Examples of SnowballFilter

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

84
85
86
87
88
89
90
      throw new RuntimeException("Error instantiating stemmer for language " + language + "from class " +stemClass, e);
    }

    if (protectedWords != null)
      input = new KeywordMarkerFilter(input, protectedWords);
    return new SnowballFilter(input, program);
  }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

143
144
145
146
147
148
149
150
    }
    result = new LowerCaseFilter(matchVersion, result);
    result = new StopFilter(matchVersion, result, stopwords);
    if(!stemExclusionSet.isEmpty())
      result = new KeywordMarkerFilter(result, stemExclusionSet);
    result = new SnowballFilter(result, new ItalianStemmer());
    return new TokenStreamComponents(source, result);
  }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

52
53
54
55
56
57
58
    try {
      program = (SnowballProgram)stemClass.newInstance();
    } catch (Exception e) {
      throw new RuntimeException("Error instantiating stemmer for language " + language + "from class " +stemClass, e);
    }
    return new SnowballFilter(input, program);
  }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

449
450
451
452
453
454
455
456
457
458
459
            @Override public String name() {
                return "snowball";
            }

            @Override public TokenStream create(TokenStream tokenStream) {
                return new SnowballFilter(tokenStream, "English");
            }
        }));
        tokenFilterFactories.put("stemmer", new PreBuiltTokenFilterFactoryFactory(new TokenFilterFactory() {
            @Override public String name() {
                return "stemmer";
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
        this.language = Strings.capitalize(settings.get("language", settings.get("name", "porter")));
    }

    @Override public TokenStream create(TokenStream tokenStream) {
        if ("armenian".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new ArmenianStemmer());
        } else if ("basque".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new BasqueStemmer());
        } else if ("catalan".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new CatalanStemmer());
        } else if ("danish".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new DanishStemmer());
        } else if ("dutch".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new DutchStemmer());
        } else if ("english".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new EnglishStemmer());
        } else if ("finnish".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new FinnishStemmer());
        } else if ("french".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new FrenchStemmer());
        } else if ("german".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new GermanStemmer());
        } else if ("german2".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new German2Stemmer());
        } else if ("hungarian".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new HungarianStemmer());
        } else if ("italian".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new ItalianStemmer());
        } else if ("kp".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new KpStemmer());
        } else if ("kstem".equalsIgnoreCase(language)) {
            return new KStemFilter(tokenStream);
        } else if ("lovins".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new LovinsStemmer());
        } else if ("norwegian".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new NorwegianStemmer());
        } else if ("porter".equalsIgnoreCase(language)) {
            return new PorterStemFilter(tokenStream);
        } else if ("porter2".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new PorterStemmer());
        } else if ("portuguese".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new PortugueseStemmer());
        } else if ("romanian".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new RomanianStemmer());
        } else if ("russian".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new RussianStemmer());
        } else if ("spanish".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new SpanishStemmer());
        } else if ("swedish".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new SwedishStemmer());
        } else if ("turkish".equalsIgnoreCase(language)) {
            return new SnowballFilter(tokenStream, new TurkishStemmer());
        } else if ("minimal_english".equalsIgnoreCase(language) || "minimalEnglish".equalsIgnoreCase(language)) {
            return new EnglishMinimalStemFilter(tokenStream);
        } else if ("possessive_english".equalsIgnoreCase(language) || "possessiveEnglish".equalsIgnoreCase(language)) {
            return new EnglishPossessiveFilter(tokenStream);
        } else if ("light_finish".equalsIgnoreCase(language) || "lightFinish".equalsIgnoreCase(language)) {
            return new FinnishLightStemFilter(tokenStream);
        } else if ("light_french".equalsIgnoreCase(language) || "lightFrench".equalsIgnoreCase(language)) {
            return new FrenchLightStemFilter(tokenStream);
        } else if ("minimal_french".equalsIgnoreCase(language) || "minimalFrench".equalsIgnoreCase(language)) {
            return new FrenchMinimalStemFilter(tokenStream);
        } else if ("light_german".equalsIgnoreCase(language) || "lightGerman".equalsIgnoreCase(language)) {
            return new GermanLightStemFilter(tokenStream);
        } else if ("minimal_german".equalsIgnoreCase(language) || "minimalGerman".equalsIgnoreCase(language)) {
            return new GermanMinimalStemFilter(tokenStream);
        } else if ("hindi".equalsIgnoreCase(language)) {
            return new HindiStemFilter(tokenStream);
        } else if ("light_hungarian".equalsIgnoreCase(language) || "lightHungarian".equalsIgnoreCase(language)) {
            return new HungarianLightStemFilter(tokenStream);
        } else if ("indonesian".equalsIgnoreCase(language)) {
            return new IndonesianStemFilter(tokenStream);
        } else if ("light_italian".equalsIgnoreCase(language) || "lightItalian".equalsIgnoreCase(language)) {
            return new ItalianLightStemFilter(tokenStream);
        } else if ("light_portuguese".equalsIgnoreCase(language) || "lightPortuguese".equalsIgnoreCase(language)) {
            return new PortugueseLightStemFilter(tokenStream);
        } else if ("minimal_portuguese".equalsIgnoreCase(language) || "minimalPortuguese".equalsIgnoreCase(language)) {
            return new PortugueseMinimalStemFilter(tokenStream);
        } else if ("portuguese".equalsIgnoreCase(language)) {
            return new PortugueseStemFilter(tokenStream);
        } else if ("light_russian".equalsIgnoreCase(language) || "lightRussian".equalsIgnoreCase(language)) {
            return new RussianLightStemFilter(tokenStream);
        } else if ("light_spanish".equalsIgnoreCase(language) || "lightSpanish".equalsIgnoreCase(language)) {
            return new SpanishLightStemFilter(tokenStream);
        } else if ("light_swedish".equalsIgnoreCase(language) || "lightSwedish".equalsIgnoreCase(language)) {
            return new SpanishLightStemFilter(tokenStream);
        }
        return new SnowballFilter(tokenStream, language);
    }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

42
43
44
45
46
47
48
        super(index, indexSettings, name, settings);
        this.language = Strings.capitalize(settings.get("language", settings.get("name", "English")));
    }

    @Override public TokenStream create(TokenStream tokenStream) {
        return new SnowballFilter(tokenStream, language);
    }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

14
15
16
17
18
19
20
21
 
  Oscar3Filters(TokenStream ts) {
    if(false) ts = new StopFilter(ts, StandardAnalyzer.STOP_WORDS, true);
    ts = new Oscar3CaseFilter(ts);
    if(false) ts = new PunctuationFilter(ts);
    if(false) ts = new SnowballFilter(ts, "Porter");
    this.ts = ts;
  }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

254
255
256
257
258
259
260
261
262
263
264
      result = new StopFilter(matchVersion, result, stoptable);
      if (!excltable.isEmpty())
        result = new KeywordMarkerFilter(result, excltable);
      if (!stemdict.isEmpty())
        result = new StemmerOverrideFilter(matchVersion, result, stemdict);
      result = new SnowballFilter(result, new org.tartarus.snowball.ext.DutchStemmer());
      return new TokenStreamComponents(source, result);
    } else {
      final Tokenizer source = new StandardTokenizer(matchVersion, aReader);
      TokenStream result = new StandardFilter(matchVersion, source);
      result = new StopFilter(matchVersion, result, stoptable);
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
        TokenStream result = new StandardFilter(matchVersion, source);
        result = new LowerCaseFilter(matchVersion, result);
        result = new StopFilter(matchVersion, result, stopwords);
        if (!stemExclusionSet.isEmpty()) result = new KeywordMarkerFilter(
            result, stemExclusionSet);
        result = new SnowballFilter(result, new org.tartarus.snowball.ext.RussianStemmer());
        return new TokenStreamComponents(source, result);
      } else {
        final Tokenizer source = new RussianLetterTokenizer(matchVersion, reader);
        TokenStream result = new LowerCaseFilter(matchVersion, source);
        result = new StopFilter(matchVersion, result, stopwords);
        if (!stemExclusionSet.isEmpty()) result = new KeywordMarkerFilter(
          result, stemExclusionSet);
        result = new SnowballFilter(result, new org.tartarus.snowball.ext.RussianStemmer());
        return new TokenStreamComponents(source, result);
      }
    }
View Full Code Here

Examples of org.apache.lucene.analysis.snowball.SnowballFilter

125
126
127
128
129
130
131
132
    TokenStream result = new StandardFilter(matchVersion, source);
    result = new TurkishLowerCaseFilter(result);
    result = new StopFilter(matchVersion, result, stopwords);
    if(!stemExclusionSet.isEmpty())
      result = new KeywordMarkerFilter(result, stemExclusionSet);
    result = new SnowballFilter(result, new TurkishStemmer());
    return new TokenStreamComponents(source, result);
  }
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.