From 050021912b4cfab3148ae5c1505e3d74945d29f7 Mon Sep 17 00:00:00 2001
From: Looly <loolly@aliyun.com>
Date: Thu, 14 Dec 2023 19:14:18 +0800
Subject: [PATCH] fix thread safe

---
 .../tokenizer/engine/TokenizerEngine.java     |  4 +--
 .../tokenizer/engine/hanlp/HanLPEngine.java   |  5 ++-
 .../engine/ikanalyzer/IKAnalyzerEngine.java   | 33 ++++++++++++-------
 .../tokenizer/engine/jcseg/JcsegEngine.java   | 28 ++++++++--------
 .../tokenizer/engine/jieba/JiebaEngine.java   |  1 +
 .../tokenizer/engine/mmseg/MmsegEngine.java   | 21 ++++++------
 .../tokenizer/engine/mynlp/MynlpEngine.java   |  4 +--
 .../tokenizer/engine/word/WordEngine.java     |  3 +-
 8 files changed, 55 insertions(+), 44 deletions(-)

diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/TokenizerEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/TokenizerEngine.java
index d6e097d0b..c0a780585 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/TokenizerEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/TokenizerEngine.java
@@ -15,10 +15,10 @@ package org.dromara.hutool.extra.tokenizer.engine;
 import org.dromara.hutool.extra.tokenizer.Result;
 
 /**
- * 分词引擎接口定义，用户通过实现此接口完成特定分词引擎的适配
+ * 分词引擎接口定义，用户通过实现此接口完成特定分词引擎的适配<br>
+ * 由于引擎使用单例模式，因此要求实现类保证线程安全
  *
  * @author looly
- *
  */
 public interface TokenizerEngine {
 
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/hanlp/HanLPEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/hanlp/HanLPEngine.java
index 7b48e2f3c..e3bdc7ad0 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/hanlp/HanLPEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/hanlp/HanLPEngine.java
@@ -21,10 +21,10 @@ import org.dromara.hutool.extra.tokenizer.Result;
 
 /**
  * HanLP分词引擎实现<br>
- * 项目地址：https://github.com/hankcs/HanLP
+ * 项目地址：https://github.com/hankcs/HanLP<br>
+ * {@link Segment#seg(String)}方法线程安全
  *
  * @author looly
- *
  */
 public class HanLPEngine implements TokenizerEngine {
 
@@ -32,7 +32,6 @@ public class HanLPEngine implements TokenizerEngine {
 
 	/**
 	 * 构造
-	 *
 	 */
 	public HanLPEngine() {
 		this(HanLP.newSegment());
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/ikanalyzer/IKAnalyzerEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/ikanalyzer/IKAnalyzerEngine.java
index 337ec2321..59f67892a 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/ikanalyzer/IKAnalyzerEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/ikanalyzer/IKAnalyzerEngine.java
@@ -12,6 +12,8 @@
 
 package org.dromara.hutool.extra.tokenizer.engine.ikanalyzer;
 
+import org.wltea.analyzer.cfg.Configuration;
+import org.wltea.analyzer.cfg.DefaultConfig;
 import org.wltea.analyzer.core.IKSegmenter;
 
 import org.dromara.hutool.core.text.StrUtil;
@@ -20,35 +22,44 @@ import org.dromara.hutool.extra.tokenizer.Result;
 
 /**
  * IKAnalyzer分词引擎实现<br>
- * 项目地址：https://github.com/yozhao/IKAnalyzer
+ * 项目地址：https://github.com/yozhao/IKAnalyzer<br>
+ * {@link IKSegmenter} 非线程全，因此每次单独创建对象
  *
  * @author looly
- *
  */
 public class IKAnalyzerEngine implements TokenizerEngine {
 
-	private final IKSegmenter seg;
+	private final Configuration cfg;
 
 	/**
 	 * 构造
-	 *
 	 */
 	public IKAnalyzerEngine() {
-		this(new IKSegmenter(null, true));
+		this(createDefaultConfig());
 	}
 
 	/**
 	 * 构造
-	 *
-	 * @param seg {@link IKSegmenter}
+	 * @param cfg 配置
 	 */
-	public IKAnalyzerEngine(final IKSegmenter seg) {
-		this.seg = seg;
+	public IKAnalyzerEngine(final Configuration cfg) {
+		cfg.setUseSmart(true);
+		this.cfg = cfg;
 	}
 
 	@Override
 	public Result parse(final CharSequence text) {
-		this.seg.reset(StrUtil.getReader(text));
-		return new IKAnalyzerResult(this.seg);
+		final IKSegmenter seg = new IKSegmenter(StrUtil.getReader(text), cfg);
+		return new IKAnalyzerResult(seg);
+	}
+
+	/**
+	 * 创建默认配置
+	 * @return {@link Configuration}
+	 */
+	private static Configuration createDefaultConfig(){
+		final Configuration configuration = DefaultConfig.getInstance();
+		configuration.setUseSmart(true);
+		return configuration;
 	}
 }
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jcseg/JcsegEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jcseg/JcsegEngine.java
index 41a277ff8..cb67d6a05 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jcseg/JcsegEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jcseg/JcsegEngine.java
@@ -26,45 +26,45 @@ import java.io.StringReader;
 
 /**
  * Jcseg分词引擎实现<br>
- * 项目地址：https://gitee.com/lionsoul/jcseg
+ * 项目地址：https://gitee.com/lionsoul/jcseg<br>
+ * {@link ISegment}非线程安全，每次单独创建
  *
  * @author looly
- *
  */
 public class JcsegEngine implements TokenizerEngine {
 
-	private final ISegment segment;
+	private final SegmenterConfig config;
+	private final ADictionary dic;
 
 	/**
 	 * 构造
 	 */
 	public JcsegEngine() {
 		// 创建SegmenterConfig分词配置实例，自动查找加载jcseg.properties配置项来初始化
-		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
-		final ADictionary dic = DictionaryFactory.createSingletonDictionary(config);
-
-		// 依据给定的ADictionary和SegmenterConfig来创建ISegment
-		this.segment = ISegment.COMPLEX.factory.create(config, dic);
+		this(new SegmenterConfig(true));
 	}
 
 	/**
 	 * 构造
 	 *
-	 * @param segment {@link ISegment}
+	 * @param config {@link SegmenterConfig}
 	 */
-	public JcsegEngine(final ISegment segment) {
-		this.segment = segment;
+	public JcsegEngine(final SegmenterConfig config) {
+		this.config = config;
+		// 创建默认单例词库实现，并且按照config配置加载词库
+		this.dic = DictionaryFactory.createSingletonDictionary(config);
 	}
 
 	@Override
 	public Result parse(final CharSequence text) {
+		// 依据给定的ADictionary和SegmenterConfig来创建ISegment
+		final ISegment segment = ISegment.COMPLEX.factory.create(config, dic);
 		try {
-			this.segment.reset(new StringReader(StrUtil.str(text)));
+			segment.reset(new StringReader(StrUtil.str(text)));
 		} catch (final IOException e) {
 			throw new TokenizerException(e);
 		}
-		return new JcsegResult(this.segment);
+		return new JcsegResult(segment);
 	}
 
 }
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jieba/JiebaEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jieba/JiebaEngine.java
index 0e5d9f859..b5a74cac6 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jieba/JiebaEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/jieba/JiebaEngine.java
@@ -22,6 +22,7 @@ import org.dromara.hutool.extra.tokenizer.Result;
 /**
  * Jieba分词引擎实现<br>
  * 项目地址：https://github.com/huaban/jieba-analysis
+ * {@link JiebaSegmenter#process(String, SegMode)} 线程安全
  *
  * @author looly
  *
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mmseg/MmsegEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mmseg/MmsegEngine.java
index 047edf29f..4999d63f0 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mmseg/MmsegEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mmseg/MmsegEngine.java
@@ -12,6 +12,7 @@
 
 package org.dromara.hutool.extra.tokenizer.engine.mmseg;
 
+import com.chenlb.mmseg4j.Seg;
 import org.dromara.hutool.core.text.StrUtil;
 import org.dromara.hutool.extra.tokenizer.Result;
 import org.dromara.hutool.extra.tokenizer.engine.TokenizerEngine;
@@ -23,37 +24,35 @@ import java.io.StringReader;
 
 /**
  * mmseg4j分词引擎实现<br>
- * 项目地址：https://github.com/chenlb/mmseg4j-core
+ * 项目地址：https://github.com/chenlb/mmseg4j-core<br>
+ * {@link MMSeg}非线程安全，故单独创建之
  *
  * @author looly
- *
  */
 public class MmsegEngine implements TokenizerEngine {
 
-	private final MMSeg mmSeg;
+	private final Seg seg;
 
 	/**
 	 * 构造
 	 */
 	public MmsegEngine() {
-		final Dictionary dict = Dictionary.getInstance();
-		final ComplexSeg seg = new ComplexSeg(dict);
-		this.mmSeg = new MMSeg(new StringReader(""), seg);
+		this(new ComplexSeg(Dictionary.getInstance()));
 	}
 
 	/**
 	 * 构造
 	 *
-	 * @param mmSeg 模式{@link MMSeg}
+	 * @param seg 模式{@link Seg}
 	 */
-	public MmsegEngine(final MMSeg mmSeg) {
-		this.mmSeg = mmSeg;
+	public MmsegEngine(final Seg seg) {
+		this.seg = seg;
 	}
 
 	@Override
 	public Result parse(final CharSequence text) {
-		this.mmSeg.reset(StrUtil.getReader(text));
-		return new MmsegResult(this.mmSeg);
+		final MMSeg mmSeg = new MMSeg(StrUtil.getReader(text), seg);
+		return new MmsegResult(mmSeg);
 	}
 
 }
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mynlp/MynlpEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mynlp/MynlpEngine.java
index c3a861177..7fca5d9b4 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mynlp/MynlpEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/mynlp/MynlpEngine.java
@@ -21,10 +21,10 @@ import org.dromara.hutool.extra.tokenizer.engine.TokenizerEngine;
 
 /**
  * MYNLP 中文NLP工具包分词实现<br>
- * 项目地址：https://github.com/mayabot/mynlp/
+ * 项目地址：https://github.com/mayabot/mynlp/<br>
+ * {@link Lexer} 线程安全
  *
  * @author looly
- *
  */
 public class MynlpEngine implements TokenizerEngine {
 
diff --git a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/word/WordEngine.java b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/word/WordEngine.java
index 497c5dde7..71722c99a 100644
--- a/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/word/WordEngine.java
+++ b/hutool-extra/src/main/java/org/dromara/hutool/extra/tokenizer/engine/word/WordEngine.java
@@ -22,7 +22,8 @@ import org.dromara.hutool.extra.tokenizer.engine.TokenizerEngine;
 
 /**
  * Word分词引擎实现<br>
- * 项目地址：https://github.com/ysc/word
+ * 项目地址：https://github.com/ysc/word<br>
+ * {@link Segmentation} 线程安全
  *
  * @author looly
  *