Package net.paoding.analysis.dictionary

Examples of net.paoding.analysis.dictionary.Word

83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
        break;
      }
    }
    int index = wordText.indexOf('[');
    if (index == -1) {
      words.add(new Word(wordText));
    }
    else {
      Word w = new Word(wordText.substring(0, index));
      int mindex = wordText.indexOf("m=", index);
      if (mindex != -1) {
        int mEndIndex = wordText.indexOf("]", mindex);
        String m = wordText.substring(mindex + "m=".length(), mEndIndex);
        w.setModifiers(Integer.parseInt(m));
        words.add(w);
      }
    }
  }
View Full Code Here
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
    }
    BufferedOutputStream out = new BufferedOutputStream(
        new FileOutputStream(dicFile), 1024 * 16);
   
    for (int i = 0; i < wordsSize; i++) {
      Word word = dictionary.get(i);
      out.write(word.getText().getBytes(charsetName));
      if (word.getModifiers() != Word.DEFAUL) {
        out.write("[m=".getBytes());
        out.write(String.valueOf(word.getModifiers()).getBytes());
        out.write(']');
      }
      out.write('\r');
      out.write('\n');
    }
View Full Code Here
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
      IOException, UnsupportedEncodingException {
    int vocabularySize = vocabularyDictionary.size();
    Word[] vocabularyWords = new Word[vocabularySize];
    char[] chs = new char[128];
    for (int i = 0; i < vocabularySize; i ++) {
      final Word curWord = vocabularyDictionary.get(i);
      curWord.getText().getChars(0, curWord.length(), chs, 0);
      chs[curWord.length()] = (char) -1;
      Beef beef = new Beef(chs, 0, curWord.length() + 1);
      final BitSet bs = new BitSet(curWord.length());
      knife.dissect(new Collector(){
        public void collect(String word, int offset, int end) {
          Hit hit = vocabularyDictionary.search(word, 0, word.length());
          if (hit.isHit() && hit.getWord().length() != curWord.length()) {
            for (int j = offset; j < end; j++) {
              bs.set(j, true);
            }
          }
        }
       
      }, beef, 0);
     
      for (int j = 0; j < curWord.length();j++) {
        if (!bs.get(j)) {
          vocabularyWords[i] = curWord;
          break;
        }
      }
View Full Code Here
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
    }
    BufferedOutputStream out = new BufferedOutputStream(
        new FileOutputStream(dicFile), 1024 * 16);
   
    for (int i = 0; i < wordsSize; i++) {
      Word word = dictionary.get(i);
      out.write(word.getText().getBytes(charsetName));
      if (word.getModifiers() != Word.DEFAUL) {
        out.write("[m=".getBytes());
        out.write(String.valueOf(word.getModifiers()).getBytes());
        out.write(']');
      }
      out.write('\r');
      out.write('\n');
    }
View Full Code Here
TOP

Related Classes of net.paoding.analysis.dictionary.Word

  • net.paoding.analysis.analyzer.impl.MostWordsModeDictionariesCompiler
  • net.paoding.analysis.analyzer.impl.SortingDictionariesCompiler
  • net.paoding.analysis.dictionary.support.filewords.SimpleReadListener
  • net.paoding.analysis.dictionary.support.filewords.SimpleReadListener2
  • net.paoding.analysis.knife.CJKKnife
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.