LUCENE-3328: specialize ConjunctionScorer if all required clauses are TermQueries

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1149547 13f79535-47bb-0310-9956-ffa450edef68

lucene/CHANGES.txt

@@ -494,6 +494,10 @@ Optimizations
   instance for merging and NRT readers, which enables directory impls
   to separately tune IO flags for each.  (Varun Thacker, Simon
   Willnauer, Mike McCandless)
+
+* LUCENE-3328: BooleanQuery now uses a specialized ConjunctionScorer if all
+  boolean clauses are required and instances of TermQuery. 
+  (Simon Willnauer, Robert Muir)
   
 Bug fixes
 

lucene/src/java/org/apache/lucene/search/BooleanQuery.java

@@ -18,10 +18,14 @@ package org.apache.lucene.search;
  */
 
 import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.index.TermsEnum;
 import org.apache.lucene.index.IndexReader.AtomicReaderContext;
 import org.apache.lucene.index.Term;
 import org.apache.lucene.util.ToStringUtils;
 import org.apache.lucene.search.BooleanClause.Occur;
+import org.apache.lucene.search.ConjunctionTermScorer.DocsAndFreqs;
+import org.apache.lucene.search.Similarity.ExactDocScorer;
+import org.apache.lucene.search.TermQuery.TermWeight;
 
 import java.io.IOException;
 import java.util.*;
@@ -166,17 +170,24 @@ public class BooleanQuery extends Query implements Iterable<BooleanClause> {
     protected ArrayList<Weight> weights;
     protected int maxCoord;  // num optional + num required
     private final boolean disableCoord;
+    private final boolean termConjunction;
 
     public BooleanWeight(IndexSearcher searcher, boolean disableCoord)
       throws IOException {
       this.similarityProvider = searcher.getSimilarityProvider();
       this.disableCoord = disableCoord;
       weights = new ArrayList<Weight>(clauses.size());
+      boolean termConjunction = clauses.isEmpty() || minNrShouldMatch != 0 ? false : true;
       for (int i = 0 ; i < clauses.size(); i++) {
         BooleanClause c = clauses.get(i);
-        weights.add(c.getQuery().createWeight(searcher));
+        Weight w = c.getQuery().createWeight(searcher);
+        if (!(c.isRequired() && (w instanceof TermWeight))) {
+          termConjunction = false;
+        }
+        weights.add(w);
         if (!c.isProhibited()) maxCoord++;
       }
+      this.termConjunction = termConjunction;
     }
 
     @Override
@@ -290,6 +301,10 @@ public class BooleanQuery extends Query implements Iterable<BooleanClause> {
     @Override
     public Scorer scorer(AtomicReaderContext context, ScorerContext scorerContext)
         throws IOException {
+      if (termConjunction) {
+        // specialized scorer for term conjunctions
+        return createConjunctionTermScorer(context);
+      }
       List<Scorer> required = new ArrayList<Scorer>();
       List<Scorer> prohibited = new ArrayList<Scorer>();
       List<Scorer> optional = new ArrayList<Scorer>();
@@ -328,6 +343,23 @@ public class BooleanQuery extends Query implements Iterable<BooleanClause> {
       // Return a BooleanScorer2
       return new BooleanScorer2(this, disableCoord, minNrShouldMatch, required, prohibited, optional, maxCoord);
     }
+
+    private Scorer createConjunctionTermScorer(AtomicReaderContext context)
+        throws IOException {
+      final DocsAndFreqs[] docsAndFreqs = new DocsAndFreqs[weights.size()];
+      for (int i = 0; i < docsAndFreqs.length; i++) {
+        final TermWeight weight = (TermWeight) weights.get(i);
+        final TermsEnum termsEnum = weight.getTermsEnum(context);
+        if (termsEnum == null) {
+          return null;
+        }
+        final ExactDocScorer docScorer = weight.createDocScorer(context);
+        docsAndFreqs[i] = new DocsAndFreqs(termsEnum.docs(
+            context.reader.getLiveDocs(), null), termsEnum.docFreq(), docScorer);
+      }
+      return new ConjunctionTermScorer(this, disableCoord ? 1.0f : coord(
+          docsAndFreqs.length, docsAndFreqs.length), docsAndFreqs);
+    }
     
     @Override
     public boolean scoresDocsOutOfOrder() {

lucene/src/java/org/apache/lucene/search/ConjunctionTermScorer.java

@@ -0,0 +1,110 @@
+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.index.DocsEnum;
+import org.apache.lucene.search.Similarity.ExactDocScorer;
+import org.apache.lucene.util.ArrayUtil;
+import java.io.IOException;
+import java.util.Comparator;
+
+/** Scorer for conjunctions, sets of terms, all of which are required. */
+final class ConjunctionTermScorer extends Scorer {
+  private final float coord;
+  private int lastDoc = -1;
+  private final DocsAndFreqs[] docsAndFreqs;
+  private final DocsAndFreqs lead;
+
+  ConjunctionTermScorer(Weight weight, float coord,
+      DocsAndFreqs[] docsAndFreqs) throws IOException {
+    super(weight);
+    this.coord = coord;
+    this.docsAndFreqs = docsAndFreqs;
+    // Sort the array the first time to allow the least frequent DocsEnum to
+    // lead the matching.
+    ArrayUtil.mergeSort(docsAndFreqs, new Comparator<DocsAndFreqs>() {
+      public int compare(DocsAndFreqs o1, DocsAndFreqs o2) {
+        return o1.freq - o2.freq;
+      }
+    });
+
+    lead = docsAndFreqs[0]; // least frequent DocsEnum leads the intersection
+  }
+
+  private int doNext(int doc) throws IOException {
+    do {
+      if (lead.doc == DocsEnum.NO_MORE_DOCS) {
+        return NO_MORE_DOCS;
+      }
+      advanceHead: do {
+        for (int i = 1; i < docsAndFreqs.length; i++) {
+          if (docsAndFreqs[i].doc < doc) {
+            docsAndFreqs[i].doc = docsAndFreqs[i].docs.advance(doc);
+          }
+          if (docsAndFreqs[i].doc > doc) {
+            // DocsEnum beyond the current doc - break and advance lead
+            break advanceHead;
+          }
+        }
+        // success - all DocsEnums are on the same doc
+        return doc;
+      } while (true);
+      // advance head for next iteration
+      doc = lead.doc = lead.docs.nextDoc();  
+    } while (true);
+  }
+
+  @Override
+  public int advance(int target) throws IOException {
+    lead.doc = lead.docs.advance(target);
+    return lastDoc = doNext(lead.doc);
+  }
+
+  @Override
+  public int docID() {
+    return lastDoc;
+  }
+
+  @Override
+  public int nextDoc() throws IOException {
+    lead.doc = lead.docs.nextDoc();
+    return lastDoc = doNext(lead.doc);
+  }
+
+  @Override
+  public float score() throws IOException {
+    float sum = 0.0f;
+    for (DocsAndFreqs docs : docsAndFreqs) {
+      sum += docs.docScorer.score(lastDoc, docs.docs.freq());
+    }
+    return sum * coord;
+  }
+
+  static final class DocsAndFreqs {
+    final DocsEnum docs;
+    final int freq;
+    final ExactDocScorer docScorer;
+    int doc = -1;
+
+    DocsAndFreqs(DocsEnum docs, int freq, ExactDocScorer docScorer) {
+      this.docs = docs;
+      this.freq = freq;
+      this.docScorer = docScorer;
+    }
+  }
+}

lucene/src/java/org/apache/lucene/search/TermQuery.java

@@ -24,6 +24,7 @@ import org.apache.lucene.index.DocsEnum;
 import org.apache.lucene.index.IndexReader;
 import org.apache.lucene.index.TermState;
 import org.apache.lucene.index.Terms;
+import org.apache.lucene.index.TermsEnum;
 import org.apache.lucene.index.IndexReader.AtomicReaderContext;
 import org.apache.lucene.index.IndexReader.ReaderContext;
 import org.apache.lucene.index.Term;
@@ -41,7 +42,7 @@ public class TermQuery extends Query {
   private int docFreq;
   private transient TermContext perReaderTermState;
 
-  private class TermWeight extends Weight {
+  final class TermWeight extends Weight {
     private final Similarity similarity;
     private final Similarity.Stats stats;
     private transient TermContext termStates;
@@ -72,17 +73,38 @@ public class TermQuery extends Query {
 
     @Override
     public Scorer scorer(AtomicReaderContext context, ScorerContext scorerContext) throws IOException {
-      final String field = term.field();
-      final IndexReader reader = context.reader;
       assert termStates.topReaderContext == ReaderUtil.getTopLevelContext(context) : "The top-reader used to create Weight (" + termStates.topReaderContext + ") is not the same as the current reader's top-reader (" + ReaderUtil.getTopLevelContext(context);
-      final TermState state = termStates.get(context.ord);
-      if (state == null) { // term is not present in that reader
-        assert termNotInReader(reader, field, term.bytes()) : "no termstate found but term exists in reader";
+      final TermsEnum termsEnum = getTermsEnum(context);
+      if (termsEnum == null) {
         return null;
       }
-      final DocsEnum docs = reader.termDocsEnum(reader.getLiveDocs(), field, term.bytes(), state);
+      // TODO should we reuse the DocsEnum here? 
+      final DocsEnum docs = termsEnum.docs(context.reader.getLiveDocs(), null);
       assert docs != null;
-      return new TermScorer(this, docs, similarity.exactDocScorer(stats, field, context));
+      return new TermScorer(this, docs, createDocScorer(context));
+    }
+    
+    /**
+     * Creates an {@link ExactDocScorer} for this {@link TermWeight}*/
+    ExactDocScorer createDocScorer(AtomicReaderContext context)
+        throws IOException {
+      return similarity.exactDocScorer(stats, term.field(), context);
+    }
+    
+    /**
+     * Returns a {@link TermsEnum} positioned at this weights Term or null if
+     * the term does not exist in the given context
+     */
+    TermsEnum getTermsEnum(AtomicReaderContext context) throws IOException {
+      final TermState state = termStates.get(context.ord);
+      if (state == null) { // term is not present in that reader
+        assert termNotInReader(context.reader, term.field(), term.bytes()) : "no termstate found but term exists in reader";
+        return null;
+      }
+      final TermsEnum termsEnum = context.reader.terms(term.field())
+          .getThreadTermsEnum();
+      termsEnum.seekExact(term.bytes(), state);
+      return termsEnum;
     }
     
     private boolean termNotInReader(IndexReader reader, String field, BytesRef bytes) throws IOException {