This is a record-aware processor built for Elasticsearch 5 and later. The index and type fields are configured with default values that can be overridden using record path operations that find an index or type value in the record set. The ID and operation type (index, update, upsert or delete) can also be extracted in a similar fashion from the record set. The following is an example of a document exercising all of these features:

        {
            "metadata": {
                "id": "12345",
                "index": "test",
                "type": "message",
                "operation": "index"
            },
            "message": "Hello, world",
            "from": "john.smith"
        }
    
        {
            "metadata": {
                "id": "12345",
                "index": "test",
                "type": "message",
                "operation": "delete"
            }
        }
    

The record path operations below would extract the relevant data:

Valid values for "operation" are: