honeymoose/OpenSearch
1
0
Fork 0
mirror of https://github.com/honeymoose/OpenSearch.git synced 2025-02-17 10:25:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

with the new stemmer analyzer, don't break when there is no language stopwords, but just create it without stopwords

Browse Source
This commit is contained in:
kimchy 2011-01-06 09:53:56 +02:00
parent 4a45df88c6
commit cc5978393b
2 changed files with 14 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
.idea/libraries/lucene.xml generated
View File

@ -17,6 +17,7 @@
<root url="file://$USER_HOME$/opt/lucene/3.0.3.src/src/java" />
<root url="file://$USER_HOME$/opt/lucene/3.0.3.src/contrib/highlighter/src/java" />
<root url="file://$USER_HOME$/opt/lucene/3.0.3.src/contrib/memory/src/java" />
<root url="file://$USER_HOME$/opt/lucene/3.0.3.src/contrib/snowball/src/java" />
</SOURCES>
</library>
</component>

44
modules/elasticsearch/src/main/java/org/elasticsearch/index/analysis/SnowballAnalyzerProvider.java
View File

@ -24,6 +24,9 @@ import org.apache.lucene.analysis.de.GermanAnalyzer;
import org.apache.lucene.analysis.fr.FrenchAnalyzer;
import org.apache.lucene.analysis.nl.DutchAnalyzer;
import org.apache.lucene.analysis.snowball.SnowballAnalyzer;
import org.elasticsearch.common.collect.ImmutableMap;
import org.elasticsearch.common.collect.ImmutableSet;
import org.elasticsearch.common.collect.MapBuilder;
import org.elasticsearch.common.inject.Inject;
import org.elasticsearch.common.inject.assistedinject.Assisted;
import org.elasticsearch.common.lucene.Lucene;
@ -44,39 +47,18 @@ import java.util.Set;
* The SnowballAnalyzer comes with a StandardFilter, LowerCaseFilter, StopFilter
* and the SnowballFilter.
*
* @author kimchy (Shay Banon)
* @author harryf (Harry Fuecks)
*/
public class SnowballAnalyzerProvider extends AbstractIndexAnalyzerProvider<SnowballAnalyzer> {
private enum SupportedAnalyzer {
DUTCH {
public Set<?> getStopwords() {
return DutchAnalyzer.getDefaultStopSet();
}
},
ENGLISH {
public Set<?> getStopwords() {
return StopAnalyzer.ENGLISH_STOP_WORDS_SET;
}
},
FRENCH {
public Set<?> getStopwords() {
return FrenchAnalyzer.getDefaultStopSet();
}
},
GERMAN {
public Set<?> getStopwords() {
return GermanAnalyzer.getDefaultStopSet();
}
},
GERMAN2 {
public Set<?> getStopwords() {
return GermanAnalyzer.getDefaultStopSet();
}
};
public abstract Set<?> getStopwords();
}
private static final ImmutableMap<String, Set<?>> defaultLanguageStopwords = MapBuilder.<String, Set<?>>newMapBuilder()
.put("English", StopAnalyzer.ENGLISH_STOP_WORDS_SET)
.put("Dutch", DutchAnalyzer.getDefaultStopSet())
.put("German", GermanAnalyzer.getDefaultStopSet())
.put("German2", GermanAnalyzer.getDefaultStopSet())
.put("French", FrenchAnalyzer.getDefaultStopSet())
.immutableMap();
private final SnowballAnalyzer analyzer;
@ -84,8 +66,8 @@ public class SnowballAnalyzerProvider extends AbstractIndexAnalyzerProvider<Snow
super(index, indexSettings, name);
String language = settings.get("language", "English");
Set<?> stopWords = Analysis.parseStopWords(settings,
SupportedAnalyzer.valueOf(language.toUpperCase()).getStopwords());
Set<?> defaultStopwords = defaultLanguageStopwords.containsKey(language) ? defaultLanguageStopwords.get(language) : ImmutableSet.<Set<?>>of();
Set<?> stopWords = Analysis.parseStopWords(settings, defaultStopwords);
analyzer = new SnowballAnalyzer(Lucene.VERSION, language, stopWords);
}