Unable to sort results using script_score function

Elastic Stack Elasticsearch
1

I have a field called price in my elasticsearch index whose values I need to convert before sorting -

{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 4,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "products",
        "_type" : "_doc",
        "_id" : "1",
        "_score" : 1.0,
        "_source" : {
          "prod_id" : 1,
          "currency" : "USD",
          "price" : 1
        }
      },
      {
        "_index" : "products",
        "_type" : "_doc",
        "_id" : "2",
        "_score" : 1.0,
        "_source" : {
          "prod_id" : 2,
          "currency" : "INR",
          "price" : 60
        }
      },
      {
        "_index" : "products",
        "_type" : "_doc",
        "_id" : "3",
        "_score" : 1.0,
        "_source" : {
          "prod_id" : 3,
          "currency" : "EUR",
          "price" : 2
        }
      },
      {
        "_index" : "products",
        "_type" : "_doc",
        "_id" : "5",
        "_score" : 1.0,
        "_source" : {
          "prod_id" : 5,
          "currency" : "MYR",
          "price" : 1
        }
      }
    ]
  }
}

As the currency field for each product is different,

I plan to convert each product's price into USD,

Using the script_score function, as recommended here -

I tried the following query -

GET products/_search
{
	"query": {
		"function_score": {
			"query": {
				"match_all": {}
			},
			"functions": [{
				"script_score": {
					"script": {
						"params": {
							"USD": 1,
							"SGD": 0.72,
							"MYR": 0.24,
							"INR": 0.014,
							"EUR": 1.12
						},
						"source": "doc['price'].value * params.doc['currency']"
					}
				}
			}]
		}
	},
	"sort": [
    {
      "_score": {
        "order": "desc"
      }
    }
  ]
}

I'm getting an error -

{
  "error": {
    "root_cause": [
      {
        "type": "script_exception",
        "reason": "runtime error",
        "script_stack": [
          "doc['price'].value * params.doc['currency']",
          "                               ^---- HERE"
        ],
        "script": "doc['price'].value * params.doc['currency']",
        "lang": "painless"
      }
    ],
    "type": "search_phase_execution_exception",
    "reason": "all shards failed",
    "phase": "query",
    "grouped": true,
    "failed_shards": [
      {
        "shard": 0,
        "index": "products",
        "node": "5-fQ27BhSUKycVJ2SwyH4A",
        "reason": {
          "type": "script_exception",
          "reason": "runtime error",
          "script_stack": [
            "doc['price'].value * params.doc['currency']",
            "                               ^---- HERE"
          ],
          "script": "doc['price'].value * params.doc['currency']",
          "lang": "painless",
          "caused_by": {
            "type": "class_cast_exception",
            "reason": "Cannot apply [*] operation to types [java.lang.Long] and [org.elasticsearch.index.fielddata.ScriptDocValues.Strings]."
          }
        }
      }
    ]
  },
  "status": 400
}

Expected product sequence -

  1. prod_id 3, _score 2.24 = 2(price) * 1.12 (USD multiplier)
  2. prod_id 1, _score 1 = 1(price) * 1 (USD multiplier) = 1
  3. prod_id 2, _score 0.84 = 60(price) * 0.014 (USD multiplier)
  4. prod_id 5, _score 0.24 = 1(price) * 0.24 (USD multiplier)

Index Mapping -

{
  "mapping": {
    "properties": {
      "currency": {
        "type": "text",
        "fields": {
          "keyword": {
            "type": "keyword",
            "ignore_above": 256
          }
        },
        "fielddata": true
      },
      "price": {
        "type": "long"
      },
      "prod_id": {
        "type": "long"
      }
    }
  }
}

Can someone please help for this use case query?

2

Your Painless script is not quite correct. Try this:

doc['price'].value * params[doc['currency.keyword'].value]
3

Apart of the main topic here, please look at the proposed new feature to enable precise retrieval of computed function value in function score query _explanation. Thanks!
:point_right: https://github.com/elastic/elasticsearch/issues/46007

4

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.