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 using System;
  17 using Lucene.Net.Index;
  18 namespace Lucene.Net.Search
  19 {
  20
  21         abstract class PhraseScorer:Scorer
  22         {
  23                 private Weight weight;
  24                 protected internal byte[] norms;
  25                 protected internal float value_Renamed;
  26
  27                 private bool firstTime = true;
  28                 private bool more = true;
  29                 protected internal PhraseQueue pq;
  30                 protected internal PhrasePositions first, last;
  31
  32                 private float freq;
  33
  34
  35                 internal PhraseScorer(Weight weight, TermPositions[] tps, int[] positions, Similarity similarity, byte[] norms) : base(similarity)
  36                 {
  37                         this.norms = norms;
  38                         this.weight = weight;
  39                         this.value_Renamed = weight.GetValue();
  40
  41                         // convert tps to a list
  42                         for (int i = 0; i < tps.Length; i++)
  43                         {
  44                                 PhrasePositions pp = new PhrasePositions(tps[i], positions[i]);
  45                                 if (last != null)
  46                                 {
  47                                         // add next to end of list
  48                                         last.next = pp;
  49                                 }
  50                                 else
  51                                         first = pp;
  52                                 last = pp;
  53                         }
  54
  55                         pq = new PhraseQueue(tps.Length); // construct empty pq
  56                 }
  57
  58                 public override int Doc()
  59                 {
  60                         return first.doc;
  61                 }
  62
  63                 public override bool Next()
  64                 {
  65                         if (firstTime)
  66                         {
  67                                 Init();
  68                                 firstTime = false;
  69                         }
  70                         else if (more)
  71                         {
  72                                 more = last.Next(); // trigger further scanning
  73                         }
  74                         return DoNext();
  75                 }
  76
  77                 // next without initial increment
  78                 private bool DoNext()
  79                 {
  80                         while (more)
  81                         {
  82                                 while (more && first.doc < last.doc)
  83                                 {
  84                                         // find doc w/ all the terms
  85                                         more = first.SkipTo(last.doc); // skip first upto last
  86                                         FirstToLast(); // and move it to the end
  87                                 }
  88
  89                                 if (more)
  90                                 {
  91                                         // found a doc with all of the terms
  92                                         freq = PhraseFreq(); // check for phrase
  93                                         if (freq == 0.0f)
  94                                         // no match
  95                                                 more = last.Next();
  96                                         // trigger further scanning
  97                                         else
  98                                                 return true; // found a match
  99                                 }
 100                         }
 101                         return false; // no more matches
 102                 }
 103
 104                 public override float Score()
 105                 {
 106                         //System.out.println("scoring " + first.doc);
 107                         float raw = GetSimilarity().Tf(freq) * value_Renamed; // raw score
 108                         return raw * Similarity.DecodeNorm(norms[first.doc]); // normalize
 109                 }
 110
 111                 public override bool SkipTo(int target)
 112                 {
 113                         for (PhrasePositions pp = first; more && pp != null; pp = pp.next)
 114                         {
 115                                 more = pp.SkipTo(target);
 116                         }
 117                         if (more)
 118                                 Sort(); // re-sort
 119                         return DoNext();
 120                 }
 121
 122                 protected internal abstract float PhraseFreq();
 123
 124                 private void  Init()
 125                 {
 126                         for (PhrasePositions pp = first; more && pp != null; pp = pp.next)
 127                                 more = pp.Next();
 128                         if (more)
 129                                 Sort();
 130                 }
 131
 132                 private void  Sort()
 133                 {
 134                         pq.Clear();
 135                         for (PhrasePositions pp = first; pp != null; pp = pp.next)
 136                                 pq.Put(pp);
 137                         PqToList();
 138                 }
 139
 140                 protected internal void  PqToList()
 141                 {
 142                         last = first = null;
 143                         while (pq.Top() != null)
 144                         {
 145                                 PhrasePositions pp = (PhrasePositions) pq.Pop();
 146                                 if (last != null)
 147                                 {
 148                                         // add next to end of list
 149                                         last.next = pp;
 150                                 }
 151                                 else
 152                                         first = pp;
 153                                 last = pp;
 154                                 pp.next = null;
 155                         }
 156                 }
 157
 158                 protected internal void  FirstToLast()
 159                 {
 160                         last.next = first; // move first to end of list
 161                         last = first;
 162                         first = first.next;
 163                         last.next = null;
 164                 }
 165
 166                 public override Explanation Explain(int doc)
 167                 {
 168                         Explanation tfExplanation = new Explanation();
 169
 170                         while (Next() && Doc() < doc)
 171                         {
 172                         }
 173
 174                         float phraseFreq = (Doc() == doc)?freq:0.0f;
 175                         tfExplanation.SetValue(GetSimilarity().Tf(phraseFreq));
 176                         tfExplanation.SetDescription("tf(phraseFreq=" + phraseFreq + ")");
 177
 178                         return tfExplanation;
 179                 }
 180
 181                 public override System.String ToString()
 182                 {
 183                         return "scorer(" + weight + ")";
 184                 }
 185         }
 186 }