import play.libs.Json;


public JsonNode createJson() {
    // {"id": 33, "values": [3, 4, 5]}
    ObjectNode rootNode = Json.newObject();
    ArrayNode listNode = Json.newArray();

    long values[] = {3, 4, 5};
    for (long val: values) {
        listNode.add(val);
    }
    
    rootNode.put("id", 33);
    rootNode.set("values", listNode);
    return rootNode;
}

import play.libs.Json;
// (...)

Loading a file from your public folder

// Note: "app" is an play.Application instance
JsonNode node = Json.parse(app.resourceAsStream("public/myjson.json"));

Load from a string

String myStr = "{\"name\": \"John Doe\"}";
JsonNode node = Json.parse(myStr);

In the following examples, json contains a JSON object with the following data:

[
  {
    "name": "John Doe",
    "work": {
      "company": {
        "name": "ASDF INC",
        "country": "USA"
      },
      "cargo": "Programmer"
    },
    "tags": ["java", "jvm", "play"]
  },
  {
    "name": "Bob Doe",
    "work": {
      "company": {
        "name": "NOPE INC",
        "country": "AUSTRALIA"
      },
      "cargo": "SysAdmin"
    },
    "tags": ["puppet", "ssh", "networking"],
    "active": true
  }
]

Get the name of some user (unsafe)

JsonNode node = json.get(0).get("name"); // --> "John Doe"
// This will throw a NullPointerException, because there is only two elements
JsonNode node = json.get(2).get("name"); // --> *crash*

Get the user name (safe way)

JsonNode node1 = json.at("/0/name"); // --> TextNode("John Doe")
JsonNode node2 = json.at("/2/name"); // --> MissingNode instance
if (! node2.isMissingNode()) {
    String name = node2.asText();    
}

Get the country where first user works

JsonNode node2 = json.at("/0/work/company/country"); // TextNode("USA")

Get every countries

List<JsonNode> d = json.findValues("country"); // List(TextNode("USA"), TextNode("AUSTRALIA"))

Find every user that contains the attribute "active"

List<JsonNode> e = json.findParents("active"); // List(ObjectNode("Bob Doe"))

By default, Jackson (the library Play JSON uses) will try to map every public field to a json field with the same name. If the object has getters/setters, it will infer the name from them. So, if you have a Book class with a private field to store the ISBN and have get/set methods named getISBN/setISBN, Jackson will

• Create a JSON object with the field "ISBN" when converting from Java to JSON
• Use the setISBN method to define the isbn field in the Java object (if the JSON object has a "ISBN" field).

Create Java object from JSON

public class Person {
    String id, name;
}

JsonNode node = Json.parse("{\"id\": \"3S2F\", \"name\", \"Salem\"}");
Person person = Json.fromJson(node, Person.class);
System.out.println("Hi " + person.name); // Hi Salem

Create JSON object from Java object

// "person" is the object from the previous example
JsonNode personNode = Json.toJson(person) 

Creating a JSON string from a JSON object

// personNode comes from the previous example
String json = personNode.toString();
// or 
String json = Json.stringify(json);

JSON pretty printing

System.out.println(personNode.toString());
/* Prints:
{"id":"3S2F","name":"Salem"}
*/   

System.out.println(Json.prettyPrint(personNode));
/* Prints:
{
  "id" : "3S2F",
  "name" : "Salem"
}
*/