elasticsearch query with Python lists returns "Can't get text on a START_ARRAY at 1:30"

Question

I'm new with elastic (any also with python) but have some first success:

es = Elasticsearch(['http://localhost:9200'], basic_auth=('user', 'pw'))
query_body = {
  "query": {
      "match": {
          "content": "English"
      }
  }
}

response = es.search(index="resumes", body=query_body)

works fine at all. But I I have a list of keywords I need a way to use list input instead of string only.

I found this sample How to make elasticsearch query with Python lists and it seems to be simple, just change code to:

query_body = {
  "query": {
      "match": {
          "content": ["English", "German"]
      }
  }
}

as at the other threat used. But I get an exception: "Can't get text on a START_ARRAY at 1:30".

Some more research point out that there was a previous bug but this is out of date I expect. https://github.com/elastic/elasticsearch/issues/15741

I'm using elastic 8.15.1 - maybe somebody know a way how to make it work?

Thank you, Andre

Answer 1

Tldr;

The match query does not accept arrays. There are ways around it.

Demo

Dataset

POST 79536558/_bulk
{"index":{}}
{"content": "this is english"}
{"index":{}}
{"content": "das ist deutsch"}
{"index":{}}
{"content": "C'est du francais"}

Solution 1 - boolean with should

GET 79536558/_search
{
  "query": {
    "bool": {
      "should": [
        {
          "match": {
            "content": "english"
          }
        },
        {
          "match": {
            "content": "deutsch"
          }
        }
      ]
    }
  }
}

Solution 2 - query string

GET 79536558/_search
{
  "query": {
    "query_string": {
      "default_field": "content",
      "query": "english OR deutsch"
    }
  }
}