有好货功能优化

@@ -3,6 +3,7 @@ package com.yoho.search.service.servicenew.impl;
 import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.Collections;
+import java.util.HashMap;
 import java.util.Iterator;
 import java.util.List;
 import java.util.Map;
@@ -68,6 +69,9 @@ public class GoodProductListService implements IGoodProductsService {
 
 	private final int maxSmallSortCount = 20;
 	private final int maxProductSknCountPerSort = 5;
+	private final int maxCountPerGroup = 10;
+	private final float firstSknScore = 300;
+	private final float recommendedSknMaxScore = 200;
 
 	@Override
 	public SearchApiResult goodProductList(Map<String, String> paramMap) {
@@ -122,17 +126,22 @@ public class GoodProductListService implements IGoodProductsService {
 		return new SearchApiResult().setData(dataMap);
 	}
 
-	private QueryBuilder builderGoodProductQueryBuilder(Map<String, String> paramMap, List<String> recommondSkns) {
+	private QueryBuilder builderGoodProductQueryBuilder(Map<String, String> paramMap, List<String> recommendedSknList) {
 		QueryBuilder queryBuilder = QueryBuilders.matchAllQuery();
 		FunctionScoreQueryBuilder functionScoreQueryBuilder = new FunctionScoreQueryBuilder(queryBuilder);
 		// 针对参数里第一个SKN加分
 		String productSkns = paramMap.get(SearchRequestParams.PARAM_SYNC_SKN);
 		if (!StringUtils.isBlank(productSkns)) {
-			functionScoreQueryBuilder.add(QueryBuilders.termsQuery("productSkn", productSkns.split(",")[0]), ScoreFunctionBuilders.weightFactorFunction(100));
+			functionScoreQueryBuilder.add(QueryBuilders.termsQuery("productSkn", productSkns.split(",")[0]), ScoreFunctionBuilders.weightFactorFunction(firstSknScore));
 		}
 		// 针对推荐出来的SKN做加分
-		if (recommondSkns != null && !recommondSkns.isEmpty()) {
-			functionScoreQueryBuilder.add(QueryBuilders.termsQuery("productSkn", recommondSkns), ScoreFunctionBuilders.weightFactorFunction(50));
+		if (recommendedSknList != null && !recommendedSknList.isEmpty()) {
+			Map<Integer, List<String>> recommondSknMap = this.splitProductSkns(recommendedSknList, maxCountPerGroup);
+			float currentGroupScore = recommendedSknMaxScore;
+			for (Map.Entry<Integer, List<String>> entry: recommondSknMap.entrySet()) {
+				functionScoreQueryBuilder.add(QueryBuilders.termsQuery("productSkn", entry.getValue()), ScoreFunctionBuilders.weightFactorFunction(currentGroupScore));
+				currentGroupScore = currentGroupScore - 10;
+			}
 		}
 		// 加上个性化打分
 		if (searchCommonHelper.isNeedPersonalSearch(paramMap)) {
@@ -291,6 +300,28 @@ public class GoodProductListService implements IGoodProductsService {
 		return JSONObject.parseArray(jsonArray.toJSONString(), clazz);
 	}
 
+	/**
+	 * 分组
+	 * 
+	 * @param recommendedSknList
+	 * @param maxGroupCount
+	 * @return
+	 */
+	private Map<Integer, List<String>> splitProductSkns(List<String> recommendedSknList, int maxCountPerGroup) {
+		int maxSize = recommendedSknList.size();
+		int groupSize = maxSize / maxCountPerGroup;
+		if (maxSize % maxCountPerGroup > 0) {
+			groupSize = groupSize + 1;
+		}
+		Map<Integer, List<String>> result = new HashMap<Integer, List<String>>();
+		for (int i = 0; i < groupSize; i++) {
+			int fromIndex = i * maxCountPerGroup;
+			int toIndex = (i + 1) * maxCountPerGroup;
+			result.put(i, recommendedSknList.subList(fromIndex, toIndex > maxSize ? maxSize : toIndex));
+		}
+		return result;
+	}
+
 	private List<String> getRecommendedSknList(MultiBucketsAggregation aggregation) {
 		Iterator<? extends Bucket> itAgg = aggregation.getBuckets().iterator();
 		// 获取聚合出来的商品
@@ -310,4 +341,16 @@ public class GoodProductListService implements IGoodProductsService {
 		Collections.shuffle(recommendedSknList);
 		return recommendedSknList;
 	}
+
+	public static void main(String[] args) {
+		List<String> list = new ArrayList<String>();
+		for (int i = 1; i <= 99; i++) {
+			list.add(i+"");
+		}
+		Map<Integer, List<String>> results = new GoodProductListService().splitProductSkns(list, 20);
+		for (Map.Entry<Integer, List<String>> entry : results.entrySet()) {
+			System.out.println(entry.getKey() + "_" + entry.getValue());
+		}
+	}
+
 }