Most efficient way to convert RDD[Status] to specific json - json

Question

I have a RDD[twitter4j.Status] (from TwitterUtils in Spark API) in my Spark streaming app that I want to convert to this json below, where Id would be (status => status.getId().toString) and Text would be (status => status.getText())
I tried a few things, but I'm not happy with the result and was wondering whether there is a really efficient way to do this.
{
"Inputs": [{
"Id": "1",
"Text": "hello world"
},
{
"Id": "2",
"Text": "hello foo world"
},
{
"Id": "three",
"Text": "hello my world"
}]
}

Answer 1

In the end I created 2 java classes Input and InputsValue and leverage com.google.code.gson:gson:2.6.2 to convert object to Json string.
Why Java classes and not scala ones? Because gson supports List, where T is a custom object, very easily with Java classes compared to spray-io for example. I might be wrong, but that's what I found so far.
public class Input {
public Input(String id, String text) {
this.Id = id;
this.Text = text;
}
public String Id;
public String Text;
}
public class InputsValue {
public InputsValue(List<Input> inputs) {
this.Inputs = inputs;
}
public List<Input> Inputs;
}
val i1 = rdd.map(o => new Input(o.getId().toString, o.getText())).collect().toList;
val iv1: InputsValue = new InputsValue(i1)
val inputs = gson.toJson(iv1)