Let's say I have index like this:

{
    "id": 6,
    "name": "some name",
    "users": [
      {
        "id": 1,
        "name": "User1",
        "isEnabled": false,
      },
      {
        "id": 2,
        "name": "User2",
        "isEnabled": false,
      },
      {
        "id": 3,
        "name": "User3,
        "isEnabled": true,
      },
    ]
}

what I need is to return that index while user searching for the name some name, but also I want to filter out all not enabled users, and if there is not enabled users omit that index.

I tried to use filters like this:

{
  "query": {
    "bool": {
      "must": {
        "match": {
          "name": "some name"
        }
      },
      "filter": {
        "term": {
          "users.isEnabled": true
        }
      }
    }
  }
}

but in such a case I see index with all users no matter if user is enabled or not. I'm a bit new but is there a way to do so??? I can filter out all that in code after getting data from elasticsearch but in such a case it can break pagination if I remove some index without enabled users from result set.

I'm a bit new to elasticsearch, but as far I can't find how to do it. Thank you in advice!

CodePudding user response：

Elasticsearch will return whole document if there is any match. If you update your mapping and make users array nested, you can achieve this by using inner hits. This is a basic example mapping that works:

{
  "mappings": {
    "properties": {
      "name": {
        "type": "text"
      },
      "users": {
        "type": "nested"
      }
    }
  }
}

And if you send a query like following, response will contain id and name from the parent document, and it will contain inner_hits that match to your user's isEnabled query.

{
  "_source": ["id", "name"], 
  "query": {
    "bool": {
      "must": [
        {
          "match": {
            "name": "some name"
          }
        },
        {
          "nested": {
            "path": "users",
            "query": {
              "term": {
                "users.isEnabled": {
                  "value": true
                }
              }
            },
            "inner_hits": {}
          }
        }
      ]
    }
  }
}

This is an example response

{
  "took" : 7,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 1,
      "relation" : "eq"
    },
    "max_score" : 1.9375811,
    "hits" : [
      {
        "_index" : "test",
        "_type" : "_doc",
        "_id" : "1",
        "_score" : 1.9375811,
        "_source" : {
          "name" : "some name",
          "id" : 6
        },
        "inner_hits" : {
          "users" : {
            "hits" : {
              "total" : {
                "value" : 1,
                "relation" : "eq"
              },
              "max_score" : 1.540445,
              "hits" : [
                {
                  "_index" : "test",
                  "_type" : "_doc",
                  "_id" : "1",
                  "_nested" : {
                    "field" : "users",
                    "offset" : 2
                  },
                  "_score" : 1.540445,
                  "_source" : {
                    "id" : 3,
                    "name" : "User3",
                    "isEnabled" : true
                  }
                }
              ]
            }
          }
        }
      }
    ]
  }
}

Then you can do the mapping in the application.