Package com.ibm.icu.text

Examples of com.ibm.icu.text.RuleBasedBreakIterator

364
365
366
367
368
369
370
371
372
373
374
375
376
377
    /**
     * Tests the method IsBoundary() of RuleBasedBreakIterator
     **/
    public void TestIsBoundary() {
        String testString1 = "Write here. \u092d\u0301\u0930\u0924 \u0938\u0941\u0902\u0926\u0930 a\u0301u";
        RuleBasedBreakIterator charIter1 = (RuleBasedBreakIterator) BreakIterator.getCharacterInstance(Locale.getDefault());
        charIter1.setText(testString1);
        int bounds1[] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 20, 21, 22, 23, 25, 26};
        doBoundaryTest(charIter1, testString1, bounds1);
        RuleBasedBreakIterator wordIter2 = (RuleBasedBreakIterator) BreakIterator.getWordInstance(Locale.getDefault());
        wordIter2.setText(testString1);
        int bounds2[] = {0, 5, 6, 10, 11, 12, 16, 17, 22, 23, 26};
        doBoundaryTest(wordIter2, testString1, bounds2);
    }
View Full Code Here
1915
1916
1917
1918
1919
1920
1921
1922
    }
   
    RBBICharMonkey  m = new RBBICharMonkey();
    BreakIterator   bi = BreakIterator.getCharacterInstance(Locale.US);
    String rules = bi.toString();
    BreakIterator rtbi = new RuleBasedBreakIterator(rules);
    RunMonkey(rtbi, m, "char", seed, loopCount);
}
View Full Code Here
1932
1933
1934
1935
1936
1937
1938
1939
   
    logln("Word Break Monkey Test");
    RBBIWordMonkey  m = new RBBIWordMonkey();
    BreakIterator   bi = BreakIterator.getWordInstance(Locale.US);
    String rules = bi.toString();
    BreakIterator rtbi = new RuleBasedBreakIterator(rules);
    RunMonkey(rtbi, m, "word", seed, loopCount);
}
View Full Code Here
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
   
    logln("Line Break Monkey Test");
    RBBILineMonkey  m = new RBBILineMonkey();
    BreakIterator   bi = BreakIterator.getLineInstance(Locale.US);
    String rules = bi.toString();
    BreakIterator rtbi = new RuleBasedBreakIterator(rules);
    if (params == null) {
        loopCount = 50;
    }
    RunMonkey(rtbi, m, "line", seed, loopCount);
}
View Full Code Here
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
   
    logln("Sentence Break Monkey Test");
    RBBISentenceMonkey  m = new RBBISentenceMonkey();
    BreakIterator   bi = BreakIterator.getSentenceInstance(Locale.US);
    String rules = bi.toString();
    BreakIterator rtbi = new RuleBasedBreakIterator(rules);
    if (params == null) {
        loopCount = 30;
    }
    RunMonkey(rtbi, m, "sent", seed, loopCount);
}
View Full Code Here
67
68
69
70
71
72
73
74
75
76
77
       * if there is a syntax error, compileRules() may succeed. the way to
       * check is to try to instantiate from the string. additionally if the
       * rules are invalid, you can get a useful syntax error.
       */
      try {
        new RuleBasedBreakIterator(rules);
      } catch (IllegalArgumentException e) {
        /*
         * do this intentionally, so you don't get a massive stack trace
         * instead, get a useful syntax error!
         */
 
View Full Code Here
118
119
120
121
122
123
124
125
126
127
128
  private static RuleBasedBreakIterator readBreakIterator(String filename) {
    InputStream is =
      DefaultICUTokenizerConfig.class.getResourceAsStream(filename);
    try {
      RuleBasedBreakIterator bi =
        RuleBasedBreakIterator.getInstanceFromCompiledRules(is);
      is.close();
      return bi;
    } catch (IOException e) {
      throw new RuntimeException(e);
View Full Code Here
139
140
141
142
143
144
145
      if ( ! line.startsWith("#"))
        rules.append(line);
      rules.append('\n');
    }
    reader.close();
    return new RuleBasedBreakIterator(rules.toString());
  }
View Full Code Here
68
69
70
71
72
73
74
75
76
77
78
       * if there is a syntax error, compileRules() may succeed. the way to
       * check is to try to instantiate from the string. additionally if the
       * rules are invalid, you can get a useful syntax error.
       */
      try {
        new RuleBasedBreakIterator(rules);
      } catch (IllegalArgumentException e) {
        /*
         * do this intentionally, so you don't get a massive stack trace
         * instead, get a useful syntax error!
         */
 
View Full Code Here
105
106
107
108
109
110
111
112
113
114
115
  private static RuleBasedBreakIterator readBreakIterator(String filename) {
    InputStream is =
      DefaultICUTokenizerConfig.class.getResourceAsStream(filename);
    try {
      RuleBasedBreakIterator bi =
        RuleBasedBreakIterator.getInstanceFromCompiledRules(is);
      is.close();
      return bi;
    } catch (IOException e) {
      throw new RuntimeException(e);
View Full Code Here
TOP

Related Classes of com.ibm.icu.text.RuleBasedBreakIterator

  • com.ibm.icu.dev.test.lang.UCharacterCaseTest
  • com.ibm.icu.dev.test.rbbi.RBBIAPITest
  • com.ibm.icu.dev.test.rbbi.RBBITest
  • com.ibm.icu.dev.test.rbbi.RBBITestMonkey
  • com.ibm.icu.dev.test.rbbi.SimpleBITest
  • com.ibm.icu.impl.CharTrie
  • org.apache.lucene.analysis.icu.RBBIRuleCompiler
  • org.apache.lucene.analysis.icu.segmentation.DefaultICUTokenizerConfig
  • org.apache.lucene.analysis.icu.segmentation.ICUTokenizerFactory
  • java.io.InputStream
    • 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.