A service object is a Ruby object that performs a single action. It encapsulates a process in your domain or business logic. Imagine that you need to create a book instance in an imaginary library application; in a plain Rails app, you'd do the following:
class BookController < ApplicationController
def create
Book.new(*args)
end
end
This is fine for simple things. However, as the app grows, you may end up with lots of boilerplate surrounding it:
class BookController < ApplicationController
def create
default_args = { genre: find_genre(), author: find_author() }
Book.new(attrs.merge(default_args))
end
private
def find_genre
// ...
end
def find_author
// ...
end
end
Service objects allow you to abstract this behavior into a separate class. Then, your code becomes simple again:
class BookController < ApplicationController
def
BookCreator.create_book
end
end
Why You Need Service Objects
Rails is designed to natively support the MVC (e.g., models, controllers, views, and helpers) organizational structure. This structure is adequate for simple applications. However, as your application grows, you may begin to see domain/business logic littered across the models and the controller. Such logics do not belong to either the controller or the model, so they make the code difficult to re-use and maintain. A Rails service object is a pattern that can help you separate business logic from controllers and models, enabling the models to be simply data layers and the controller entry point to your API.
We get a lot of benefits when we introduce services to encapsulate business logic, including the following:
Lean Rails controller - The controller is only responsible for understanding requests and turning the request params, sessions, and cookies into arguments that are passed into the service object to act. The controller then redirects or renders according to the service response. Even in large applications, controller actions using service objects are usually not more than 10 lines of code.
Testable controllers - Since the controllers are lean and serve as collaborators to the service, it becomes really easy to test, as we can only check whether certain methods within the controller are called when a certain action occurs.
Ability to test business process in isolation - Services are easy and fast to test since they are small Ruby objects that have been separated from their environment. We can easily stub all collaborators and only check whether certain steps are performed within our service.
Re-usable services - Service objects can be called by controllers, other service objects, queued jobs, etc.
Separation between the framework and business domain - Rails controllers only see services and interact with the domain object using them. This decrease in coupling makes scalability easier, especially when you want to move from a monolith to a microservice. Your services can easily be extracted and moved to a new service with minimal modification.
Creating a Service Object
First, let’s create a new BookCreator in a new folder called app/services for an imaginary library management application:
$ mkdir app/services && touch app/services/book_creator.rb
Next, let’s just dump all our logic inside a new Ruby class:
# app/services/book_creator.rb
class BookCreator
def initialize(title:, description:, author_id:, genre_id:)
@title = title
@description = description
@author_id = author_id
@genre_id = genre_id
end
def create_book
Boook.create!(
title: @title
description: @description
author_id: @author_id
genre_id: @genre_id
)
rescue ActiveRecord::RecordNotUnique => e
# handle duplicate entry
end
end
end
Then, we can call the service object in the controller or anywhere within the application:
class BookController < ApplicationController
def create
BookCreator.new(title: params[:title], description: params[:description], author_id: params[:author_id], genre_id: params[:genre_id]).create_book
end
end
Service Object Syntactic Sugar
We can simplify the BookCreator.new(arguments).create
chain by adding a class method that instantiates the BookCreator
and calls the create
method for us:
# app/services/book_creator.rb
class BookCreator
def initialize(title:, description:, author_id:, genre_id:)
@title = title
@description = description
@author_id = author_id
@genre_id = genre_id
end
def call(*args)
new(*args).create_book
end
private
def create_book
Boook.create!(
title: @title
description: @description
author_id: @author_id
genre_id: @genre_id
)
rescue ActiveRecord::RecordNotUnique => e
# handle duplicate entry
end
end
end
In the controller, the book creator can now be called as follows:
class BookController < ApplicationController
def create
BookCreator.call(
title: params[:title],
description: params[:description],
author_id: params[:author_id],
genre_id: params[:genre_id])
end
end
To keep our code DRY(Don't Repeat Yourself) and reuse this behavior with other service objects, we can abstract the call
method into a base ApplicationService
class that every service object will inherit from:
class ApplicationService
self.call(*args)
new(*args).call
end
end
With this code, we can refactor the BookCreator
to inherit from the ApplicationService
:
# app/services/book_creator.rb
class BookCreator < ApplicationService
def initialize(title:, description:, author_id:, genre_id:)
@title = title
@description = description
@author_id = author_id
@genre_id = genre_id
end
def call
create_book
end
private
def create_book
# ...
end
end
Creating Service Objects Using the BusinessProcess Gem
With the BusinessProcess gem, you don't have to create a base application service class or define the initialize
method because the gem has all these configurations built into it. Your service object just has to inherit from the BusinessProcess::Base
.
In your gem file, add the following:
gem 'business_process'
And then run the bundle
command in your terminal
class BookCreator < BusinessProcess::Base
# Specify requirements
needs :title
needs :description
needs :author_id
needs :genre_id
# Specify process (action)
def call
create_book
end
private
def create_book
# ...
end
end
Guides to Creating Good Service Objects
One public method
A service object is supposed to perform one business action and do it well, so it should only expose a single public method for doing that. Other methods should be private and called by the public method. You can choose to name the public method whatever you prefer, as long as the naming is consistent across all service objects. In our example, we have named it call
. Other conventional names are perform
and execute
.
Name Service Objects According to the Role They Perform
The name of a service object should indicate what it does. There is a popular way of naming service objects with words ending with “or” and "er". For instance, if the job of the service object is to create a book, name it BookCreator, and if the job is to read a book, name it BookReader.
Do not instantiate service objects directly
Use abstractions like the syntactic sugar pattern or gems like BusinessProcess to shorten the notation of calling service objects. Using this approach would allow you simplify BookCreator.new(*args).call
or BookCreator.new.call(*args)
into BookCreator.call(*args),
which is shorter and more readable.
Group service objects in namespaces
Introducing service objects, especially in a large application, means that you would grow from one service object to tens of service objects. To improve code organization, it is a good practice to group common service objects into namespaces. In the library application, for instance, we would group all book-related services together and group all author-related services in a separate namespace. Our folder structure will now look like this:
services
├── application_service.rb
└── book
├── book_creator.rb
└── book_reader.rb
Our service objects would look like this:
# services/book/book_creator.rb
module Book
class BookCreator < ApplicationService
...
end
end
# services/twitter_manager/book_reader.rb
module Book
class BookReader < ApplicationService
...
end
end
Our calls will now become Book::BookCreator.call(*args) and Book::BookReader.call(*args)
.
One responsibility per service object
Having a service object that does more than one thing goes against the "business action" mindset of service objects. Conventionally, having a generic service object that performs multiple actions is discouraged. If you want to share code among service objects, create a base or helper module and use mixins to include in your service objects.
Rescue exceptions and raise custom exceptions
The purpose of a service object is to encapsulate implementation details inside it, such as interactions between third-party services or libraries or database manipulation with Rails ActiveRecord. If an error occurs, such as ActiveRecord::RecordNotUnique, while interacting with an ActiveRecord, the service needs to rescue the exception properly. Errors should not be allowed to propagate up the call stack. If it can't be handled within the rescue block, raise a custom-defined exception specific to that service object.
Conclusion
The service object pattern can greatly improve your application's overall design as you add new features to your application. It will make your codebase more expressive, easier to maintain, and less painful to test.