CODEWITHSUNDEEP
elasticsearch
MishraJi
MishraJi

Reputation: 313

Elastic search concatenation of fields

I am new to elastic search and wanted to solve an use case with below problem

I have some documents as below format

{
 "member_id":"number"
 "e_id":"number",
 "c_id":"number",
 "salary":"float"
}

In my Java application, I will get a String input and need to find the matching document's member_id. The algorithm to find the document is as below

STEP 1: Take input string as "keyInput".
STEP 2: Concat "e_id","c_id" and ceil("salary") as "concatedData"
STEP 3: if "concatedData" equals to "keyInput" THEN return the "member_id" of the matching document.

STEP 2 need to be performed in Elasticsearch.

Can anyone help to figure out if this kind of actions can be performed on ES or not, if Yes then how?

Thanks in Advance.

Upvotes: 3

Views: 16755

Answers (1)

Nishant
Nishant

Reputation: 7874

There are two approaches to solve this.

1. Using script:

Suggested Mapping:

"mappings": {
    "_doc": {
      "properties": {
        "member_id": {
          "type": "integer"
        },
        "e_id": {
          "type": "integer"
        },
        "c_id": {
          "type": "integer"
        },
        "salary": {
          "type": "float"
        }
      }
    }
  }

Query:

{
  "query": {
    "bool": {
      "filter": {
        "script": {
          "script": {
            "source": "String.valueOf(doc['e_id'].value) + doc['c_id'].value + Math.ceil(doc['salary'].value).intValue() == params.inp",
            "lang": "painless",
            "params": {
              "inp": "12345678"
            }
          }
        }
      }
    }
  }
}

Note: Scripting should be avoided as it cause performance hit. I would strongly recommend the second approach below.

2. By concatenating value at index time

For this all you need to do is add a new field with datatype as keyword to store the concatenated value. Then use term query to filter the documents accordingly.

Upvotes: 4

Related Questions