Java 如何按字段类型对匹配的文档字段进行分组？

我正在使用术语聚合来计算字段值，但在进行聚合之前，我正在基于结果聚合进行筛选搜索。。最后，我需要id和聚合计数作为示例 若异常id为1，若它得到匹配项，那个么我需要输出为

1->“键”：“转移” “单据计数”：2

2->“键”：“库存” “单据计数”：4

我在下面突出显示了异常id，我希望它作为指向每个bucket的指针

我如何在弹性搜索中做到这一点？我已经附上了示例响应

{
  "took": 250,
  "timed_out": false,
  "_shards": {
    "total": 1,
    "successful": 1,
    "skipped": 0,
    "failed": 0
  },
  "hits": {
    "total": {
      "value": 1,
      "relation": "eq"
    },
    "max_score": 0.0,
    "hits": [
      {
        "_index": "america",
        "_type": "_doc",
        "_id": "1",
        "_score": 0.0,
        "_source": {
          "clusterId": "1",
          "rank": 1,
          "events": [
            {
              "eventId": "1",
              "eventType": "Delayed",
              "metaInfo": {
                "batch_id": "batch_1"
              },
              "recommendationData": [
                {
                  ***"exceptionId": "1",***
                  "item": "Item1",
                  "location": "DC1",
                  "dueDate": "2019-01-10T05:30:00.000+0530",
                  "quantity": 100,
                  "metaInfo": {
                    "batch_id": "batch_1",
                    "dummy_id": "dummy_1"
                  },
                  "rank": 1,
                  "recommendations": [
                    {
                      "rank": 1,
                      "recommendationType": "transfer",
                      "customerName": "Walmart",
                      "stockTransfer": {
                        "primaryRecommendation": true,
                        "priority": 1,
                        "sourceLocation": "DC3",
                        "metaInfo": 40,
                        "shipDate": "2019-01-09T05:30:00.000+0530",
                        "arrivalDate": "2019-01-10T05:30:00.000+0530",
                        "transportMode": "Road",
                        "transferCost": 200.0,
                        "maxQtyAvailableForTransfer": 40,
                        "totalQtyAtSource": 40
                      },
                      "expedite": null
                    },
                    {
                      "rank": 1,
                      "recommendationType": "transfer",
                      "customerName": "Walmart",
                      "stockTransfer": {
                        "primaryRecommendation": true,
                        "priority": 2,
                        "sourceLocation": "DC2",
                        "transferQuantity": 60,
                        "shipDate": "2019-01-09T05:30:00.000+0530",
                        "arrivalDate": "2019-01-10T05:30:00.000+0530",
                        "transportMode": "Road",
                        "transferCost": 600.0,
                        "maxQtyAvailableForTransfer": 100,
                        "totalQtyAtSource": 100
                      },
                      "expedite": null
                    }
                  ]
                }
              ]
            }
          ]
        }
      }
    ]
  },
  "aggregations": {
    "recommendationTypes": {
      "doc_count": 2,
      "recommendationTypes": {
        "doc_count_error_upper_bound": 0,
        "sum_other_doc_count": 0,
        "buckets": [
          {
            "key": "transfer",
            "doc_count": 2
          }
        ]
      }
    }
  }
}

---

如果要在任何exceptionId或recommendationType上进行聚合（两者都位于嵌套对象内），则需要使用嵌套聚合

例如，如果您有一个包含两个嵌套文档的文档，例外ID为1和2。如果要在exceptionId为2的嵌套文档上进行聚合，则即使在“查询”部分使用嵌套查询进行筛选，也需要使用嵌套聚合，因为即使嵌套对象匹配，也会返回整个文档，并且必须在聚合中明确指出要在特定嵌套对象上进行聚合。 质疑

结果:

"aggregations" : {
    "recommendations" : {
      "doc_count" : 1,
      "exception" : {
        "doc_count" : 1,
        "exceptionIds" : {
          "doc_count_error_upper_bound" : 0,
          "sum_other_doc_count" : 0,
          "buckets" : [
            {
              "key" : "2",
              "doc_count" : 1,
              "recommendations" : {
                "doc_count" : 2,
                "recommendationType" : {
                  "doc_count_error_upper_bound" : 0,
                  "sum_other_doc_count" : 0,
                  "buckets" : [
                    {
                      "key" : "transfer",
                      "doc_count" : 2
                    }
                  ]
                }
              }
            }
          ]
        }
      }
    }
  }

---

您是否需要上面的java代码用于ealstic search dsl？是否有任何方法可以在弹性搜索中替代java？请告诉我是否正确理解您的问题。您需要匹配嵌套对象的计数，但取而代之的是所有嵌套对象的计数？否。除了术语bucket外，我还需要匹配的id（异常id），它应该指向bucket。。。好的，例如，如果id是1，并且它在文档中匹配，那么连同术语bucket[key and count]我需要这个1来指向bucket。。希望你基本上了解我需要什么。。对于匹配的id，我需要匹配id以及术语桶。。

"aggregations" : {
    "recommendations" : {
      "doc_count" : 1,
      "exception" : {
        "doc_count" : 1,
        "exceptionIds" : {
          "doc_count_error_upper_bound" : 0,
          "sum_other_doc_count" : 0,
          "buckets" : [
            {
              "key" : "2",
              "doc_count" : 1,
              "recommendations" : {
                "doc_count" : 2,
                "recommendationType" : {
                  "doc_count_error_upper_bound" : 0,
                  "sum_other_doc_count" : 0,
                  "buckets" : [
                    {
                      "key" : "transfer",
                      "doc_count" : 2
                    }
                  ]
                }
              }
            }
          ]
        }
      }
    }
  }