beagled/Lucene.Net/Search/QueryTermVector.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 Analyzer = Lucene.Net.Analysis.Analyzer;
  18 using Token = Lucene.Net.Analysis.Token;
  19 using TokenStream = Lucene.Net.Analysis.TokenStream;
  20 using TermFreqVector = Lucene.Net.Index.TermFreqVector;
  21 namespace Lucene.Net.Search
  22 {
  23
  24         /// <summary>
  25         ///
  26         ///
  27         /// </summary>
  28         public class QueryTermVector : TermFreqVector
  29         {
  30                 private System.String[] terms = new System.String[0];
  31                 private int[] termFreqs = new int[0];
  32
  33                 public virtual System.String GetField()
  34                 {
  35                         return null;
  36                 }
  37
  38                 /// <summary> </summary>
  39                 /// <param name="queryTerms">The original list of terms from the query, can contain duplicates
  40                 /// </param>
  41                 public QueryTermVector(System.String[] queryTerms)
  42                 {
  43
  44                         ProcessTerms(queryTerms);
  45                 }
  46
  47                 public QueryTermVector(System.String queryString, Analyzer analyzer)
  48                 {
  49                         if (analyzer != null)
  50                         {
  51                                 TokenStream stream = analyzer.TokenStream("", new System.IO.StringReader(queryString));
  52                                 if (stream != null)
  53                                 {
  54                                         Token next = null;
  55                                         System.Collections.ArrayList terms = new System.Collections.ArrayList();
  56                                         try
  57                                         {
  58                                                 while ((next = stream.Next()) != null)
  59                                                 {
  60                                                         terms.Add(next.TermText());
  61                                                 }
  62                                                 ProcessTerms((System.String[]) terms.ToArray(typeof(System.String)));
  63                                         }
  64                                         catch (System.IO.IOException)
  65                                         {
  66                                         }
  67                                 }
  68                         }
  69                 }
  70
  71                 private void  ProcessTerms(System.String[] queryTerms)
  72                 {
  73                         if (queryTerms != null)
  74                         {
  75                                 System.Array.Sort(queryTerms);
  76                                 System.Collections.IDictionary tmpSet = new System.Collections.Hashtable(queryTerms.Length);
  77                                 //filter out duplicates
  78                                 System.Collections.ArrayList tmpList = new System.Collections.ArrayList(queryTerms.Length);
  79                                 System.Collections.ArrayList tmpFreqs = new System.Collections.ArrayList(queryTerms.Length);
  80                                 int j = 0;
  81                                 for (int i = 0; i < queryTerms.Length; i++)
  82                                 {
  83                                         System.String term = queryTerms[i];
  84                     System.Object tmpPosition = tmpSet[term];
  85                                         if (tmpPosition == null)
  86                                         {
  87                                                 tmpSet[term] = (System.Int32) j++;
  88                                                 tmpList.Add(term);
  89                                                 tmpFreqs.Add(1);
  90                                         }
  91                                         else
  92                                         {
  93                         System.Int32 position = (System.Int32) tmpSet[term];
  94                                                 System.Int32 integer = (System.Int32) tmpFreqs[position];
  95                                                 tmpFreqs[position] = (System.Int32) (integer + 1);
  96                                         }
  97                                 }
  98                 terms = (System.String[]) tmpList.ToArray(typeof(System.String));
  99                                 //termFreqs = (int[])tmpFreqs.toArray(termFreqs);
 100                                 termFreqs = new int[tmpFreqs.Count];
 101                                 int i2 = 0;
 102                                 for (System.Collections.IEnumerator iter = tmpFreqs.GetEnumerator(); iter.MoveNext(); )
 103                                 {
 104                                         System.Int32 integer = (System.Int32) iter.Current;
 105                                         termFreqs[i2++] = integer;
 106                                 }
 107                         }
 108                 }
 109
 110                 public override System.String ToString()
 111                 {
 112                         System.Text.StringBuilder sb = new System.Text.StringBuilder();
 113                         sb.Append('{');
 114                         for (int i = 0; i < terms.Length; i++)
 115                         {
 116                                 if (i > 0)
 117                                         sb.Append(", ");
 118                                 sb.Append(terms[i]).Append('/').Append(termFreqs[i]);
 119                         }
 120                         sb.Append('}');
 121                         return sb.ToString();
 122                 }
 123
 124
 125                 public virtual int Size()
 126                 {
 127                         return terms.Length;
 128                 }
 129
 130                 public virtual System.String[] GetTerms()
 131                 {
 132                         return terms;
 133                 }
 134
 135                 public virtual int[] GetTermFrequencies()
 136                 {
 137                         return termFreqs;
 138                 }
 139
 140                 public virtual int IndexOf(System.String term)
 141                 {
 142                         int res = System.Array.BinarySearch(terms, term);
 143                         return res >= 0?res:- 1;
 144                 }
 145
 146                 public virtual int[] IndexesOf(System.String[] terms, int start, int len)
 147                 {
 148                         int[] res = new int[len];
 149
 150                         for (int i = 0; i < len; i++)
 151                         {
 152                                 res[i] = IndexOf(terms[i]);
 153                         }
 154                         return res;
 155                 }
 156         }
 157 }