Examples of PhoneticEngine

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

62
63
64
65
66
67
68
69
70
71
72
    String ruleTypeArg = args.get("ruleType");
    RuleType ruleType = (ruleTypeArg == null) ? RuleType.APPROX : RuleType.valueOf(ruleTypeArg);
   
    boolean concat = getBoolean("concat", true);
    engine = new PhoneticEngine(nameType, ruleType, concat);
   
    // LanguageSet: defaults to automagic, otherwise a comma-separated list.
    String languageSetArg = args.get("languageSet");
    if (languageSetArg == null || languageSetArg.equals("auto")) {
      languageSet = null;
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

73
74
75
76
77
78
79
80
81
82
83
      @Override
      protected TokenStreamComponents createComponents(String fieldName, Reader reader) {
        Tokenizer tokenizer = new MockTokenizer(reader, MockTokenizer.WHITESPACE, false);
        return new TokenStreamComponents(tokenizer,
            new BeiderMorseFilter(tokenizer,
                new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true), languages));
      }
    };
    assertAnalyzesTo(analyzer, "Angelo",
        new String[] { "andZelo", "angelo", "anxelo" },
        new int[] { 0, 0, 0, },
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

101
102
103
104
105
106
107
108
109
110
  public void testEmptyTerm() throws IOException {
    Analyzer a = new Analyzer() {
      @Override
      protected TokenStreamComponents createComponents(String fieldName, Reader reader) {
        Tokenizer tokenizer = new KeywordTokenizer(reader);
        return new TokenStreamComponents(tokenizer, new BeiderMorseFilter(tokenizer, new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true)));
      }
    };
    checkOneTerm(a, "", "");
  }
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

110
111
112
113
114
115
116
117
118
119
120
  }
 
  public void testCustomAttribute() throws IOException {
    TokenStream stream = new MockTokenizer(new StringReader("D'Angelo"), MockTokenizer.KEYWORD, false);
    stream = new PatternKeywordMarkerFilter(stream, Pattern.compile(".*"));
    stream = new BeiderMorseFilter(stream, new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true));
    KeywordAttribute keyAtt = stream.addAttribute(KeywordAttribute.class);
    stream.reset();
    int i = 0;
    while(stream.incrementToken()) {
      assertTrue(keyAtt.isKeyword());
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

73
74
75
76
77
78
79
80
81
82
83
      @Override
      protected TokenStreamComponents createComponents(String fieldName, Reader reader) {
        Tokenizer tokenizer = new MockTokenizer(reader, MockTokenizer.WHITESPACE, false);
        return new TokenStreamComponents(tokenizer,
            new BeiderMorseFilter(tokenizer,
                new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true), languages));
      }
    };
    assertAnalyzesTo(analyzer, "Angelo",
        new String[] { "andZelo", "angelo", "anxelo" },
        new int[] { 0, 0, 0, },
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

101
102
103
104
105
106
107
108
109
110
  public void testEmptyTerm() throws IOException {
    Analyzer a = new Analyzer() {
      @Override
      protected TokenStreamComponents createComponents(String fieldName, Reader reader) {
        Tokenizer tokenizer = new KeywordTokenizer(reader);
        return new TokenStreamComponents(tokenizer, new BeiderMorseFilter(tokenizer, new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true)));
      }
    };
    checkOneTermReuse(a, "", "");
  }
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

110
111
112
113
114
115
116
117
118
119
120
  }
 
  public void testCustomAttribute() throws IOException {
    TokenStream stream = new KeywordTokenizer(new StringReader("D'Angelo"));
    stream = new PatternKeywordMarkerFilter(stream, Pattern.compile(".*"));
    stream = new BeiderMorseFilter(stream, new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true));
    KeywordAttribute keyAtt = stream.addAttribute(KeywordAttribute.class);
    stream.reset();
    int i = 0;
    while(stream.incrementToken()) {
      assertTrue(keyAtt.isKeyword());
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

68
69
70
71
72
73
74
75
76
77
78
      @Override
      protected TokenStreamComponents createComponents(String fieldName, Reader reader) {
        Tokenizer tokenizer = new MockTokenizer(reader, MockTokenizer.WHITESPACE, false);
        return new TokenStreamComponents(tokenizer,
            new BeiderMorseFilter(tokenizer,
                new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true), languages));
      }
    };
    assertAnalyzesTo(analyzer, "Angelo",
        new String[] { "andZelo", "angelo", "anxelo" },
        new int[] { 0, 0, 0, },
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

96
97
98
99
100
101
102
103
104
105
  public void testEmptyTerm() throws IOException {
    Analyzer a = new Analyzer() {
      @Override
      protected TokenStreamComponents createComponents(String fieldName, Reader reader) {
        Tokenizer tokenizer = new KeywordTokenizer(reader);
        return new TokenStreamComponents(tokenizer, new BeiderMorseFilter(tokenizer, new PhoneticEngine(NameType.GENERIC, RuleType.EXACT, true)));
      }
    };
    checkOneTermReuse(a, "", "");
  }
View Full Code Here

Examples of org.apache.commons.codec.language.bm.PhoneticEngine

51
52
53
54
55
56
57
58
59
60
61
    // we use common-codec's defaults: GENERIC + APPROX + true
    NameType nameType = NameType.valueOf(get(args, "nameType", NameType.GENERIC.toString()));
    RuleType ruleType = RuleType.valueOf(get(args, "ruleType", RuleType.APPROX.toString()));
   
    boolean concat = getBoolean(args, "concat", true);
    engine = new PhoneticEngine(nameType, ruleType, concat);
   
    // LanguageSet: defaults to automagic, otherwise a comma-separated list.
    Set<String> langs = getSet(args, "languageSet");
    languageSet = (null == langs || (1 == langs.size() && langs.contains("auto"))) ? null : LanguageSet.from(langs);
    if (!args.isEmpty()) {
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. 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.