Over a million developers have joined DZone.

Scala's Self Type Annotations

DZone's Guide to

Scala's Self Type Annotations

This overview covers self type annotations/references in Scala and their usefulness for dependency injection while also touching on the cake pattern that uses them.

· Java Zone ·
Free Resource

Download Microservices for Java Developers: A hands-on introduction to frameworks and containers. Brought to you in partnership with Red Hat.

Self type annotations/references allow you to redefine this and is a way to declare the dependencies required by a component. Using a trait mixin, you can inject various implementations of dependencies.

In layman terms, self type annotations/references are to ensure that the class cannot be instantiated without mixing in the trait explicitly specified in the notation, and its members can be used in the class exactly like extending/mixing in a trait.

The syntax goes like this:

trait A{
    ...some defs and vals

class B{
    this: A =>
    .....rest of the class
    .....here we can use def and vals of A too

In this example, if we try to instantiate our class B, it won’t:

val b = new B //gives error: class B cannot be instantiated because it does not conform to its self-type B with A

It needs A to be mixed in to be instantiated:

val b = new B with A

Now, what’s “this: A =>”

It's the self type annotation, and you can use ‘self’ or any identifier instead of ‘this’ (more details). Using A here means class B started with mixing in trait A.


The point here is the advantage that we get from using self types instead of mixing in a trait. So let's consider the following snippet that displays a class extending a trait:

trait ReaderWriter{
    def read: String = "data from file"
    def write: String = "written to a file"
class Service extends ReaderWriter
    def reading = read()
    def writing = write()
val service = new Service

The problem here is that we are stuck with the ReaderWriter trait forever in case of class Service. We can’t change the extended trait without changing our class definition


Now here, we’ll use self type annotation:

trait ReadWrite{
    def read: String = "data from file"
    def write: String = "data written to file"
class Service{
    this: ReaderWriter =>
    def reading: String = read()
    def writing : String = write()

We can instantiate our class Service by mixing in ReadWrite:

val service = new Service with ReadWrite

Now if some trait ReadWriteDB extends ReaderWriter...

trait ReadWriteDB extends ReadWrite {
    def read: String = "data from DB"
    def write: String = "data written to DB"

... then we can instantiate our service class with ReadWriteDB:

val service = new Service with ReadWriteDB

Hence we can mixin child traits explicitly specified in class definitions at the time of instantiation.

Using Multiple Traits in Self References

One more trait:

trait UpdateDelete{
    def update: String = "update file data"
    def delete: String = "deleted data from file"

Now the new class definition

class Service{
    this: ReadWrite with UpdateDelete =>
    .....class definition

Class instantiation

val service = new Service with ReadWrite with UpdateDelete

So that was it for self type annotations/references.


Self type annotations/references are used in cake patterns, which we didn't cover above. The cake pattern is a means of dependency injection, along with the Constructor Pattern and Google Guice. We’ll discuss those future posts.

Hope it was helpful!

Download Building Reactive Microservices in Java: Asynchronous and Event-Based Application Design. Brought to you in partnership with Red Hat

java ,scala ,self type annotations ,dependency injection ,tutorial

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}