Elasticsearch-如何使用嵌套字段过滤数据

Question

Below are few documents with tags as nested field. Is it possible to filter the results to get the documents which has both tags.id = 21 and tags.id = 22 ?

{
  "title": "Nokia",
  "tags": [
    {
      "id": 21
    },
    {
      "id": 22
    }
  ]  
}

{
  "title": "HTC",
  "tags": [
    {
      "id": 21
    },
    {
      "id": 23
    }
  ]  
}

{
  "title": "Samsung",
  "tags": [
    {
      "id": 21
    },
    {
      "id": 22
    }
  ]  
}

In this case the result should return the documents with title Nokia and Samsung

Answer 1

You can use a boolean query:

GET my_index/_search
{
    "query": {
        "bool": {
            "filter": [
                {"match": {"tags.id": 21}},
                {"match": {"tags.id": 22}}
            ]
        }
    }
}

Answer 2

Below is the query for searching in the situation as mentioned in the question

{
    "query": {
        "bool": {
            "filter": [
                {
                    "nested":{
                        "path":"tags",
                        "query":{
                            "bool":{
                                "filter":[
                                    {"term": {"tags.id": 21}}
                                ]
                            }
                        }
                    }
                },
                {
                    "nested":{
                        "path":"tags",
                        "query":{
                            "bool":{
                                "filter":[
                                    {"term": {"tags.id": 22}}
                                ]
                            }
                        }
                    }
                }

            ]
        }
    }
}

The mistake that i was doing was putting in the 2 match blocks under the same filter block.

"filter":[
    {"term": {"tags.id": 21}},
    {"term": {"tags.id": 22}}
]