es 根据父文档，对子文档聚合

最新推荐文章于 2024-07-30 17:00:13 发布

zwy_

最新推荐文章于 2024-07-30 17:00:13 发布

阅读量1.5k

点赞数 1

分类专栏： es Java 文章标签： elasticsearch es

本文链接：https://blog.csdn.net/momo_joanne/article/details/115757883

版权

Java 同时被 2 个专栏收录

12 篇文章 0 订阅

订阅专栏

9 篇文章 1 订阅

订阅专栏

https://www.elastic.co/guide/en/elasticsearch/reference/current/search-aggregations-bucket-children-aggregation.html

上面链接是官方文档对于父子文档的聚合示例，解释还是很清楚的。

下是我自己根据官网进行的聚合（就不提供数据了，只是我的模仿而已）

    /*
     * ******************************* 此处ES查询语句形如 **************************************/
         GET company_info/_search
        {
          "size": 0,
          "aggregations": {

            "top_district":{
              "terms": {
                "field": "districtName"
              } ,
              "aggs": {
                "to-person": {
                  "children": {
                    "type": "person"
                  } ,
                  "aggs": {
                    "per_state": {
                      "terms": {
                        "field": "personCerts"
                      }
                    }
                  }
                }
              }
            }
          }
        }

结果，为了一张截图展示更多，对部分结果做了折叠处理，折叠部分和下面未折叠部分数据格式一致。

Java 父子聚合用到的是

具体代码如下

    public Map<String,List<Map<String,Long>>> aggregationDistrictPerson() throws Exception {
        TermsAggregationBuilder perStats = AggregationBuilders.terms("per_state").field("personCerts").size(10);
        //这句就是针对子文档的聚合了
        ChildrenAggregationBuilder childrenAggregation = JoinAggregationBuilders.children("to-person","person").subAggregation(perStats) ;

        TermsAggregationBuilder districtStats = AggregationBuilders.terms("top_district").field("districtName").subAggregation(childrenAggregation) ;

        Aggregations aggregations  = esSearchService.aggregate(districtStats,matchAllQuery(),ES_INDEX);

        Map<String, Aggregation> aggMap = aggregations.asMap();
        Terms termAgg = (Terms) aggMap.get("top_district");

        List<? extends Terms.Bucket> buckets = termAgg.getBuckets();

        Map<String,List<Map<String,Long>>> resultMap = new HashMap<>();

        buckets.forEach(bucket -> {
            List<Map<String,Long>> listMap = new ArrayList<>();
            Aggregations bucketAggregations = bucket.getAggregations();
            Map<String, Aggregation> aggregationMap = bucketAggregations.asMap() ;
            ParsedChildren toPerson = (ParsedChildren) aggregationMap.get("to-person");
            Aggregations termAggregations = toPerson.getAggregations();
            Map<String, Aggregation> termAggregationMap = termAggregations.asMap() ;

            Terms perState = (Terms) termAggregationMap.get("per_state");
            List<? extends Terms.Bucket> perStateBuckets = perState.getBuckets();

            Map<String,Long> tbMap = new HashMap<>();
            perStateBuckets.forEach(tb->tbMap.put(tb.getKey().toString(),tb.getDocCount()));
            listMap.add(tbMap);
            resultMap.put(bucket.getKey().toString(),listMap);
        });

        log.info(buckets);
        log.info(aggregations);

        return resultMap ;
    }

整理后，结果形如