es-object vs nested vs has_child and has_parent

1.object类型

增加object类型的索引

PUT /joining_query_object_test/_doc/1
{
   
  "obj1": {
   
    "name": "blue",
    "count": 5
  },
  "tags": [
    "龙虎榜",
    "金牛奖"
  ],
  "comments": [
    {
   
      "name": "司马懿",
      "comment": "智务其实,愚争虚名",
      "age": 28,
      "stars": 4,
      "date": "2022-05-01"
    },
    {
   
      "name": "刘伯温",
      "comment": "平而后清,清而后明",
      "age": 31,
      "stars": 5,
      "date": "2022-05-11"
    }
  ],
  "group" : "fans",
  "user" : [ 
    {
   
      "first" : "John",
      "last" :  "Smith"
    },
    {
   
      "first" : "Alice",
      "last" :  "White"
    }
  ]
}

PUT /joining_query_object_test/_doc/2
{
   
  "obj1": {
   
    "name": "red",
    "count": 3
  },
    "tags": [
    "1h热股",
    "24h热股"
  ],
  "comments": [
    {
   
      "name": "刘基",
      "comment": "有酒且尽欢，听我薤露歌",
      "age": 42,
      "stars": 3,
      "date": "2022-05-02"
    },
    {
   
      "name": "袁天罡",
      "comment": "兴也洛水，衰也洛水",
      "age": 32,
      "stars": 4,
      "date": "2022-05-12"
    }
  ],
  "group" : "fans",
  "user" : [ 
    {
   
      "first" : "xiaoping",
      "last" :  "liu"
    },
    {
   
      "first" : "tianji",
      "last" :  "li"
    }
  ]
}

查看mapping

GET /joining_query_object_test/_mapping/

##查询年龄是32岁的刘基评论
##理论上应该hit 1，结果 hit 2
##刘基时42岁，32岁的是袁天罡，根本不是同一条评论
##原因：object类型底层数据结构，会将一个json数组中的数据，进行扁平化
GET /joining_query_object_test/_search
{
   
  "query": {
   
    "bool": {
   
      "must": [
        {
   
          "match": {
   
            "comments.name": "刘基"
          }
        },
        {
   
          "match": {
   
            "comments.age": 32
          }
        }
      ]
    }
  }
}

## 验证：
##+(+(comments.name:刘 comments.name:基) +comments.age:[32 TO 32]) #MatchNoDocsQuery("")
GET /joining_query_object_test/_mapping/_validate/query?explain
{
   
  "query": {
   
    "bool": {
   
      "must": [
        {
   
          "match": {
   
            "comments.name": "刘基"
          }
        },
        {
   
          "match": {
   
            "comments.age": 32
          }
        }
      ]
    }
  }
}

##查询first=Alice and last=Smith 出现数据不准的问题
GET joining_query_object_test/_search
{
   
  "query": {
   
    "bool": {
   
      "must": [
        {
    "match": {
    "user.first": "Alice" }},
        {
    "match": {
    "user.last":  "Smith" }}
      ]
    }
  }
}

2.nested query

• doc必须是nested类型。
• nested类型就是为了解决object类型在对象数组上丢失关联性的问题的

（1）nested query

##设置mapping
DELETE joining_query_nested_test
PUT /joining_query_nested_test
{
   
  "mappings": {
   
    "properties": {
   
      "obj1": {
   
        "type": "nested"
      },
      "comments": {