How to write complex DSL query in Elasticsearch?

Question

This is what I have used for creating index in multi field mapping

curl -X PUT http://localhost:9200/mapping_log
{ "mappings":{ "properties":{"data1:{"type": "text","fields":{"keyword":{"type":"keyword"}}}, {"data2":{"type": "text","fields":{"keyword":{"type":"keyword"}}}, {"data3":{"type": "text","fields":{"keyword":{"type":"keyword"}}}, } } }  

This is my data in my elastic search and I want to do search like for example when I search for support I should get values data and server and when I search for developer I should get geo and graph , and I want suggester for data3 like suppose if I enter g then I should get option of graph and geo ,,can someone help me with this ..

{
        "took": 14,
        "timed_out": false,
        "_shards": {
            "total": 1,
            "successful": 1,
            "skipped": 0,
            "failed": 0
        },
        "hits": {
            "total": {
                "value": 4,
                "relation": "eq"
            },
            "max_score": 1,
            "hits": [
                {
                    "_index": "mapping_log",
                    "_type": "properties",
                    "_id": "1",
                    "_score": 1,
                    "_source": {
                        "data1": "support",
                        "data2": "data",
                        "data3": "datapos"
                    }
                },
                {
                    "_index": "mapping_log",
                    "_type": "properties",
                    "_id": "2",
                    "_score": 1,
                    "_source": {
                        "data1": "support",
                        "data2": "server",
                        "data3": "serverpos"
                    }
                },
                {
                    "_index": "mapping_log",
                    "_type": "properties",
                    "_id": "3",
                    "_score": 1,
                    "_source": {
                        "data1": "developer",
                        "data2": "graph",
                        "data3": "graphpos"
                    }
                },
                {
                    "_index": "mapping_log",
                    "_type": "properties",
                    "_id": "4",
                    "_score": 1,
                    "_source": {
                        "data1": "developer",
                        "data2": "geo",
                        "data3": "geopos"
                    }
                }
            ]
        }
    }

Answer 1

You need to use edge_ngram tokenizer along with the source filtering

Adding a working example with index data (Same as that of question), index mapping, search query and search result

Index Mapping:

{
  "settings": {
    "analysis": {
      "analyzer": {
        "my_analyzer": {
          "tokenizer": "my_tokenizer"
        }
      },
      "tokenizer": {
        "my_tokenizer": {
          "type": "edge_ngram",
          "min_gram": 1,
          "max_gram": 10,
          "token_chars": [
            "letter",
            "digit"
          ]
        }
      }
    }
  },
  "mappings": {
    "properties": {
      "data1": {
        "type": "text"
      },
      "data2": {
        "type": "text"
      },
      "data3": {
        "type": "text",
        "analyzer":"my_analyzer"
      }
    }
  }
}

Search Query:

when I search for support I should get values data and server

{
  "_source": [
    "data2"
  ],
  "query": {
    "multi_match": {
      "query": "support",
      "fields": [
        "data1",
        "data2"
      ]
    }
  }
}

Search Result:

"hits": [
      {
        "_index": "67260491",
        "_type": "_doc",
        "_id": "1",
        "_score": 1.4094054,
        "_source": {
          "data2": "server"
        }
      },
      {
        "_index": "67260491",
        "_type": "_doc",
        "_id": "4",
        "_score": 0.44183272,
        "_source": {
          "data2": "data"
        }
      }
    ]

---

Search Query:

{
  "_source": [
    "data2"
  ],
  "query": {
    "multi_match": {
      "query": "developer",
      "fields": [
        "data1",
        "data2"
      ]
    }
  }
}

Search Result:

 "hits": [
      {
        "_index": "67260491",
        "_type": "_doc",
        "_id": "2",
        "_score": 1.0296195,
        "_source": {
          "data2": "graph"
        }
      },
      {
        "_index": "67260491",
        "_type": "_doc",
        "_id": "3",
        "_score": 1.0296195,
        "_source": {
          "data2": "geo"
        }
      }
    ]

---

Search Query:

I want suggester for data3 like suppose if I enter g then I should get option of graph and geo

{
  "_source": [
    "data3"
  ],
  "query": {
    "match": {
      "data3": "g"
    }
  }
}

Search Result:

"hits": [
      {
        "_index": "67260491",
        "_type": "_doc",
        "_id": "3",
        "_score": 1.1123568,
        "_source": {
          "data3": "geopos"
        }
      },
      {
        "_index": "67260491",
        "_type": "_doc",
        "_id": "2",
        "_score": 0.99270093,
        "_source": {
          "data3": "graphpos"
        }
      }
    ]

---

Update 1:

You can update the index mapping as

PUT /<index-name>/_mapping

{
  "properties": {
    "data2": {
      "type": "text",
      "fields": {
        "keyword": {
          "type": "keyword"
        }
      }
    }
  }
}