通过开关控制个性化采用老的还是新的基于特征向量的

@@ -2,7 +2,7 @@ package com.yoho.search.service.personalized;
 
 import com.yoho.search.service.service.SearchDynamicConfigService;
 import org.apache.commons.lang3.StringUtils;
-import org.elasticsearch.index.query.functionscore.ScoreFunctionBuilder;
+import org.elasticsearch.index.query.functionscore.FunctionScoreQueryBuilder;
 import org.elasticsearch.index.query.functionscore.ScoreFunctionBuilders;
 import org.elasticsearch.script.Script;
 import org.elasticsearch.script.ScriptService;
@@ -30,17 +30,17 @@ public class PersonalVectorFeatureSearch {
     @Autowired
     private PersonalizedRedisService personalizedRedisService;
 
-    public ScoreFunctionBuilder build(String uid) {
+    public void addPersonalizedScriptScore(FunctionScoreQueryBuilder functionScoreQueryBuilder, String uid) {
         // 1. 获取特征向量版本（即生成时间，该时间需要与skn的生成时间一致才有意义）
         String vectorFeatureVersion = searchDynamicConfigService.personalizedSearchVersion();
         if (StringUtils.isEmpty(vectorFeatureVersion) || "-1".equals(vectorFeatureVersion)) {
-            return null;
+            return;
         }
 
         // 2. 获取用户的特征向量
         String userVectorFeature = personalizedRedisService.getUserVectorFeature(uid, vectorFeatureVersion);
         if (StringUtils.isEmpty(userVectorFeature)) {
-            return null;
+            return;
         }
 
         // 3. 传入参数调用脚本
@@ -56,7 +56,7 @@ public class PersonalVectorFeatureSearch {
         scriptParams.put("factorConstant", FACTOR_CONSTANT);
         logger.info("[PersonalVectorFeatureSearch.build][scriptParams={}]", scriptParams);
         Script script = new Script("feature_factor_vector_score", ScriptService.ScriptType.INLINE, "native", scriptParams);
-        return ScoreFunctionBuilders.scriptFunction(script);
+        functionScoreQueryBuilder.add(ScoreFunctionBuilders.scriptFunction(script));
     }
 
     /**

@@ -28,6 +28,15 @@ public class SearchDynamicConfigService {
 	public boolean openPersonalized() {
 		return configReader.getBoolean("search.degrade.open.personalized", true);
 	}
+
+	/**
+	 * 是否采用基于特征向量的个性化模式
+	 *
+	 * @return
+	 */
+	public boolean openVectorFeaturePersonalized() {
+		return configReader.getBoolean("search.personalized.vectorfeature.open", true);
+	}
 	
 	/**
 	 * 是否开启个性化

@@ -7,7 +7,6 @@ import org.elasticsearch.common.lucene.search.function.CombineFunction;
 import org.elasticsearch.index.query.QueryBuilder;
 import org.elasticsearch.index.query.QueryBuilders;
 import org.elasticsearch.index.query.functionscore.FunctionScoreQueryBuilder;
-import org.elasticsearch.index.query.functionscore.ScoreFunctionBuilder;
 import org.elasticsearch.index.query.functionscore.ScoreFunctionBuilders;
 import org.elasticsearch.index.query.functionscore.weight.WeightBuilder;
 import org.springframework.beans.factory.annotation.Autowired;
@@ -36,11 +35,12 @@ public class FunctionScoreSearchHelper {
 	public QueryBuilder buildFunctionScoreQueryBuild(QueryBuilder queryBuilder, Map<String, String> paramMap) {
 		FunctionScoreQueryBuilder functionScoreQueryBuilder = new FunctionScoreQueryBuilder(queryBuilder);
 		if (searchCommonHelper.isNeedPersonalSearch(paramMap)) {
-			//personalizedSearch.addPersonalizedUserFetureFactor(functionScoreQueryBuilder, paramMap);
-			//personalizedSearch.addPersonalizedFieldValueFactor(functionScoreQueryBuilder, paramMap);
-			ScoreFunctionBuilder scoreFunctionBuilder = personalVectorFeatureSearch.build(paramMap.get("uid"));
-			if (scoreFunctionBuilder != null){
-				functionScoreQueryBuilder.add(scoreFunctionBuilder);
+			if(dynamicConfig.openVectorFeaturePersonalized()) {
+				personalVectorFeatureSearch.addPersonalizedScriptScore(functionScoreQueryBuilder, paramMap.get("uid"));
+			}
+			else {
+				personalizedSearch.addPersonalizedUserFetureFactor(functionScoreQueryBuilder, paramMap);
+				personalizedSearch.addPersonalizedFieldValueFactor(functionScoreQueryBuilder, paramMap);
 			}
 		}
 		if (searchCommonHelper.containGlobal(paramMap)) {

@@ -6,6 +6,9 @@ search.degrade.direct.downgrade=false
 #search open personalized[true:open personalized, false:close personalized]
 search.degrade.open.personalized=true
 
+#search open personalized based vector feature
+search.personalized.vectorfeature.open=true
+
 #search open personalized[true:open personalized, false:close personalized]
 search.degrade.personalized.maxcount=20