/src/java/org/apache/lucene/search/Explanation.java new file mode 100644 index 00000000000..e96020c8b41 --- /dev/null +++ b/src/java/org/apache/lucene/search/Explanation.java @@ -0,0 +1,145 @@ +package org.apache.lucene.search; + +/* ==================================================================== + * The Apache Software License, Version 1.1 + * + * Copyright (c) 2003 The Apache Software Foundation. All rights + * reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * 3. The end-user documentation included with the redistribution, + * if any, must include the following acknowledgment: + * "This product includes software developed by the + * Apache Software Foundation (http://www.apache.org/)." + * Alternately, this acknowledgment may appear in the software itself, + * if and wherever such third-party acknowledgments normally appear. + * + * 4. The names "Apache" and "Apache Software Foundation" and + * "Apache Lucene" must not be used to endorse or promote products + * derived from this software without prior written permission. For + * written permission, please contact apache@apache.org. + * + * 5. Products derived from this software may not be called "Apache", + * "Apache Lucene", nor may "Apache" appear in their name, without + * prior written permission of the Apache Software Foundation. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES + * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE + * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR + * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, + * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT + * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * ==================================================================== + * + * This software consists of voluntary contributions made by many + * individuals on behalf of the Apache Software Foundation. For more + * information on the Apache Software Foundation, please see + * . + */ + +import java.util.ArrayList; + +/** Expert: Describes the score computation for document and query. */ +public class Explanation implements java.io.Serializable { + private float value; // the value of this node + private String description; // what it represents + private ArrayList details; // sub-explanations + + public Explanation() {} + + public Explanation(float value, String description) { + this.value = value; + this.description = description; + } + + /** The value assigned to this explanation node. */ + public float getValue() { return value; } + /** Sets the value assigned to this explanation node. */ + public void setValue(float value) { this.value = value; } + + /** A description of this explanation node. */ + public String getDescription() { return description; } + /** Sets the description of this explanation node. */ + public void setDescription(String description) { + this.description = description; + } + + /** The sub-nodes of this explanation node. */ + public Explanation[] getDetails() { + if (details == null) + return null; + return (Explanation[])details.toArray(new Explanation[0]); + } + + /** Adds a sub-node to this explanation node. */ + public void addDetail(Explanation detail) { + if (details == null) + details = new ArrayList(); + details.add(detail); + } + + /** Render an explanation as HTML. */ + public String toString() { + return toString(0); + } + private String toString(int depth) { + StringBuffer buffer = new StringBuffer(); + for (int i = 0; i < depth; i++) { + buffer.append(" "); + } + buffer.append(getValue()); + buffer.append(" = "); + buffer.append(getDescription()); + buffer.append("\n"); + + Explanation[] details = getDetails(); + if (details != null) { + for (int i = 0 ; i < details.length; i++) { + buffer.append(details[i].toString(depth+1)); + } + } + + return buffer.toString(); + } + + + /** Render an explanation as HTML. */ + public String toHtml() { + StringBuffer buffer = new StringBuffer(); + buffer.append("
    \n"); + + buffer.append("
  • "); + buffer.append(getValue()); + buffer.append(" = "); + buffer.append(getDescription()); + buffer.append("
    • \n"); + + Explanation[] details = getDetails(); + if (details != null) { + for (int i = 0 ; i < details.length; i++) { + buffer.append(details[i].toHtml()); + } + } + + buffer.append("
\n"); + + return buffer.toString(); + } +} diff --git a/src/java/org/apache/lucene/search/FuzzyQuery.java b/src/java/org/apache/lucene/search/FuzzyQuery.java index afb4bf3a726..96dbb1f7b9f 100644 --- a/src/java/org/apache/lucene/search/FuzzyQuery.java +++ b/src/java/org/apache/lucene/search/FuzzyQuery.java @@ -60,20 +60,15 @@ import java.io.IOException; /** Implements the fuzzy search query */ public final class FuzzyQuery extends MultiTermQuery { - private Term fuzzyTerm; + public FuzzyQuery(Term term) { + super(term); + } - public FuzzyQuery(Term term) { - super(term); - fuzzyTerm = term; - } + protected FilteredTermEnum getEnum(IndexReader reader) throws IOException { + return new FuzzyTermEnum(reader, getTerm()); + } - final void prepare(IndexReader reader) { - try { - setEnum(new FuzzyTermEnum(reader, fuzzyTerm)); - } catch (IOException e) {} - } - - public String toString(String field) { - return super.toString(field) + '~'; - } + public String toString(String field) { + return super.toString(field) + '~'; + } } diff --git a/src/java/org/apache/lucene/search/Hits.java b/src/java/org/apache/lucene/search/Hits.java index 1588e720e53..207e29ac035 100644 --- a/src/java/org/apache/lucene/search/Hits.java +++ b/src/java/org/apache/lucene/search/Hits.java @@ -93,8 +93,8 @@ public final class Hits { ScoreDoc[] scoreDocs = topDocs.scoreDocs; float scoreNorm = 1.0f; - if (length > 0 && scoreDocs[0].score > 1.0f) - scoreNorm = 1.0f / scoreDocs[0].score; +// if (length > 0 && scoreDocs[0].score > 1.0f) +// scoreNorm = 1.0f / scoreDocs[0].score; int end = scoreDocs.length < length ? scoreDocs.length : length; for (int i = hitDocs.size(); i < end; i++) diff --git a/src/java/org/apache/lucene/search/IndexSearcher.java b/src/java/org/apache/lucene/search/IndexSearcher.java index b6886042308..0a712be306f 100644 --- a/src/java/org/apache/lucene/search/IndexSearcher.java +++ b/src/java/org/apache/lucene/search/IndexSearcher.java @@ -122,7 +122,7 @@ public class IndexSearcher extends Searcher implements Searchable { */ public TopDocs search(Query query, Filter filter, final int nDocs) throws IOException { - Scorer scorer = Query.scorer(query, this, reader); + Scorer scorer = query.weight(this).scorer(reader); if (scorer == null) return new TopDocs(0, new ScoreDoc[0]); @@ -181,10 +181,25 @@ public class IndexSearcher extends Searcher implements Searchable { }; } - Scorer scorer = Query.scorer(query, this, reader); + Scorer scorer = query.weight(this).scorer(reader); if (scorer == null) return; scorer.score(collector, reader.maxDoc()); } + /** */ + public Query rewrite(Query original) throws IOException { + Query query = original; + for (Query rewrittenQuery = query.rewrite(reader); rewrittenQuery != query; + rewrittenQuery = query.rewrite(reader)) { + query = rewrittenQuery; + } + return query; + } + + /** */ + public Explanation explain(Query query, int doc) throws IOException { + return query.weight(this).scorer(reader).explain(doc); + } + } diff --git a/src/java/org/apache/lucene/search/MultiSearcher.java b/src/java/org/apache/lucene/search/MultiSearcher.java index 2fab6955df0..5000832b0ab 100644 --- a/src/java/org/apache/lucene/search/MultiSearcher.java +++ b/src/java/org/apache/lucene/search/MultiSearcher.java @@ -203,4 +203,21 @@ public class MultiSearcher extends Searcher implements Searchable { } } + + /** */ + public Query rewrite(Query original) throws IOException { + Query[] queries = new Query[searchables.length]; + for (int i = 0; i < searchables.length; i++) { + queries[i] = searchables[i].rewrite(original); + } + return original.combine(queries); + } + + + /** */ + public Explanation explain(Query query, int doc) throws IOException { + int i = subSearcher(doc); // find searcher index + return searchables[i].explain(query,doc-starts[i]); // dispatch to searcher + } + } diff --git a/src/java/org/apache/lucene/search/MultiTermQuery.java b/src/java/org/apache/lucene/search/MultiTermQuery.java index 351deeb6249..2acb27e15a6 100644 --- a/src/java/org/apache/lucene/search/MultiTermQuery.java +++ b/src/java/org/apache/lucene/search/MultiTermQuery.java @@ -73,81 +73,55 @@ import org.apache.lucene.index.TermEnum; * MultiTermQuery to provide {@link WildcardTermEnum} and * {@link FuzzyTermEnum}, respectively. */ -public class MultiTermQuery extends Query { +public abstract class MultiTermQuery extends Query { private Term term; - private FilteredTermEnum enum; - private BooleanQuery query; - - /** Enable or disable lucene style toString(field) format */ - private static boolean LUCENE_STYLE_TOSTRING = false; /** Constructs a query for terms matching term. */ public MultiTermQuery(Term term) { this.term = term; } - /** Set the TermEnum to be used */ - protected void setEnum(FilteredTermEnum enum) { - this.enum = enum; + /** Returns the pattern term. */ + public Term getTerm() { return term; } + + /** Construct the enumeration to be used, expanding the pattern term. */ + protected abstract FilteredTermEnum getEnum(IndexReader reader) + throws IOException; + + public Query rewrite(IndexReader reader) throws IOException { + FilteredTermEnum enum = getEnum(reader); + BooleanQuery query = new BooleanQuery(); + try { + do { + Term t = enum.term(); + if (t != null) { + TermQuery tq = new TermQuery(t); // found a match + tq.setBoost(getBoost() * enum.difference()); // set the boost + query.add(tq, false, false); // add to query + } + } while (enum.next()); + } finally { + enum.close(); + } + return query; } - final float sumOfSquaredWeights(Searcher searcher) throws IOException { - return getQuery().sumOfSquaredWeights(searcher); + public Query combine(Query[] queries) { + return Query.mergeBooleanQueries(queries); } - - final void normalize(float norm) { - try { - getQuery().normalize(norm); - } catch (IOException e) { - throw new RuntimeException(e.toString()); - } - } - - final Scorer scorer(IndexReader reader, Similarity similarity) - throws IOException { - return getQuery().scorer(reader, similarity); - } - - private final BooleanQuery getQuery() throws IOException { - if (query == null) { - BooleanQuery q = new BooleanQuery(); - try { - do { - Term t = enum.term(); - if (t != null) { - TermQuery tq = new TermQuery(t); // found a match - tq.setBoost(boost * enum.difference()); // set the boost - q.add(tq, false, false); // add to q - } - } while (enum.next()); - } finally { - enum.close(); - } - query = q; - } - return query; - } - + + /** Prints a user-readable version of this query. */ public String toString(String field) { - if (!LUCENE_STYLE_TOSTRING) { - Query q = null; - try { - q = getQuery(); - } catch (Exception e) {} - if (q != null) { - return "(" + q.toString(field) + ")"; - } - } StringBuffer buffer = new StringBuffer(); if (!term.field().equals(field)) { buffer.append(term.field()); buffer.append(":"); } buffer.append(term.text()); - if (boost != 1.0f) { + if (getBoost() != 1.0f) { buffer.append("^"); - buffer.append(Float.toString(boost)); + buffer.append(Float.toString(getBoost())); } return buffer.toString(); } diff --git a/src/java/org/apache/lucene/search/PhrasePrefixQuery.java b/src/java/org/apache/lucene/search/PhrasePrefixQuery.java index 36a85539e48..d7658233345 100644 --- a/src/java/org/apache/lucene/search/PhrasePrefixQuery.java +++ b/src/java/org/apache/lucene/search/PhrasePrefixQuery.java @@ -62,6 +62,7 @@ import org.apache.lucene.index.IndexReader; import org.apache.lucene.index.MultipleTermPositions; import org.apache.lucene.index.Term; import org.apache.lucene.index.TermPositions; +import org.apache.lucene.index.TermDocs; import org.apache.lucene.search.Query; /** @@ -75,180 +76,177 @@ import org.apache.lucene.search.Query; * @author Anders Nielsen * @version 1.0 */ -public class PhrasePrefixQuery - extends Query -{ - private String _field; - private ArrayList _termArrays = new ArrayList(); +public class PhrasePrefixQuery extends Query { + private String field; + private ArrayList termArrays = new ArrayList(); - private float _idf = 0.0f; - private float _weight = 0.0f; + private float idf = 0.0f; + private float weight = 0.0f; - private int _slop = 0; + private int slop = 0; - /** - * Creates a new PhrasePrefixQuery instance. - * - */ - public PhrasePrefixQuery() - { + /* Sets the phrase slop for this query. + * @see PhraseQuery#setSlop(int) + */ + public void setSlop(int s) { slop = s; } + + /* Sets the phrase slop for this query. + * @see PhraseQuery#getSlop() + */ + public int getSlop() { return slop; } + + /* Add a single term at the next position in the phrase. + * @see PhraseQuery#add(Term) + */ + public void add(Term term) { add(new Term[]{term}); } + + /* Add multiple terms at the next position in the phrase. Any of the terms + * may match. + * + * @see PhraseQuery#add(Term) + */ + public void add(Term[] terms) { + if (termArrays.size() == 0) + field = terms[0].field(); + + for (int i=0; isetSlop method here. - * - * @param s an int value - */ - public void setSlop(int s) - { - _slop = s; + termArrays.add(terms); + } + + private class PhrasePrefixWeight implements Weight { + private Searcher searcher; + private float value; + private float idf; + private float queryNorm; + + public PhrasePrefixWeight(Searcher searcher) { + this.searcher = searcher; } - /** - * Describe getSlop method here. - * - * @return an int value - */ - public int getSlop() - { - return _slop; + public Query getQuery() { return PhrasePrefixQuery.this; } + public float getValue() { return value; } + + public float sumOfSquaredWeights() throws IOException { + Iterator i = termArrays.iterator(); + while (i.hasNext()) { + Term[] terms = (Term[])i.next(); + for (int j=0; jadd method here. - * - * @param term a Term value - */ - public void add(Term term) - { - add(new Term[]{term}); + public void normalize(float norm) { + queryNorm = norm; + queryNorm *= idf; // factor from document + value *= queryNorm; // normalize for query } - /** - * Describe add method here. - * - * @param terms a Term[] value - */ - public void add(Term[] terms) - { - if (_termArrays.size() == 0) - _field = terms[0].field(); + public Scorer scorer(IndexReader reader) throws IOException { + if (termArrays.size() == 0) // optimize zero-term case + return null; + + if (termArrays.size() == 1) { // optimize one-term case + Term[] terms = (Term[])termArrays.get(0); + + BooleanScorer bos = new BooleanScorer(searcher.getSimilarity()); + for (int i=0; i 1) + p = new MultipleTermPositions(reader, terms); + else + p = reader.termPositions(terms[0]); + + if (p == null) + return null; + + tps[i] = p; + } + + if (slop == 0) + return new ExactPhraseScorer(this, tps, searcher.getSimilarity(), + reader.norms(field)); + else + return new SloppyPhraseScorer(this, tps, searcher.getSimilarity(), + slop, reader.norms(field)); + } + + public Explanation explain() throws IOException { + Query q = getQuery(); - _termArrays.add(terms); + Explanation result = new Explanation(); + result.setDescription("weight(" + getQuery() + "), product of:"); + + Explanation boostExpl = new Explanation(getBoost(), "boost"); + if (getBoost() != 1.0f) + result.addDetail(boostExpl); + + Explanation idfExpl = new Explanation(idf, "idf"); + result.addDetail(idfExpl); + + Explanation normExpl = new Explanation(queryNorm, "queryNorm"); + result.addDetail(normExpl); + + result.setValue(boostExpl.getValue() * + idfExpl.getValue() * + normExpl.getValue()); + + return result; + } + } + + protected Weight createWeight(Searcher searcher) { + return new PhrasePrefixWeight(searcher); + } + + /** Prints a user-readable version of this query. */ + public final String toString(String f) { + StringBuffer buffer = new StringBuffer(); + if (!field.equals(f)) { + buffer.append(field); + buffer.append(":"); } - Scorer scorer(IndexReader reader, Similarity similarity) - throws IOException - { - if (_termArrays.size() == 0) // optimize zero-term case - return null; + buffer.append("\""); + Iterator i = termArrays.iterator(); + while (i.hasNext()) { + Term[] terms = (Term[])i.next(); + buffer.append(terms[0].text() + (terms.length > 0 ? "*" : "")); + } + buffer.append("\""); - if (_termArrays.size() == 1) // optimize one-term case - { - Term[] terms = (Term[])_termArrays.get(0); - - BooleanQuery boq = new BooleanQuery(); - for (int i=0; i 1) - p = new MultipleTermPositions(reader, terms); - else - p = reader.termPositions(terms[0]); - - if (p == null) - return null; - - tps[i] = p; - } - - if (_slop == 0) - return new ExactPhraseScorer(tps, similarity, - reader.norms(_field), _weight); - else - return new SloppyPhraseScorer(tps, similarity, _slop, - reader.norms(_field), _weight); + if (slop != 0) { + buffer.append("~"); + buffer.append(slop); } - float sumOfSquaredWeights(Searcher searcher) - throws IOException - { - Iterator i = _termArrays.iterator(); - while (i.hasNext()) - { - Term[] terms = (Term[])i.next(); - for (int j=0; jtoString method here. - * - * This method assumes that the first term in a array of terms is the - * prefix for the whole array. That might not necessarily be so. - * - * @param f a String value - * @return a String value - */ - public final String toString(String f) - { - StringBuffer buffer = new StringBuffer(); - if (!_field.equals(f)) - { - buffer.append(_field); - buffer.append(":"); - } - - buffer.append("\""); - Iterator i = _termArrays.iterator(); - while (i.hasNext()) - { - Term[] terms = (Term[])i.next(); - buffer.append(terms[0].text() + (terms.length > 0 ? "*" : "")); - } - buffer.append("\""); - - if (_slop != 0) - { - buffer.append("~"); - buffer.append(_slop); - } - - if (boost != 1.0f) - { - buffer.append("^"); - buffer.append(Float.toString(boost)); - } - - return buffer.toString(); - } + return buffer.toString(); + } } diff --git a/src/java/org/apache/lucene/search/PhraseQuery.java b/src/java/org/apache/lucene/search/PhraseQuery.java index 9ac6e08218d..7fb1611d45f 100644 --- a/src/java/org/apache/lucene/search/PhraseQuery.java +++ b/src/java/org/apache/lucene/search/PhraseQuery.java @@ -68,15 +68,10 @@ import org.apache.lucene.index.IndexReader; public class PhraseQuery extends Query { private String field; private Vector terms = new Vector(); - private float idf = 0.0f; - private float weight = 0.0f; - private int slop = 0; - /** Constructs an empty phrase query. */ - public PhraseQuery() { - } + public PhraseQuery() {} /** Sets the number of other words permitted between words in query phrase. If zero, then this is an exact phrase search. For larger values this works @@ -107,48 +102,103 @@ public class PhraseQuery extends Query { terms.addElement(term); } - final float sumOfSquaredWeights(Searcher searcher) throws IOException { - idf = searcher.getSimilarity().idf(terms, searcher); - weight = idf * boost; - return weight * weight; // square term weights + /** Returns the set of terms in this phrase. */ + public Term[] getTerms() { + return (Term[])terms.toArray(new Term[0]); } - final void normalize(float norm) { - weight *= norm; // normalize for query - weight *= idf; // factor from document - } + private class PhraseWeight implements Weight { + private Searcher searcher; + private float value; + private float idf; + private float queryNorm; - final Scorer scorer(IndexReader reader, Similarity similarity) - throws IOException { - if (terms.size() == 0) // optimize zero-term case - return null; - if (terms.size() == 1) { // optimize one-term case - Term term = (Term)terms.elementAt(0); - TermDocs docs = reader.termDocs(term); - if (docs == null) - return null; - return new TermScorer(docs, similarity, - reader.norms(term.field()), weight); + public PhraseWeight(Searcher searcher) { + this.searcher = searcher; } - TermPositions[] tps = new TermPositions[terms.size()]; - for (int i = 0; i < terms.size(); i++) { - TermPositions p = reader.termPositions((Term)terms.elementAt(i)); - if (p == null) - return null; - tps[i] = p; + public Query getQuery() { return PhraseQuery.this; } + public float getValue() { return value; } + + public float sumOfSquaredWeights() throws IOException { + idf = searcher.getSimilarity().idf(terms, searcher); + value = idf * getBoost(); + return value * value; // square term weights } - if (slop == 0) // optimize exact case - return new ExactPhraseScorer(tps, similarity, - reader.norms(field), weight); - else - return - new SloppyPhraseScorer(tps, similarity, slop, - reader.norms(field), weight); + public void normalize(float norm) { + queryNorm = norm; + queryNorm *= idf; // factor from document + value *= queryNorm; // normalize for query + } + public Scorer scorer(IndexReader reader) throws IOException { + if (terms.size() == 0) // optimize zero-term case + return null; + if (terms.size() == 1) { // optimize one-term case + Term term = (Term)terms.elementAt(0); + TermDocs docs = reader.termDocs(term); + if (docs == null) + return null; + return new TermScorer(this, docs, searcher.getSimilarity(), + reader.norms(term.field())); + } + + TermPositions[] tps = new TermPositions[terms.size()]; + for (int i = 0; i < terms.size(); i++) { + TermPositions p = reader.termPositions((Term)terms.elementAt(i)); + if (p == null) + return null; + tps[i] = p; + } + + if (slop == 0) // optimize exact case + return new ExactPhraseScorer(this, tps, searcher.getSimilarity(), + reader.norms(field)); + else + return + new SloppyPhraseScorer(this, tps, searcher.getSimilarity(), slop, + reader.norms(field)); + + } + + public Explanation explain() throws IOException { + Query q = getQuery(); + + Explanation result = new Explanation(); + result.setDescription("weight(" + getQuery() + "), product of:"); + + Explanation boostExpl = new Explanation(getBoost(), "boost"); + if (getBoost() != 1.0f) + result.addDetail(boostExpl); + + StringBuffer docFreqs = new StringBuffer(); + for (int i = 0; i < terms.size(); i++) { + if (i != 0) docFreqs.append(" "); + docFreqs.append(((Term)terms.elementAt(i)).text()); + docFreqs.append("="); + docFreqs.append(searcher.docFreq((Term)terms.elementAt(i))); + } + Explanation idfExpl = + new Explanation(idf, "idf(" + field + ": " + docFreqs + ")"); + result.addDetail(idfExpl); + + Explanation normExpl = new Explanation(queryNorm, "queryNorm"); + result.addDetail(normExpl); + + result.setValue(boostExpl.getValue() * + idfExpl.getValue() * + normExpl.getValue()); + + return result; + } } + protected Weight createWeight(Searcher searcher) { + return new PhraseWeight(searcher); + } + + /** Prints a user-readable version of this query. */ public String toString(String f) { StringBuffer buffer = new StringBuffer(); @@ -170,11 +220,29 @@ public class PhraseQuery extends Query { buffer.append(slop); } - if (boost != 1.0f) { + if (getBoost() != 1.0f) { buffer.append("^"); - buffer.append(Float.toString(boost)); + buffer.append(Float.toString(getBoost())); } return buffer.toString(); } + + /** Returns true iff o is equal to this. */ + public boolean equals(Object o) { + if (!(o instanceof PhraseQuery)) + return false; + PhraseQuery other = (PhraseQuery)o; + return (this.getBoost() == other.getBoost()) + && (this.slop == other.slop) + && this.terms.equals(other.terms); + } + + /** Returns a hash code value for this object.*/ + public int hashCode() { + return Float.floatToIntBits(getBoost()) + ^ Float.floatToIntBits(slop) + ^ terms.hashCode(); + } + } diff --git a/src/java/org/apache/lucene/search/PhraseScorer.java b/src/java/org/apache/lucene/search/PhraseScorer.java index 853ed7294a1..2bd5ee5b6b7 100644 --- a/src/java/org/apache/lucene/search/PhraseScorer.java +++ b/src/java/org/apache/lucene/search/PhraseScorer.java @@ -60,17 +60,21 @@ import org.apache.lucene.util.*; import org.apache.lucene.index.*; abstract class PhraseScorer extends Scorer { + private Weight weight; protected byte[] norms; - protected float weight; + protected float value; protected PhraseQueue pq; protected PhrasePositions first, last; - PhraseScorer(TermPositions[] tps, Similarity similarity, - byte[] norms, float weight) throws IOException { + private float freq; + + PhraseScorer(Weight weight, TermPositions[] tps, Similarity similarity, + byte[] norms) throws IOException { super(similarity); this.norms = norms; this.weight = weight; + this.value = weight.getValue(); // use PQ to build a sorted list of PhrasePositions pq = new PhraseQueue(tps.length); @@ -79,7 +83,7 @@ abstract class PhraseScorer extends Scorer { pqToList(); } - final void score(HitCollector results, int end) throws IOException { + public final void score(HitCollector results, int end) throws IOException { Similarity similarity = getSimilarity(); while (last.doc < end) { // find doc w/ all the terms while (first.doc < last.doc) { // scan forward in first @@ -92,10 +96,10 @@ abstract class PhraseScorer extends Scorer { } // found doc with all terms - float freq = phraseFreq(); // check for phrase + freq = phraseFreq(); // check for phrase if (freq > 0.0) { - float score = similarity.tf(freq)*weight; // compute score + float score = similarity.tf(freq)*value; // compute score score *= Similarity.decodeNorm(norms[first.doc]); // normalize results.collect(first.doc, score); // add to results } @@ -124,4 +128,37 @@ abstract class PhraseScorer extends Scorer { first = first.next; last.next = null; } + + public Explanation explain(final int doc) throws IOException { + Explanation result = new Explanation(); + PhraseQuery query = (PhraseQuery)weight.getQuery(); + + result.setDescription("phraseScore(" + query + "), product of:"); + + Explanation weightExplanation = weight.explain(); + result.addDetail(weightExplanation); + + Explanation tfExplanation = new Explanation(); + score(new HitCollector() { + public final void collect(int d, float score) {} + }, doc+1); + + float phraseFreq = (first.doc == doc) ? freq : 0.0f; + tfExplanation.setValue(getSimilarity().tf(phraseFreq)); + tfExplanation.setDescription("tf(phraseFreq=" + phraseFreq + ")"); + result.addDetail(tfExplanation); + + Explanation normExplanation = new Explanation(); + normExplanation.setValue(Similarity.decodeNorm(norms[doc])); + String field = query.getTerms()[0].field(); + normExplanation.setDescription("norm(field="+field + ", doc="+doc + ")"); + result.addDetail(normExplanation); + + result.setValue(weightExplanation.getValue() * + tfExplanation.getValue() * + normExplanation.getValue()); + + return result; + } + } diff --git a/src/java/org/apache/lucene/search/PrefixQuery.java b/src/java/org/apache/lucene/search/PrefixQuery.java index cc5f63d71b8..b8a6c66a7a3 100644 --- a/src/java/org/apache/lucene/search/PrefixQuery.java +++ b/src/java/org/apache/lucene/search/PrefixQuery.java @@ -63,65 +63,41 @@ import org.apache.lucene.index.IndexReader; /** A Query that matches documents containing terms with a specified prefix. */ public class PrefixQuery extends Query { private Term prefix; - private IndexReader reader; - private BooleanQuery query; /** Constructs a query for terms starting with prefix. */ public PrefixQuery(Term prefix) { this.prefix = prefix; - this.reader = reader; } - final void prepare(IndexReader reader) { - this.query = null; - this.reader = reader; - } - - final float sumOfSquaredWeights(Searcher searcher) - throws IOException { - return getQuery().sumOfSquaredWeights(searcher); - } - - void normalize(float norm) { + public Query rewrite(IndexReader reader) throws IOException { + BooleanQuery query = new BooleanQuery(); + TermEnum enum = reader.terms(prefix); try { - getQuery().normalize(norm); - } catch (IOException e) { - throw new RuntimeException(e.toString()); - } - } - - Scorer scorer(IndexReader reader, Similarity similarity) throws IOException { - return getQuery().scorer(reader, similarity); - } - - private BooleanQuery getQuery() throws IOException { - if (query == null) { - BooleanQuery q = new BooleanQuery(); - TermEnum enum = reader.terms(prefix); - try { - String prefixText = prefix.text(); - String prefixField = prefix.field(); - do { - Term term = enum.term(); - if (term != null && - term.text().startsWith(prefixText) && - term.field() == prefixField) { - TermQuery tq = new TermQuery(term); // found a match - tq.setBoost(boost); // set the boost - q.add(tq, false, false); // add to q - //System.out.println("added " + term); - } else { - break; - } - } while (enum.next()); - } finally { - enum.close(); - } - query = q; + String prefixText = prefix.text(); + String prefixField = prefix.field(); + do { + Term term = enum.term(); + if (term != null && + term.text().startsWith(prefixText) && + term.field() == prefixField) { + TermQuery tq = new TermQuery(term); // found a match + tq.setBoost(getBoost()); // set the boost + query.add(tq, false, false); // add to query + //System.out.println("added " + term); + } else { + break; + } + } while (enum.next()); + } finally { + enum.close(); } return query; } + public Query combine(Query[] queries) { + return Query.mergeBooleanQueries(queries); + } + /** Prints a user-readable version of this query. */ public String toString(String field) { StringBuffer buffer = new StringBuffer(); @@ -131,9 +107,9 @@ public class PrefixQuery extends Query { } buffer.append(prefix.text()); buffer.append('*'); - if (boost != 1.0f) { + if (getBoost() != 1.0f) { buffer.append("^"); - buffer.append(Float.toString(boost)); + buffer.append(Float.toString(getBoost())); } return buffer.toString(); } diff --git a/src/java/org/apache/lucene/search/Query.java b/src/java/org/apache/lucene/search/Query.java index b3b84727e5c..ed8cc467707 100644 --- a/src/java/org/apache/lucene/search/Query.java +++ b/src/java/org/apache/lucene/search/Query.java @@ -55,8 +55,10 @@ package org.apache.lucene.search; */ import java.io.IOException; -import java.util.Hashtable; -import org.apache.lucene.document.Document; + +import java.util.HashSet; +import java.util.Iterator; + import org.apache.lucene.index.IndexReader; /** The abstract base class for queries. @@ -76,52 +78,93 @@ import org.apache.lucene.index.IndexReader;
  • {@link org.apache.lucene.queryParser.QueryParser QueryParser} */ -public abstract class Query implements java.io.Serializable -{ - // query boost factor - protected float boost = 1.0f; +public abstract class Query implements java.io.Serializable { + private float boost = 1.0f; // query boost factor - // query weighting - abstract float sumOfSquaredWeights(Searcher searcher) throws IOException; - abstract void normalize(float norm); + /** Sets the boost for this query clause to b. Documents + * matching this clause will (in addition to the normal weightings) have + * their score multiplied by b. + */ + public void setBoost(float b) { boost = b; } - // query evaluation - abstract Scorer scorer(IndexReader reader, Similarity similarity) - throws IOException; + /** Gets the boost for this clause. Documents matching + * this clause will (in addition to the normal weightings) have their score + * multiplied by b. The boost is 1.0 by default. + */ + public float getBoost() { return boost; } - void prepare(IndexReader reader) {} + /** Prints a query to a string, with field as the default field + * for terms.

    The representation used is one that is readable by {@link + * org.apache.lucene.queryParser.QueryParser QueryParser} (although, if the + * query was created by the parser, the printed representation may not be + * exactly what was parsed). + */ + public abstract String toString(String field); - static Scorer scorer(Query query, Searcher searcher, IndexReader reader) - throws IOException { - Similarity similarity = searcher.getSimilarity(); - query.prepare(reader); - float sum = query.sumOfSquaredWeights(searcher); - float norm = similarity.queryNorm(sum); - query.normalize(norm); - return query.scorer(reader, similarity); + /** Prints a query to a string. */ + public String toString() { + return toString(""); + } + + /** Expert: Constructs an appropriate Weight implementation for this query. + * + *

    Only implemented by primitive queries, which re-write to themselves. + */ + protected Weight createWeight(Searcher searcher) { + throw new UnsupportedOperationException(); + } + + /** Expert: Constructs an initializes a Weight for a top-level query. */ + public Weight weight(Searcher searcher) + throws IOException { + Query query = searcher.rewrite(this); + Weight weight = query.createWeight(searcher); + float sum = weight.sumOfSquaredWeights(); + float norm = searcher.getSimilarity().queryNorm(sum); + weight.normalize(norm); + return weight; + } + + /** Expert: called to re-write queries into primitive queries. + * + *

    Only implemented by derived queries, with no {@link + * #createWeight(Searcher)} implementatation.. + */ + public Query rewrite(IndexReader reader) throws IOException { + return this; + } + + /** Expert: called when re-writing queries under MultiSearcher. + * + *

    Only implemented by derived queries, with no {@link + * #createWeight(Searcher)} implementatation.. + */ + public Query combine(Query[] queries) { + throw new UnsupportedOperationException(); + } + + + /** Expert: merges the clauses of a set of BooleanQuery's into a single + * BooleanQuery. + * + *

    A utility for use by {@link #combine(Query[])} implementations. + */ + public static Query mergeBooleanQueries(Query[] queries) { + HashSet allClauses = new HashSet(); + for (int i = 0; i < queries.length; i++) { + BooleanClause[] clauses = ((BooleanQuery)queries[i]).getClauses(); + for (int j = 0; j < clauses.length; j++) { + allClauses.add(clauses[j]); + } } - /** - * Sets the boost for this term to b. Documents containing - * this term will (in addition to the normal weightings) have their score - * multiplied by b. - */ - public void setBoost(float b) { boost = b; } + BooleanQuery result = new BooleanQuery(); + Iterator i = allClauses.iterator(); + while (i.hasNext()) { + result.add((BooleanClause)i.next()); + } + return result; + } - /** - * Gets the boost for this term. Documents containing - * this term will (in addition to the normal weightings) have their score - * multiplied by b. The boost is 1.0 by default. - */ - public float getBoost() { return boost; } - /** - * Prints a query to a string, with field as the default field - * for terms. - *

    The representation used is one that is readable by - * {@link org.apache.lucene.queryParser.QueryParser QueryParser} - * (although, if the query was created by the parser, the printed - * representation may not be exactly what was parsed). - */ - public abstract String toString(String field); } diff --git a/src/java/org/apache/lucene/search/RangeQuery.java b/src/java/org/apache/lucene/search/RangeQuery.java index 277e174d289..6eaa1af9568 100644 --- a/src/java/org/apache/lucene/search/RangeQuery.java +++ b/src/java/org/apache/lucene/search/RangeQuery.java @@ -66,8 +66,6 @@ public class RangeQuery extends Query private Term lowerTerm; private Term upperTerm; private boolean inclusive; - private IndexReader reader; - private BooleanQuery query; /** Constructs a query selecting all terms greater than * lowerTerm but less than upperTerm. @@ -89,99 +87,59 @@ public class RangeQuery extends Query this.upperTerm = upperTerm; this.inclusive = inclusive; } - - final void prepare(IndexReader reader) - { - this.query = null; - this.reader = reader; - } - - final float sumOfSquaredWeights(Searcher searcher) throws IOException - { - return getQuery().sumOfSquaredWeights(searcher); - } - - void normalize(float norm) - { - try - { - getQuery().normalize(norm); - } - catch (IOException e) - { - throw new RuntimeException(e.toString()); - } - } - - Scorer scorer(IndexReader reader, Similarity similarity) throws IOException - { - return getQuery().scorer(reader, similarity); - } - - private BooleanQuery getQuery() throws IOException - { - if (query == null) - { - BooleanQuery q = new BooleanQuery(); - // if we have a lowerTerm, start there. otherwise, start at beginning - if (lowerTerm == null) lowerTerm = new Term(getField(), ""); - TermEnum enum = reader.terms(lowerTerm); - try - { - String lowerText = null; - String field; - boolean checkLower = false; - if (!inclusive) // make adjustments to set to exclusive - { - if (lowerTerm != null) - { - lowerText = lowerTerm.text(); - checkLower = true; - } - if (upperTerm != null) - { - // set upperTerm to an actual term in the index - TermEnum uppEnum = reader.terms(upperTerm); - upperTerm = uppEnum.term(); - } - } - String testField = getField(); - do - { - Term term = enum.term(); - if (term != null && term.field() == testField) - { - if (!checkLower || term.text().compareTo(lowerText) > 0) - { - checkLower = false; - if (upperTerm != null) - { - int compare = upperTerm.compareTo(term); - /* if beyond the upper term, or is exclusive and - * this is equal to the upper term, break out */ - if ((compare < 0) || (!inclusive && compare == 0)) break; - } - TermQuery tq = new TermQuery(term); // found a match - tq.setBoost(boost); // set the boost - q.add(tq, false, false); // add to q - } - } - else - { - break; - } - } - while (enum.next()); - } - finally - { - enum.close(); + + public Query rewrite(IndexReader reader) throws IOException { + BooleanQuery query = new BooleanQuery(); + // if we have a lowerTerm, start there. otherwise, start at beginning + if (lowerTerm == null) lowerTerm = new Term(getField(), ""); + TermEnum enum = reader.terms(lowerTerm); + try { + String lowerText = null; + String field; + boolean checkLower = false; + if (!inclusive) { // make adjustments to set to exclusive + if (lowerTerm != null) { + lowerText = lowerTerm.text(); + checkLower = true; } - query = q; - } - return query; + if (upperTerm != null) { + // set upperTerm to an actual term in the index + TermEnum uppEnum = reader.terms(upperTerm); + upperTerm = uppEnum.term(); + } + } + String testField = getField(); + do { + Term term = enum.term(); + if (term != null && term.field() == testField) { + if (!checkLower || term.text().compareTo(lowerText) > 0) { + checkLower = false; + if (upperTerm != null) { + int compare = upperTerm.compareTo(term); + /* if beyond the upper term, or is exclusive and + * this is equal to the upper term, break out */ + if ((compare < 0) || (!inclusive && compare == 0)) break; + } + TermQuery tq = new TermQuery(term); // found a match + tq.setBoost(getBoost()); // set the boost + query.add(tq, false, false); // add to query + } + } + else { + break; + } + } + while (enum.next()); + } finally { + enum.close(); + } + return query; } + public Query combine(Query[] queries) { + return Query.mergeBooleanQueries(queries); + } + private String getField() { return (lowerTerm != null ? lowerTerm.field() : upperTerm.field()); @@ -201,10 +159,10 @@ public class RangeQuery extends Query buffer.append("-"); buffer.append(upperTerm != null ? upperTerm.text() : "null"); buffer.append(inclusive ? "]" : "}"); - if (boost != 1.0f) + if (getBoost() != 1.0f) { buffer.append("^"); - buffer.append(Float.toString(boost)); + buffer.append(Float.toString(getBoost())); } return buffer.toString(); } diff --git a/src/java/org/apache/lucene/search/RemoteSearchable.java b/src/java/org/apache/lucene/search/RemoteSearchable.java index 5059c9f45f9..62474f1ab42 100644 --- a/src/java/org/apache/lucene/search/RemoteSearchable.java +++ b/src/java/org/apache/lucene/search/RemoteSearchable.java @@ -102,6 +102,14 @@ public class RemoteSearchable return local.doc(i); } + public Query rewrite(Query original) throws IOException { + return local.rewrite(original); + } + + public Explanation explain(Query query, int doc) throws IOException { + return local.explain(query, doc); + } + /** Exports a searcher for the index in args[0] named * "//localhost/Searchable". */ public static void main(String args[]) throws Exception { diff --git a/src/java/org/apache/lucene/search/Scorer.java b/src/java/org/apache/lucene/search/Scorer.java index bd04dfb2580..68d03a41de0 100644 --- a/src/java/org/apache/lucene/search/Scorer.java +++ b/src/java/org/apache/lucene/search/Scorer.java @@ -56,16 +56,27 @@ package org.apache.lucene.search; import java.io.IOException; -abstract class Scorer { +/** Expert: Implements scoring for a class of queries. */ +public abstract class Scorer { private Similarity similarity; + /** Constructs a Scorer. */ protected Scorer(Similarity similarity) { this.similarity = similarity; } + /** Returns the Similarity implementation used by this scorer. */ public Similarity getSimilarity() { return this.similarity; } - abstract void score(HitCollector hc, int maxDoc) throws IOException; + /** Scores hits and passes them to a collector. Stops at the last document + * before maxDoc. If called repeatedly, will restart at point + * where it last left off. + */ + public abstract void score(HitCollector hc, int maxDoc) throws IOException; + + /** Returns an explanation of the score for doc. */ + public abstract Explanation explain(int doc) throws IOException; + } diff --git a/src/java/org/apache/lucene/search/Searchable.java b/src/java/org/apache/lucene/search/Searchable.java index 0c2ff72a3b2..c31938dd864 100644 --- a/src/java/org/apache/lucene/search/Searchable.java +++ b/src/java/org/apache/lucene/search/Searchable.java @@ -112,4 +112,12 @@ public interface Searchable extends java.rmi.Remote { * @see IndexReader#document(int). */ Document doc(int i) throws IOException; + + /** */ + Query rewrite(Query query) throws IOException; + + /** */ + Explanation explain(Query query, int doc) throws IOException; + + } diff --git a/src/java/org/apache/lucene/search/SloppyPhraseScorer.java b/src/java/org/apache/lucene/search/SloppyPhraseScorer.java index c3afa75b485..effce5f1343 100644 --- a/src/java/org/apache/lucene/search/SloppyPhraseScorer.java +++ b/src/java/org/apache/lucene/search/SloppyPhraseScorer.java @@ -62,9 +62,9 @@ import org.apache.lucene.index.*; final class SloppyPhraseScorer extends PhraseScorer { private int slop; - SloppyPhraseScorer(TermPositions[] tps, Similarity similarity, - int slop, byte[] norms, float weight) throws IOException { - super(tps, similarity, norms, weight); + SloppyPhraseScorer(Weight weight, TermPositions[] tps, Similarity similarity, + int slop, byte[] norms) throws IOException { + super(weight, tps, similarity, norms); this.slop = slop; } diff --git a/src/java/org/apache/lucene/search/TermQuery.java b/src/java/org/apache/lucene/search/TermQuery.java index 3e666a49764..d6f6b3d6df5 100644 --- a/src/java/org/apache/lucene/search/TermQuery.java +++ b/src/java/org/apache/lucene/search/TermQuery.java @@ -64,34 +64,77 @@ import org.apache.lucene.index.IndexReader; */ public class TermQuery extends Query { private Term term; - private float idf = 0.0f; - private float weight = 0.0f; + + private class TermWeight implements Weight { + private Searcher searcher; + private float value; + private float idf; + private float queryNorm; + + public TermWeight(Searcher searcher) { + this.searcher = searcher; + } + + public Query getQuery() { return TermQuery.this; } + public float getValue() { return value; } + + public float sumOfSquaredWeights() throws IOException { + idf = searcher.getSimilarity().idf(term, searcher); + value = idf * getBoost(); + return value * value; // square term weights + } + + public void normalize(float norm) { + queryNorm = norm; + queryNorm *= idf; // factor from document + value *= queryNorm; // normalize for query + } + + public Scorer scorer(IndexReader reader) throws IOException { + TermDocs termDocs = reader.termDocs(term); + + if (termDocs == null) + return null; + + return new TermScorer(this, termDocs, searcher.getSimilarity(), + reader.norms(term.field())); + } + + public Explanation explain() throws IOException { + Query q = getQuery(); + + Explanation result = new Explanation(); + result.setDescription("weight(" + getQuery() + "), product of:"); + + Explanation boostExpl = new Explanation(getBoost(), "boost"); + if (getBoost() != 1.0f) + result.addDetail(boostExpl); + + Explanation idfExpl = + new Explanation(idf, "idf(docFreq=" + searcher.docFreq(term) + ")"); + result.addDetail(idfExpl); + + Explanation normExpl = new Explanation(queryNorm,"queryNorm"); + result.addDetail(normExpl); + + result.setValue(boostExpl.getValue() * + idfExpl.getValue() * + normExpl.getValue()); + + return result; + } + } /** Constructs a query for the term t. */ public TermQuery(Term t) { term = t; } - final float sumOfSquaredWeights(Searcher searcher) throws IOException { - idf = searcher.getSimilarity().idf(term, searcher); - weight = idf * boost; - return weight * weight; // square term weights - } + /** Returns the term of this query. */ + public Term getTerm() { return term; }; - final void normalize(float norm) { - weight *= norm; // normalize for query - weight *= idf; // factor from document - } - - Scorer scorer(IndexReader reader, Similarity similarity) - throws IOException { - TermDocs termDocs = reader.termDocs(term); - - if (termDocs == null) - return null; - - return new TermScorer(termDocs, similarity, - reader.norms(term.field()), weight); + protected Weight createWeight(Searcher searcher) { + return new TermWeight(searcher); } /** Prints a user-readable version of this query. */ @@ -102,10 +145,25 @@ public class TermQuery extends Query { buffer.append(":"); } buffer.append(term.text()); - if (boost != 1.0f) { + if (getBoost() != 1.0f) { buffer.append("^"); - buffer.append(Float.toString(boost)); + buffer.append(Float.toString(getBoost())); } return buffer.toString(); } + + /** Returns true iff o is equal to this. */ + public boolean equals(Object o) { + if (!(o instanceof TermQuery)) + return false; + TermQuery other = (TermQuery)o; + return (this.getBoost() == other.getBoost()) + && this.term.equals(other.term); + } + + /** Returns a hash code value for this object.*/ + public int hashCode() { + return Float.floatToIntBits(getBoost()) ^ term.hashCode(); + } + } diff --git a/src/java/org/apache/lucene/search/TermScorer.java b/src/java/org/apache/lucene/search/TermScorer.java index 7582c66e7ec..45b1cdde066 100644 --- a/src/java/org/apache/lucene/search/TermScorer.java +++ b/src/java/org/apache/lucene/search/TermScorer.java @@ -55,12 +55,14 @@ package org.apache.lucene.search; */ import java.io.IOException; +import org.apache.lucene.index.Term; import org.apache.lucene.index.TermDocs; final class TermScorer extends Scorer { + private Weight weight; private TermDocs termDocs; private byte[] norms; - private float weight; + private float weightValue; private int doc; private final int[] docs = new int[32]; // buffered doc numbers @@ -71,15 +73,16 @@ final class TermScorer extends Scorer { private static final int SCORE_CACHE_SIZE = 32; private float[] scoreCache = new float[SCORE_CACHE_SIZE]; - TermScorer(TermDocs td, Similarity similarity, byte[] norms, float weight) - throws IOException { + TermScorer(Weight weight, TermDocs td, Similarity similarity, + byte[] norms) throws IOException { super(similarity); + this.weight = weight; this.termDocs = td; this.norms = norms; - this.weight = weight; + this.weightValue = weight.getValue(); for (int i = 0; i < SCORE_CACHE_SIZE; i++) - scoreCache[i] = getSimilarity().tf(i) * weight; + scoreCache[i] = getSimilarity().tf(i) * weightValue; pointerMax = termDocs.read(docs, freqs); // fill buffers @@ -91,7 +94,7 @@ final class TermScorer extends Scorer { } } - final void score(HitCollector c, final int end) throws IOException { + public final void score(HitCollector c, final int end) throws IOException { int d = doc; // cache doc in local Similarity similarity = getSimilarity(); // cache sim in local while (d < end) { // for docs in window @@ -99,7 +102,7 @@ final class TermScorer extends Scorer { float score = // compute tf(f)*weight f < SCORE_CACHE_SIZE // check cache ? scoreCache[f] // cache hit - : similarity.tf(f)*weight; // cache miss + : similarity.tf(f)*weightValue; // cache miss score *= Similarity.decodeNorm(norms[d]); // normalize for field @@ -119,4 +122,45 @@ final class TermScorer extends Scorer { } doc = d; // flush cache } + + public Explanation explain(int doc) throws IOException { + Explanation result = new Explanation(); + TermQuery query = (TermQuery)weight.getQuery(); + + result.setDescription("termScore(" + query + "), product of:"); + + Explanation weightExplanation = weight.explain(); + result.addDetail(weightExplanation); + + Explanation tfExplanation = new Explanation(); + int tf = 0; + while (pointer < pointerMax) { + if (docs[pointer] == doc) + tf = freqs[pointer]; + pointer++; + } + if (tf == 0) { + while (termDocs.next()) { + if (termDocs.doc() == doc) { + tf = termDocs.freq(); + } + } + } + termDocs.close(); + tfExplanation.setValue(getSimilarity().tf(tf)); + tfExplanation.setDescription("tf(termFreq("+query.getTerm()+")="+tf+")"); + result.addDetail(tfExplanation); + + Explanation normExplanation = new Explanation(); + normExplanation.setValue(Similarity.decodeNorm(norms[doc])); + String field = query.getTerm().field(); + normExplanation.setDescription("norm(field="+field + ", doc="+doc + ")"); + result.addDetail(normExplanation); + + result.setValue(weightExplanation.getValue() * + tfExplanation.getValue() * + normExplanation.getValue()); + + return result; + } } diff --git a/src/java/org/apache/lucene/search/Weight.java b/src/java/org/apache/lucene/search/Weight.java new file mode 100644 index 00000000000..6cf58b13237 --- /dev/null +++ b/src/java/org/apache/lucene/search/Weight.java @@ -0,0 +1,88 @@ +package org.apache.lucene.search; + +/* ==================================================================== + * The Apache Software License, Version 1.1 + * + * Copyright (c) 2003 The Apache Software Foundation. All rights + * reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * 3. The end-user documentation included with the redistribution, + * if any, must include the following acknowledgment: + * "This product includes software developed by the + * Apache Software Foundation (http://www.apache.org/)." + * Alternately, this acknowledgment may appear in the software itself, + * if and wherever such third-party acknowledgments normally appear. + * + * 4. The names "Apache" and "Apache Software Foundation" and + * "Apache Lucene" must not be used to endorse or promote products + * derived from this software without prior written permission. For + * written permission, please contact apache@apache.org. + * + * 5. Products derived from this software may not be called "Apache", + * "Apache Lucene", nor may "Apache" appear in their name, without + * prior written permission of the Apache Software Foundation. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES + * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE + * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR + * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, + * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT + * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * ==================================================================== + * + * This software consists of voluntary contributions made by many + * individuals on behalf of the Apache Software Foundation. For more + * information on the Apache Software Foundation, please see + * . + */ + +import java.io.IOException; + +import org.apache.lucene.index.IndexReader; + +/** Expert: Calculate query weights and build query scorers. + * + *

    A Weight is constructed by a query, given a Searcher ({@link + * Query#createWeight(Searcher)}). The {@link #sumOfSquaredWeights()} method + * is then called on the top-level query to compute the query normalization + * factor (@link Similarity#queryNorm(float)}). This factor is then passed to + * {@link #normalize(float)}. At this point the weighting is complete and a + * scorer may be constructed by calling {@link #scorer(IndexReader)}. + */ +public interface Weight extends java.io.Serializable { + /** The query that this concerns. */ + Query getQuery(); + + /** The weight for this query. */ + float getValue(); + + /** The sum of squared weights of contained query clauses. */ + float sumOfSquaredWeights() throws IOException; + + /** Assigns the query normalization factor to this. */ + void normalize(float norm); + + /** Constructs a scorer for this. */ + Scorer scorer(IndexReader reader) throws IOException; + + /** An explanation of this weight computation. */ + Explanation explain() throws IOException; +} diff --git a/src/java/org/apache/lucene/search/WildcardQuery.java b/src/java/org/apache/lucene/search/WildcardQuery.java index 644ff6cdef0..f3418438dd7 100644 --- a/src/java/org/apache/lucene/search/WildcardQuery.java +++ b/src/java/org/apache/lucene/search/WildcardQuery.java @@ -60,17 +60,12 @@ import java.io.IOException; /** Implements the wildcard search query */ public class WildcardQuery extends MultiTermQuery { - private Term wildcardTerm; + public WildcardQuery(Term term) { + super(term); + } - public WildcardQuery(Term term) { - super(term); - wildcardTerm = term; - } - - final void prepare(IndexReader reader) { - try { - setEnum(new WildcardTermEnum(reader, wildcardTerm)); - } catch (IOException e) {} - } + protected FilteredTermEnum getEnum(IndexReader reader) throws IOException { + return new WildcardTermEnum(reader, getTerm()); + } } diff --git a/src/test/org/apache/lucene/queryParser/TestQueryParser.java b/src/test/org/apache/lucene/queryParser/TestQueryParser.java index 5435e7c2093..e6cdb3956d5 100644 --- a/src/test/org/apache/lucene/queryParser/TestQueryParser.java +++ b/src/test/org/apache/lucene/queryParser/TestQueryParser.java @@ -190,6 +190,8 @@ public class TestQueryParser extends TestCase { assertTrue(getQuery("\"hello there\"", null) instanceof PhraseQuery); assertQueryEquals("germ term^2.0", null, "germ term^2.0"); + assertQueryEquals("(term)^2.0", null, "term^2.0"); + assertQueryEquals("(germ term)^2.0", null, "(germ term)^2.0"); assertQueryEquals("term^2.0", null, "term^2.0"); assertQueryEquals("term^2", null, "term^2.0"); assertQueryEquals("\"germ term\"^2.0", null, "\"germ term\"^2.0");