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
  17 using System;
  18 using Analyzer = Lucene.Net.Analysis.Analyzer;
  19 using Token = Lucene.Net.Analysis.Token;
  20 using TokenStream = Lucene.Net.Analysis.TokenStream;
  21 using TermFreqVector = Lucene.Net.Index.TermFreqVector;
  22
  23 namespace Lucene.Net.Search
  24 {
  25
  26         /// <summary>
  27         ///
  28         ///
  29         /// </summary>
  30         public class QueryTermVector : TermFreqVector
  31         {
  32                 private System.String[] terms = new System.String[0];
  33                 private int[] termFreqs = new int[0];
  34
  35                 public virtual System.String GetField()
  36                 {
  37                         return null;
  38                 }
  39
  40                 /// <summary> </summary>
  41                 /// <param name="queryTerms">The original list of terms from the query, can contain duplicates
  42                 /// </param>
  43                 public QueryTermVector(System.String[] queryTerms)
  44                 {
  45
  46                         ProcessTerms(queryTerms);
  47                 }
  48
  49                 public QueryTermVector(System.String queryString, Analyzer analyzer)
  50                 {
  51                         if (analyzer != null)
  52                         {
  53                                 TokenStream stream = analyzer.TokenStream("", new System.IO.StringReader(queryString));
  54                                 if (stream != null)
  55                                 {
  56                                         Token next = null;
  57                                         System.Collections.ArrayList terms = new System.Collections.ArrayList();
  58                                         try
  59                                         {
  60                                                 while ((next = stream.Next()) != null)
  61                                                 {
  62                                                         terms.Add(next.TermText());
  63                                                 }
  64                                                 ProcessTerms((System.String[]) terms.ToArray(typeof(System.String)));
  65                                         }
  66                                         catch (System.IO.IOException)
  67                                         {
  68                                         }
  69                                 }
  70                         }
  71                 }
  72
  73                 private void  ProcessTerms(System.String[] queryTerms)
  74                 {
  75                         if (queryTerms != null)
  76                         {
  77                                 System.Array.Sort(queryTerms);
  78                                 System.Collections.IDictionary tmpSet = new System.Collections.Hashtable(queryTerms.Length);
  79                                 //filter out duplicates
  80                                 System.Collections.ArrayList tmpList = new System.Collections.ArrayList(queryTerms.Length);
  81                                 System.Collections.ArrayList tmpFreqs = new System.Collections.ArrayList(queryTerms.Length);
  82                                 int j = 0;
  83                                 for (int i = 0; i < queryTerms.Length; i++)
  84                                 {
  85                                         System.String term = queryTerms[i];
  86                                         System.Object tmpPosition = tmpSet[term];
  87                                         if (tmpPosition == null)
  88                                         {
  89                                                 tmpSet[term] = (System.Int32) j++;
  90                                                 tmpList.Add(term);
  91                                                 tmpFreqs.Add(1);
  92                                         }
  93                                         else
  94                                         {
  95                         System.Int32 position = (System.Int32) tmpSet[term];
  96                                                 System.Int32 integer = (System.Int32) tmpFreqs[position];
  97                                                 tmpFreqs[position] = (System.Int32) (integer + 1);
  98                                         }
  99                                 }
 100                                 terms = (System.String[]) tmpList.ToArray(typeof(System.String));
 101                                 //termFreqs = (int[])tmpFreqs.toArray(termFreqs);
 102                                 termFreqs = new int[tmpFreqs.Count];
 103                                 int i2 = 0;
 104                                 for (System.Collections.IEnumerator iter = tmpFreqs.GetEnumerator(); iter.MoveNext(); )
 105                                 {
 106                                         System.Int32 integer = (System.Int32) iter.Current;
 107                                         termFreqs[i2++] = integer;
 108                                 }
 109                         }
 110                 }
 111
 112                 public override System.String ToString()
 113                 {
 114                         System.Text.StringBuilder sb = new System.Text.StringBuilder();
 115                         sb.Append('{');
 116                         for (int i = 0; i < terms.Length; i++)
 117                         {
 118                                 if (i > 0)
 119                                         sb.Append(", ");
 120                                 sb.Append(terms[i]).Append('/').Append(termFreqs[i]);
 121                         }
 122                         sb.Append('}');
 123                         return sb.ToString();
 124                 }
 125
 126
 127                 public virtual int Size()
 128                 {
 129                         return terms.Length;
 130                 }
 131
 132                 public virtual System.String[] GetTerms()
 133                 {
 134                         return terms;
 135                 }
 136
 137                 public virtual int[] GetTermFrequencies()
 138                 {
 139                         return termFreqs;
 140                 }
 141
 142                 public virtual int IndexOf(System.String term)
 143                 {
 144                         int res = System.Array.BinarySearch(terms, term);
 145                         return res >= 0?res:- 1;
 146                 }
 147
 148                 public virtual int[] IndexesOf(System.String[] terms, int start, int len)
 149                 {
 150                         int[] res = new int[len];
 151
 152                         for (int i = 0; i < len; i++)
 153                         {
 154                                 res[i] = IndexOf(terms[i]);
 155                         }
 156                         return res;
 157                 }
 158         }
 159 }