ElasticSearch dynamic default mapping for nested objects

README.md

First start ES with /etc/elasticsearch/default-mapping.json, then index several documents with inner objects

curl -XPOST -d '{"key":{"subkey1":1, "subkey2":1}}' localhost:9200/toto/toto/1
curl -XPOST -d '{"key2":{"subkey1":1, "subkey2":1}}' localhost:9200/toto/toto/2
curl -XPOST -d '{"key3":{"subkey1":1, "subkey2":1}}' localhost:9200/toto/toto/3

Finally, launch the query

$ curl -XPOST -d @query.json localhost:9200/toto/_search
{"took":7,"timed_out":false,"_shards":{"total":5,"successful":5,"failed":0},"hits":{"total":1,"max_score":1.0,"hits":[{"_index":"toto","_type":"toto","_id":"2","_score":1.0,"_source":{"key2":{"subkey1":1, "subkey2":1}}}]}}

default-mapping.json

{
    "_default_" : {
        "dynamic_templates": [
            {
                "objects": {
                    "match": "*",
                    "match_mapping_type" : "object",
                    "mapping": {
                        "type": "nested",
                        "include_in_parent": true
                    }
                }
            }
        ]
    }
}

query.json

{
  "query": {
    "filtered": {
      "query": {
        "match_all": {}
      },
      "filter": {
        "nested": {
          "path": "key2",
          "filter": {
            "bool": {
              "must": [
                {
                  "term": {
                    "key2.subkey1": 1
                  }
                },
                {
                  "term": {
                    "key2.subkey2": 1
                  }
                }
              ]
            }
          }
        }
      }
    }
  }
}

But if there is multiple nested objects then in this case how we can put the mapping file