Newline based resultFormat for Druid native queries

I have a usecase wherein a Druid query results in millions of rows which makes it obvious to consume the response as a stream of rows with a new line for each.

The Druid SQL over HTTP has very good support for this with a resultFormat config supporting objectLines, arrayLines, & csv enabling stream-based response consumption.

However, I’m using Druid’s native queries which don’t seem to consider resultFormat config. The response is a JSONArray just like -

[
  {
    "version" : "v1",
    "timestamp" : "2012-01-01T00:00:00.000Z",
    "event" : {
      "country" : <some_dim_value_one>,
      "device" : <some_dim_value_two>,
      "total_usage" : <some_value_one>,
      "data_transfer" :<some_value_two>,
      "avg_usage" : <some_avg_usage_value>
    }
  },
  {
    "version" : "v1",
    "timestamp" : "2012-01-01T00:00:12.000Z",
    "event" : {
      "dim1" : <some_other_dim_value_one>,
      "dim2" : <some_other_dim_value_two>,
      "sample_name1" : <some_other_value_one>,
      "sample_name2" :<some_other_value_two>,
      "avg_usage" : <some_other_avg_usage_value>
    }
  },
...
]

This response doesn’t contain newline and is difficult to process as a stream. It needs to download entire resultset and then parse using tools like jq to extract row per line.

P.S I’ve tested jd --stream which is again not very fast.

Hence, want to know if there is anything to get native query results as newline-separated rows of json documents.