mirror of
https://gitee.com/chinabugotech/hutool.git
synced 2025-05-09 23:51:34 +08:00
fix jcseg code
This commit is contained in:
parent
6992ea9ee1
commit
b59e0c1b4f
@ -3,12 +3,10 @@ package cn.hutool.extra.tokenizer.engine.jcseg;
|
|||||||
import java.io.IOException;
|
import java.io.IOException;
|
||||||
import java.io.StringReader;
|
import java.io.StringReader;
|
||||||
|
|
||||||
import org.lionsoul.jcseg.tokenizer.core.ADictionary;
|
import org.lionsoul.jcseg.ISegment;
|
||||||
import org.lionsoul.jcseg.tokenizer.core.DictionaryFactory;
|
import org.lionsoul.jcseg.dic.ADictionary;
|
||||||
import org.lionsoul.jcseg.tokenizer.core.ISegment;
|
import org.lionsoul.jcseg.dic.DictionaryFactory;
|
||||||
import org.lionsoul.jcseg.tokenizer.core.JcsegException;
|
import org.lionsoul.jcseg.segmenter.SegmenterConfig;
|
||||||
import org.lionsoul.jcseg.tokenizer.core.JcsegTaskConfig;
|
|
||||||
import org.lionsoul.jcseg.tokenizer.core.SegmentFactory;
|
|
||||||
|
|
||||||
import cn.hutool.core.util.StrUtil;
|
import cn.hutool.core.util.StrUtil;
|
||||||
import cn.hutool.extra.tokenizer.TokenizerEngine;
|
import cn.hutool.extra.tokenizer.TokenizerEngine;
|
||||||
@ -30,19 +28,13 @@ public class JcsegEngine implements TokenizerEngine {
|
|||||||
* 构造
|
* 构造
|
||||||
*/
|
*/
|
||||||
public JcsegEngine() {
|
public JcsegEngine() {
|
||||||
// 创建JcsegTaskConfig分词配置实例,自动查找加载jcseg.properties配置项来初始化
|
// 创建SegmenterConfig分词配置实例,自动查找加载jcseg.properties配置项来初始化
|
||||||
JcsegTaskConfig config = new JcsegTaskConfig(true);
|
final SegmenterConfig config = new SegmenterConfig(true);
|
||||||
// 创建默认单例词库实现,并且按照config配置加载词库
|
// 创建默认单例词库实现,并且按照config配置加载词库
|
||||||
ADictionary dic = DictionaryFactory.createSingletonDictionary(config);
|
final ADictionary dic = DictionaryFactory.createSingletonDictionary(config);
|
||||||
|
|
||||||
// 依据给定的ADictionary和JcsegTaskConfig来创建ISegment
|
// 依据给定的ADictionary和SegmenterConfig来创建ISegment
|
||||||
try {
|
this.segment = ISegment.COMPLEX.factory.create(config, dic);
|
||||||
this.segment = SegmentFactory.createJcseg(//
|
|
||||||
JcsegTaskConfig.COMPLEX_MODE, //
|
|
||||||
new Object[] { config, dic });
|
|
||||||
} catch (JcsegException e) {
|
|
||||||
throw new TokenizerException(e);
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
|
||||||
/**
|
/**
|
||||||
|
@ -1,15 +1,14 @@
|
|||||||
package cn.hutool.extra.tokenizer.engine.jcseg;
|
package cn.hutool.extra.tokenizer.engine.jcseg;
|
||||||
|
|
||||||
import java.io.IOException;
|
|
||||||
import java.util.Iterator;
|
|
||||||
import java.util.NoSuchElementException;
|
|
||||||
|
|
||||||
import org.lionsoul.jcseg.tokenizer.core.ISegment;
|
|
||||||
import org.lionsoul.jcseg.tokenizer.core.IWord;
|
|
||||||
|
|
||||||
import cn.hutool.extra.tokenizer.Result;
|
import cn.hutool.extra.tokenizer.Result;
|
||||||
import cn.hutool.extra.tokenizer.TokenizerException;
|
import cn.hutool.extra.tokenizer.TokenizerException;
|
||||||
import cn.hutool.extra.tokenizer.Word;
|
import cn.hutool.extra.tokenizer.Word;
|
||||||
|
import org.lionsoul.jcseg.ISegment;
|
||||||
|
import org.lionsoul.jcseg.IWord;
|
||||||
|
|
||||||
|
import java.io.IOException;
|
||||||
|
import java.util.Iterator;
|
||||||
|
import java.util.NoSuchElementException;
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Jcseg分词结果包装<br>
|
* Jcseg分词结果包装<br>
|
||||||
@ -64,6 +63,7 @@ public class JcsegResult implements Result{
|
|||||||
throw new UnsupportedOperationException("Jcseg result not allow to remove !");
|
throw new UnsupportedOperationException("Jcseg result not allow to remove !");
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@SuppressWarnings("NullableProblems")
|
||||||
@Override
|
@Override
|
||||||
public Iterator<Word> iterator() {
|
public Iterator<Word> iterator() {
|
||||||
return this;
|
return this;
|
||||||
|
@ -1,8 +1,7 @@
|
|||||||
package cn.hutool.extra.tokenizer.engine.jcseg;
|
package cn.hutool.extra.tokenizer.engine.jcseg;
|
||||||
|
|
||||||
import org.lionsoul.jcseg.tokenizer.core.IWord;
|
|
||||||
|
|
||||||
import cn.hutool.extra.tokenizer.Word;
|
import cn.hutool.extra.tokenizer.Word;
|
||||||
|
import org.lionsoul.jcseg.IWord;
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Jcseg分词中的一个单词包装
|
* Jcseg分词中的一个单词包装
|
||||||
|
Loading…
x
Reference in New Issue
Block a user