Package org.tartarus.snowball.ext

Examples of org.tartarus.snowball.ext.spanishStemmer

133
134
135
136
137
138
139
140
141
    if(!stemExclusionSet.isEmpty())
      result = new SetKeywordMarkerFilter(result, stemExclusionSet);
    if (matchVersion.onOrAfter(Version.LUCENE_36)) {
      result = new SpanishLightStemFilter(result);
    } else {
      result = new SnowballFilter(result, new SpanishStemmer());
    }
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
133
134
135
136
137
138
139
140
141
    if(!stemExclusionSet.isEmpty())
      result = new KeywordMarkerFilter(result, stemExclusionSet);
    if (matchVersion.onOrAfter(Version.LUCENE_36)) {
      result = new SpanishLightStemFilter(result);
    } else {
      result = new SnowballFilter(result, new SpanishStemmer());
    }
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
134
135
136
137
138
139
140
141
142
    if(!stemExclusionSet.isEmpty())
      result = new SetKeywordMarkerFilter(result, stemExclusionSet);
    if (matchVersion.onOrAfter(Version.LUCENE_3_6)) {
      result = new SpanishLightStemFilter(result);
    } else {
      result = new SnowballFilter(result, new SpanishStemmer());
    }
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
134
135
136
137
138
139
140
141
142
    if(!stemExclusionSet.isEmpty())
      result = new SetKeywordMarkerFilter(result, stemExclusionSet);
    if (matchVersion.onOrAfter(Version.LUCENE_36)) {
      result = new SpanishLightStemFilter(result);
    } else {
      result = new SnowballFilter(result, new SpanishStemmer());
    }
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
124
125
126
127
128
129
130
131
    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 SpanishStemmer());
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
123
124
125
126
127
128
129
130
    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 SpanishStemmer());
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
134
135
136
137
138
139
140
141
142
    if(!stemExclusionSet.isEmpty())
      result = new KeywordMarkerFilter(result, stemExclusionSet);
    if (matchVersion.onOrAfter(Version.LUCENE_36)) {
      result = new SpanishLightStemFilter(result);
    } else {
      result = new SnowballFilter(result, new SpanishStemmer());
    }
    return new TokenStreamComponents(source, result);
  }
View Full Code Here
120
121
122
123
124
125
126
127
128
129
130
      this.stemmers.put("hu", new hungarianStemmer());
      this.stemmers.put("it", new italianStemmer());
      this.stemmers.put("no", new norwegianStemmer());
      this.stemmers.put("pt", new portugueseStemmer());
      this.stemmers.put("ru", new russianStemmer());
      this.stemmers.put("es", new spanishStemmer());
      this.stemmers.put("sw", new swedishStemmer());
    } catch (Exception ex) {
      throw new ResourceInitializationException(ex);
    }
View Full Code Here
136
137
138
139
140
141
142
143
144
145
146
      this.stemmers.put("hu", new hungarianStemmer());
      this.stemmers.put("it", new italianStemmer());
      this.stemmers.put("no", new norwegianStemmer());
      this.stemmers.put("pt", new portugueseStemmer());
      this.stemmers.put("ru", new russianStemmer());
      this.stemmers.put("es", new spanishStemmer());
      this.stemmers.put("sw", new swedishStemmer());
    } catch (Exception ex) {
      throw new AnnotatorInitializationException(ex);
    }
View Full Code Here
47
48
49
50
51
52
53
54
55
56
    stemmedStopwords = readInput(fs, stemmedStopwordsFile);
    isStopwordRemoval = !stopwords.isEmpty();
   
    isStemming = conf.getBoolean(Constants.Stemming, true);
    if (isStemming) {
      stemmer = new spanishStemmer();
    }
   
    LOG.warn("Stemming is " + isStemming + "; Stopword removal is " + isStopwordRemoval +"; number of stopwords: " + stopwords.size() +"; stemmed: " + stemmedStopwords.size());
  }
View Full Code Here
TOP

Related Classes of org.tartarus.snowball.ext.spanishStemmer

  • ivory.core.tokenize.LuceneSpanishAnalyzer
  • org.apache.lucene.analysis.es.SpanishAnalyzer
  • org.apache.uima.annotator.SnowballAnnotator
    • 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.