- group id : com.github.lucene-gosen
- artifact id : lucene-gosen
There are three types of jar files:
- lucene-gosen-.jar : Only java library, not include dictionary.
- lucene-gosen--ipadic.jar : Java library with IPA dictionary.
- lucene-gosen--naist-chasen.jar : Java library with Naist Chasen dictionary
- Download jar file from Maven Central Repository
- Create <your_solr_home>/<collection_dir>/lib and put this jar file in it.
- Copy stopwords_ja.txt and stoptags_ja.txt into <your_solr_home>/<collection_dir>/conf/lang
- Add "text_ja_gosen" fieldtype: see example/schema.xml.snippet for example configuration.
Please refer to example/
for an example japanese configuration with comments explaining
what the various configuration options are.
Add dependency to pom.xml.
<dependencies>
<dependency>
<groupId>com.github.lucene-gosen</groupId>
<artifactId>lucene-gosen</artifactId>
<version>7.4.0</version>
<classifier>ipadic</classifier>
</dependency>
...
</dependencies>
- Download jar file from Maven Central Repository
- Add this jar file to your classpath, and use GosenAnalyzer, or make your own analyzer from the various filters. Its recommended you extend ReusableAnalyzerBase to make any custom analyzer!
You can build the project using Gradle. And you should use gradlew
command.
Build only jar file without dictionary
$ ./gradlew jar
Build jar file with IPA dictionary
$ ./gradlew jarWithIpaDic
Build jar file with Naist Chasen dictionary
$ ./gradlew jarWithNaistChasen
Please note that you should modify the following line in gradle.properties
if you want to build the Gosen for Solr 7.1 or before.
luceneVersion = 7.4.0