support 7.9.1

README.md

@@ -6,7 +6,8 @@
 
 Plugin 版本 | ES 版本 | THULAC 版本 |  Link
 -----------|-----------|----------|------------
-master | 6.x -> master | lite      |
+master | 7.x -> master | lite      |
+7.9.1  | 7.9.1         | lite      |[下载](https://github.com/microbun/elasticsearch-thulac-plugin/releases/download/7.9.1/elasticsearch-thulac-plugin-7.9.1.zip)
 6.4.1-181027 | 6.4.1          | lite      |[下载](https://github.com/microbun/elasticsearch-thulac-plugin/releases/download/6.4.1-181027/elasticsearch-thulac-plugin-6.4.1-181027.zip)
 6.4.0-181027 | 6.4.0          | lite      |[下载](https://github.com/microbun/elasticsearch-thulac-plugin/releases/download/6.4.0-181027/elasticsearch-thulac-plugin-6.4.0-181027.zip)
 6.3.0-181027 | 6.3.0          | lite      |[下载](https://github.com/microbun/elasticsearch-thulac-plugin/releases/download/6.3.0-181027/elasticsearch-thulac-plugin-6.3.0-181027.zip)
@@ -30,15 +31,15 @@ cd elasticsearch-thulac-plugin
 
 2.安装到elasticsearch
 ```
-cp build/distributions/elasticsearch-thulac-plugin-6.1.0.zip ${ES_HOME}/plugins
+cp build/distributions/elasticsearch-thulac-plugin-7.9.1.zip ${ES_HOME}/plugins
 cd ${ES_HOME}/plugins
-unzip elasticsearch-thulac-plugin-6.1.0.zip
-rm elasticsearch-thulac-plugin-6.1.0.zip
+unzip elasticsearch-thulac-plugin-7.9.1.zip
+rm elasticsearch-thulac-plugin-7.9.1.zip
 ```
 解压后在plugins目录下会有一个thulac文件夹。
 ```
 thulac
- |-elasticsearch-thulac-plugin-6.1.0.jar
+ |-elasticsearch-thulac-plugin-7.9.1.jar
  |-models #算法模型目录
  |-plugin-descriptor.properties
  |-plugin.xml
@@ -55,9 +56,7 @@ thulac
 ```bash
 curl -H "Content-Type:application/json" -XPUT http://localhost:9200/index -d'
 {
-  "settings": {
-  },
-  "mapping": {
+  "mappings": {
     "properties": {
       "text": {
         "type": "text",
@@ -93,15 +92,15 @@ curl -H "Content-Type:application/json" -XPUT http://localhost:9200/index -d'
       }
     }
   },
-  "mapping": {
+  "mappings": {
     "properties": {
       "text": {
         "type": "text",
-        "analyzer": "custom_thulac_tokenizer"
+        "analyzer": "custom_thulac_analyzer"
       }
     }
   }
-}
+}'
 ```
 
 | 参数名称 | 含义 | 值 |

build.gradle

@@ -1,5 +1,5 @@
 group 'org.elasticsearch.thulac'
-version '6.4.1-181027'
+version '7.9.1'
 
 apply plugin: 'java'
 
@@ -35,7 +35,7 @@ sourceSets {
 
 dependencies {
     testCompile group: 'junit', name: 'junit', version: '4.11'
-    compile 'org.elasticsearch:elasticsearch:6.4.1'
+    compile 'org.elasticsearch:elasticsearch:7.9.1'
 }
 
 task release_full(type: Zip, dependsOn: [':jar']) {

src/main/java/org/elasticsearch/index/ThulacAnalyzerProvider.java

@@ -21,7 +21,7 @@ public class ThulacAnalyzerProvider extends AbstractIndexAnalyzerProvider<Thulac
      */
     public ThulacAnalyzerProvider(IndexSettings indexSettings, Environment environment, String name, Settings settings) {
         super(indexSettings, name, settings);
-        Configuration configuration = new Configuration(environment, indexSettings, settings);
+        Configuration configuration = new Configuration(environment,indexSettings, settings);
         thulacAnalyzer = new ThulacAnalyzer(configuration);
     }
 

src/main/java/org/elasticsearch/index/ThulacTokenizerFactory.java

@@ -14,7 +14,7 @@ public class ThulacTokenizerFactory extends AbstractTokenizerFactory {
     private Configuration configuration;
 
     public ThulacTokenizerFactory(IndexSettings indexSettings, Environment environment, String name, Settings settings) {
-        super(indexSettings, name, settings);
+        super(indexSettings, settings, name);
         configuration = new Configuration(environment, indexSettings, settings);
     }
 

src/main/java/org/elasticsearch/thulac/Configuration.java

@@ -1,7 +1,6 @@
 package org.elasticsearch.thulac;
 
 import org.apache.logging.log4j.Logger;
-import org.elasticsearch.common.logging.ESLoggerFactory;
 import org.elasticsearch.common.logging.Loggers;
 import org.elasticsearch.common.settings.Settings;
 import org.elasticsearch.env.Environment;
@@ -23,7 +22,7 @@ public class Configuration {
     private Environment environment;
     private IndexSettings indexSettings;
     private Settings settings;
-    private Logger logger = ESLoggerFactory.getLogger(getClass());
+    private Logger logger = Loggers.getLogger(getClass(),"thulac");
 
     public Configuration() {
     }
@@ -38,7 +37,7 @@ public Configuration(Environment environment, IndexSettings indexSettings, Setti
         filter = settings.getAsBoolean("filter", false);
         modelPath = environment.pluginsFile().resolve("thulac/models");
 //        logger.info("thulac settings: path={}", modelPath.toAbsolutePath().toString());
-//        logger.info("thulac settings: user_dict={} use_t2s={} seg_only={} use_filter={} ", userDict, segOnly, useFilter);
+//        logger.info("thulac settings: user_dict={} use_t2s={} seg_only={} use_filter={} ", userDict, t2s, segOnly, useFilter);
     }
 
     public Environment getEnvironment() {

src/main/java/org/elasticsearch/thulac/ThulacLiteSegment.java

@@ -26,15 +26,14 @@ public class ThulacLiteSegment {
     private static final Map<String, CBTaggingDecoder> decoder = new HashMap<>();
 
     private static final Map<Configuration, ThulacLiteSegment> cache = new ConcurrentHashMap<>();
-    private Logger logger;
+    ;
     private CBTaggingDecoder taggingDecoder;
     // preprocess passes
     private List<IPreprocessPass> pre = new ArrayList<>();
     // postprocess passes
     private List<IPostprocessPass> post = new ArrayList<>();
 
     private ThulacLiteSegment(Configuration configuration) throws IOException {
-        logger = Loggers.getLogger(getClass());
         synchronized (decoder) {
             init(configuration);
         }

src/main/resources/plugin-descriptor.properties

@@ -34,7 +34,7 @@
 description=A thulac analysis of plugins for Elasticsearch
 #
 # 'version': plugin's version
-version=6.4.1-181027
+version=7.9.1
 #
 # 'name': the plugin name
 name=analysis-thulac-plugin
@@ -68,7 +68,7 @@ java.version=1.8
 # elasticsearch release. This version is checked when the plugin
 # is loaded so Elasticsearch will refuse to start in the presence of
 # plugins with the incorrect elasticsearch.version.
-elasticsearch.version=6.4.1
+elasticsearch.version=7.9.1
 #
 ### deprecated elements for jvm plugins :
 #

src/test/java/TestThulac.java

@@ -12,8 +12,8 @@ public class TestThulac {
 
     @Test
     public void test2() throws IOException {
-        ThulacLiteTokenizerScanner tokenizer = new ThulacLiteTokenizerScanner(new Configuration(environment));
-        InputStreamReader isr =new InputStreamReader(getClass().getClassLoader().getResource("input").openStream());
+        ThulacLiteTokenizerScanner tokenizer = new ThulacLiteTokenizerScanner(new Configuration());
+        InputStreamReader isr = new InputStreamReader(getClass().getClassLoader().getResource("input").openStream());
         tokenizer.reset(isr);
         while (tokenizer.hasNext()) {
             TaggedWord token = tokenizer.next();