beagled/Lucene.Net/Search/PhraseScorer.cs

   1 /*
   2  * Copyright 2004 The Apache Software Foundation
   3  *
   4  * Licensed under the Apache License, Version 2.0 (the "License");
   5  * you may not use this file except in compliance with the License.
   6  * You may obtain a copy of the License at
   7  *
   8  * http://www.apache.org/licenses/LICENSE-2.0
   9  *
  10  * Unless required by applicable law or agreed to in writing, software
  11  * distributed under the License is distributed on an "AS IS" BASIS,
  12  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13  * See the License for the specific language governing permissions and
  14  * limitations under the License.
  15  */
  16
  17 using System;
  18 using Lucene.Net.Index;
  19
  20 namespace Lucene.Net.Search
  21 {
  22
  23         abstract class PhraseScorer : Scorer
  24         {
  25                 private Weight weight;
  26                 protected internal byte[] norms;
  27                 protected internal float value_Renamed;
  28
  29                 private bool firstTime = true;
  30                 private bool more = true;
  31                 protected internal PhraseQueue pq;
  32                 protected internal PhrasePositions first, last;
  33
  34                 private float freq;
  35
  36
  37                 internal PhraseScorer(Weight weight, TermPositions[] tps, int[] positions, Similarity similarity, byte[] norms) : base(similarity)
  38                 {
  39                         this.norms = norms;
  40                         this.weight = weight;
  41                         this.value_Renamed = weight.GetValue();
  42
  43                         // convert tps to a list
  44                         for (int i = 0; i < tps.Length; i++)
  45                         {
  46                                 PhrasePositions pp = new PhrasePositions(tps[i], positions[i]);
  47                                 if (last != null)
  48                                 {
  49                                         // add next to end of list
  50                                         last.next = pp;
  51                                 }
  52                                 else
  53                                         first = pp;
  54                                 last = pp;
  55                         }
  56
  57                         pq = new PhraseQueue(tps.Length); // construct empty pq
  58                 }
  59
  60                 public override int Doc()
  61                 {
  62                         return first.doc;
  63                 }
  64
  65                 public override bool Next()
  66                 {
  67                         if (firstTime)
  68                         {
  69                                 Init();
  70                                 firstTime = false;
  71                         }
  72                         else if (more)
  73                         {
  74                                 more = last.Next(); // trigger further scanning
  75                         }
  76                         return DoNext();
  77                 }
  78
  79                 // next without initial increment
  80                 private bool DoNext()
  81                 {
  82                         while (more)
  83                         {
  84                                 while (more && first.doc < last.doc)
  85                                 {
  86                                         // find doc w/ all the terms
  87                                         more = first.SkipTo(last.doc); // skip first upto last
  88                                         FirstToLast(); // and move it to the end
  89                                 }
  90
  91                                 if (more)
  92                                 {
  93                                         // found a doc with all of the terms
  94                                         freq = PhraseFreq(); // check for phrase
  95                                         if (freq == 0.0f)
  96                                         // no match
  97                                                 more = last.Next();
  98                                         // trigger further scanning
  99                                         else
 100                                                 return true; // found a match
 101                                 }
 102                         }
 103                         return false; // no more matches
 104                 }
 105
 106                 public override float Score()
 107                 {
 108                         //System.out.println("scoring " + first.doc);
 109                         float raw = GetSimilarity().Tf(freq) * value_Renamed; // raw score
 110                         return raw * Similarity.DecodeNorm(norms[first.doc]); // normalize
 111                 }
 112
 113                 public override bool SkipTo(int target)
 114                 {
 115                         for (PhrasePositions pp = first; more && pp != null; pp = pp.next)
 116                         {
 117                                 more = pp.SkipTo(target);
 118                         }
 119                         if (more)
 120                                 Sort(); // re-sort
 121                         return DoNext();
 122                 }
 123
 124                 protected internal abstract float PhraseFreq();
 125
 126                 private void  Init()
 127                 {
 128                         for (PhrasePositions pp = first; more && pp != null; pp = pp.next)
 129                                 more = pp.Next();
 130                         if (more)
 131                                 Sort();
 132                 }
 133
 134                 private void  Sort()
 135                 {
 136                         pq.Clear();
 137                         for (PhrasePositions pp = first; pp != null; pp = pp.next)
 138                                 pq.Put(pp);
 139                         PqToList();
 140                 }
 141
 142                 protected internal void  PqToList()
 143                 {
 144                         last = first = null;
 145                         while (pq.Top() != null)
 146                         {
 147                                 PhrasePositions pp = (PhrasePositions) pq.Pop();
 148                                 if (last != null)
 149                                 {
 150                                         // add next to end of list
 151                                         last.next = pp;
 152                                 }
 153                                 else
 154                                         first = pp;
 155                                 last = pp;
 156                                 pp.next = null;
 157                         }
 158                 }
 159
 160                 protected internal void  FirstToLast()
 161                 {
 162                         last.next = first; // move first to end of list
 163                         last = first;
 164                         first = first.next;
 165                         last.next = null;
 166                 }
 167
 168                 public override Explanation Explain(int doc)
 169                 {
 170                         Explanation tfExplanation = new Explanation();
 171
 172                         while (Next() && Doc() < doc)
 173                         {
 174                         }
 175
 176                         float phraseFreq = (Doc() == doc)?freq:0.0f;
 177                         tfExplanation.SetValue(GetSimilarity().Tf(phraseFreq));
 178                         tfExplanation.SetDescription("tf(phraseFreq=" + phraseFreq + ")");
 179
 180                         return tfExplanation;
 181                 }
 182
 183                 public override System.String ToString()
 184                 {
 185                         return "scorer(" + weight + ")";
 186                 }
 187         }
 188 }