子级聚合

children 聚合将父文档与其相关的子文档连接起来。这使得您可以在单个查询中分析不同类型数据之间的关系，而无需运行多个查询并手动合并结果。

示例索引、样本数据和子级聚合查询

例如，如果作者、帖子和评论之间存在父子关系，您可以在单个查询中分析不同数据类型（authors、posts 和 comments）之间的关系。

authors 聚合按 author.keyword 字段对文档进行分组。这使您可以查看与每位作者关联的文档数量。

在每个作者组中，children 聚合检索关联的帖子。这为您提供了每位作者所写文章的细分。

在 posts 聚合中，另一个 children 聚合获取与每个帖子关联的评论。这为您提供了一种查看每个帖子的评论的方法。

在 comments 聚合中，value_count 聚合统计每个帖子的评论数量。这使您可以通过查看每个帖子收到的评论数量来衡量其互动水平。

示例索引

PUT /blog-sample
{
  "mappings": {
    "properties": {
      "type": { "type": "keyword" },
      "name": { "type": "keyword" },
      "title": { "type": "text" },
      "content": { "type": "text" },
      "author": { "type": "keyword" },
      "post_id": { "type": "keyword" },
      "join_field": {
        "type": "join",
        "relations": {
          "author": "post",
          "post": "comment"
        }
      }
    }
  }
}

示例文档

POST /blog-sample/_doc/1?routing=1
{
  "type": "author",
  "name": "John Doe",
  "join_field": "author"
}

POST /blog-sample/_doc/2?routing=1
{
  "type": "post",
  "title": "Introduction to UDB-SX",
  "content": "UDB-SX is a powerful search and analytics engine...",
  "author": "John Doe",
  "join_field": {
    "name": "post",
    "parent": "1"
  }
}

POST /blog-sample/_doc/3?routing=1
{
  "type": "comment",
  "content": "Great article! Very informative.",
  "join_field": {
    "name": "comment",
    "parent": "2"
  }
}

POST /blog-sample/_doc/4?routing=1
{
  "type": "comment",
  "content": "Thanks for the clear explanation.",
  "join_field": {
    "name": "comment",
    "parent": "2"
  }
}

示例子级聚合查询

GET /blog-sample/_search
{
  "size": 0,
  "aggs": {
    "authors": {
      "terms": {
        "field": "name.keyword"
      },
      "aggs": {
        "posts": {
          "children": {
            "type": "post"
          },
          "aggs": {
            "post_titles": {
              "terms": {
                "field": "title.keyword"
              },
              "aggs": {
                "comments": {
                  "children": {
                    "type": "comment"
                  },
                  "aggs": {
                    "comment_count": {
                      "value_count": {
                        "field": "_id"
                      }
                    }
                  }
                }
              }
            }
          }
        }
      }
    }
  }
}

示例响应

响应应类似于以下示例：

{
  "took": 30,
  "timed_out": false,
  "_shards": {
    "total": 1,
    "successful": 1,
    "skipped": 0,
    "failed": 0
  },
  "hits": {
    "total": {
      "value": 4,
      "relation": "eq"
    },
    "max_score": null,
    "hits": []
  },
  "aggregations": {
    "authors": {
      "doc_count_error_upper_bound": 0,
      "sum_other_doc_count": 0,
      "buckets": []
    }
  }
}