Handling JSON

Scalatra makes JSON handling quick and easy. By adding a few library imports and several lines of code, you can get automatic JSON serialization and deserialization for any Scala case class.

Setup

Let’s say you’ve generated a brand-new Scalatra project, and accepted all the defaults, except you’ve used “FlowersController” instead of “MyScalatraServlet” for the servlet name. Then you trim things down so you’ve got a bare controller like this one:

package com.example.app

import org.scalatra._

class FlowersController extends ScalatraServlet {

}

This assumes that your ScalatraBootstrap file, in src/main/scala/ScalatraBootstrap.scala, looks like this:

import com.example.app._
import org.scalatra._
import javax.servlet.ServletContext

class ScalatraBootstrap extends LifeCycle {
  override def init(context: ServletContext) {
    context.mount(new FlowersController, "/*")
  }
}

The data infrastructure

We can quickly add a Flower model underneath our controller class, so that we’ve got a data model to play with:

// A Flower object to use as a faked-out data model
case class Flower(slug: String, name: String)

If you wanted your application to be more structured, you might put that in its own file in a models directory. However, we’re not too concerned about that for the purposes of this tutorial, so let’s just put it in FlowersController.scala for now.

Now that you’ve got a model, let’s add a data store. Instead of taking a detour to talk about persistence libraries, we can just make a List of Flowers, and hang them off of an object. Put this at the bottom of FlowersController.scala:

object FlowerData {

  /**
   * Some fake flowers data so we can simulate retrievals.
   */
  var all = List(
      Flower("yellow-tulip", "Yellow Tulip"),
      Flower("red-rose", "Red Rose"),
      Flower("black-rose", "Black Rose"))
}

Let’s see if everything is working. Add the following action to your FlowersController:

get("/"){
  FlowerData.all
}

If you hit your local server on http://localhost:8080, you should see the following output:

List(Flower(yellow-tulip,Yellow Tulip), Flower(red-rose,Red Rose), Flower(black-rose, Black Rose))

Defaulting to JSON output

What’s going on here? Scalatra has converted the FlowerData.all value to a string and rendered its Scala source representation as the response. This is the default behaviour, but in fact we don’t want things to work this way - we want to use JSON as our data interchange format.

Let’s get that working.

In order to use Scalatra’s JSON features, we’ll need to add a couple of library dependencies so that our application can access some new code.

The first thing you’ll need is Scalatra’s JSON handling library. The second thing you’ll need is json4s, which is a unified wrapper around the various Scala JSON libraries. We’ll use the json4s-jackson variant, which uses the jackson library as the basis of its json support.

In the root of your generated project, you’ll find a file called build.sbt. Open that up, and add the following two lines to the libraryDependencies sequence, after the other scalatra-related lines:

  "org.scalatra" %% "scalatra-json" % "2.8.2",
  "org.json4s"   %% "json4s-jackson" % "4.0.1",

Restart sbt to download the new jars.

Add the following imports to the top of your FlowersController file, in order to make the new JSON libraries available:

// JSON-related libraries
import org.json4s.{DefaultFormats, Formats}

// JSON handling support from Scalatra
import org.scalatra.json._

Now we can add a bit of magic to the FlowersController. Putting this line of code right underneath the controller class definition will allow your controller to automatically convert Scalatra action results to JSON:

  // Sets up automatic case class to JSON output serialization, required by
  // the JValueResult trait.
  protected implicit lazy val jsonFormats: Formats = DefaultFormats

To serialize fractional numbers as BigDecimal instead of Double, use DefaultFormats.withBigDecimal:

  protected implicit lazy val jsonFormats: Formats = DefaultFormats.withBigDecimal

Watch out! That lazy on the protected implicit lazy val is necessary, not optional, when using JacksonJsonSupport.

Just like its Sinatra forebear, Scalatra has a rich set of constructs for running things before and after requests to your controllers. A before filter runs before all requests. Add a before filter to set all output for this controller to set the content type for all action results to JSON:

  // Before every action runs, set the content type to be in JSON format.
  before() {
    contentType = formats("json")
  }

Now mix JacksonJsonSupport into your servlet so your controller declaration looks like this:

class FlowersController extends ScalatraServlet with JacksonJsonSupport {

Your code should compile again at this point. Refresh your browser at http://localhost:8080, and suddenly the output of your / action has changed to JSON:

[{"slug":"yellow-tulip","name":"Yellow Tulip"},{"slug":"red-rose","name":"Red Rose"},{"slug":"black-rose","name":"Black Rose"}]

The JacksonJsonSupport trait which we mixed into the controller, combined with the implicit val jsonFormats, are now turning all Scalatra action result values into JSON.

Receiving JSON

Inbound JSON works in a similar way.

When a json request comes, which is identified by the Content-Type header or format param then scalatra will try to read the json body into an AST.

You can then extract a case class from this json AST.

case class Person(id: Int, name: String)

post("/create") {
  parsedBody.extract[Person]
}

Manipulating the JSON

You can transform the JSON AST before when it’s being received by overriding the method transformRequestBody

protected override def transformRequestBody(body: JValue): JValue = body.camelizeKeys

Likewise you can also transform the JSON AST right before sending it by overriding the method transformResponseBody

protected override def transformResponseBody(body: JValue): JValue = body.underscoreKeys