If you've worked with Ruby, you've likely encounter attr methods such as attr_accessor, attr_reader, and attr_writer sprinkled at the top of the class.

To truly understand what each of these attr methods are and how to use them, we must understand the very basics of how data is stored in a Ruby class. (If you already have the basics down, skip to here!)

How is data stored and access in a Ruby class?

Let’s say we’re creating a Ruby on Rails web app for a veterinary hospital 😸🐶. We want to store information about each dog patient. To start with, we will need a Dog class that stores name.

In Ruby, this class may look something like the following:

  class Dog
    def initialize(name)
      @name = name
    end
  end

Here, we have a Dog class. This Dog class has an initialize method which expects a name parameter. Within this initialize method we instantiate an instance variable @name which stores the name
parameter that is passed in when a Dog instance is created.

What is an instance variable?

In Ruby, instance variables are prefixed with @. You may recognize instance variables being used to pass data between the controller and the view in a Ruby on Rails app. (Under the hood, Ruby on Rails is set up so that an instance variable in a controller method is accessible in the corresponding view.)

Even outside of a Rails-y context, instance variables are used to store data for a regular Ruby class.

An instance variable can be accessed by any method within the class.

Here's an Example:

  class Dog
    def initialize(name)
      @name = name
    end

    def barks
      "#{@name} says WOOF" # This method has access to @name
    end
  end

The catch here is that instance variables are accessible by any method within a class, but are NOT accessible outside the class.

If instantiate a Dog instance and try to call .name, we get a NoMethodError:

dog = Dog.new('Tomi')
#=> <Dog:0x00007fb72cd0aa30 @name="Tomi">
dog.name
#=> NoMethodError: undefined method `name` for #<Dog:0x00007fb72eab6bb0 @name="Tomi">

If we want to read the @name value stored on a Dog instance, we'll need to access it through a method.

Let's define a simple method that returns @name.

  class Dog
    def initialize(name)
      @name = name
    end

    # Method that returns @name
    def name
      @name
    end
  end

Now we are able to access the value of @name by calling the name method!

dog = Dog.new('Tomi')
dog.name
#=> "Tomi"

This simple name method we just defined is what we call a getter method.

A getter method returns the value of an instance variable.

In the previous example, we had the name method simply return @name. However, if you wanted @name to be returned in a certain format, you can easily achieve that by customizing your getter method.

Let's write a custom getter method that returns dog.name in all caps:

  class Dog
    def initialize(name)
      @name = name
    end

    def name
      @name.upcase
    end
  end

dog = Dog.new('Tomi')
#=> #<Dog:0x00007fb7475df010 @name="Tomi">
dog.name
#=> "TOMI"

Notice that here, @name itself is not modified. We’ve simply defined a method that returns @name in the desired format.

Say we initialize a Dog instance but realize we have made a typo! If we try to reset name, we'll get a NoMethodError:

dog = Dog.new('Tomu') # Whoops!
dog.name = 'Tomi'
#=> NoMethodError: undefined method `name=' for #<Dog:0x00007fb7483dd9a0 @name="Tomu">

Similar to how we need to define a getter method to read the @name value, we also need a method to set @name a value. Though we're setting the value of @name on initialize, we don't have anything in place to change it later.

We need a setter method!

A setter method sets the value of an instance variable.

Let's create a setter method that allows us to change the name attribute:

  class Dog
    def initialize(name)
      @name = name
    end

    # Getter method
    def name
      @name
    end

    # Setter method
    def name=(name)
      @name = name
    end
  end

Notice that a setter method follows the format def attribute=(attribute). Now with a setter method, we can easily update the value of @name:

dog = Dog.new('Tomuu')
dog.name = 'Tomi'
dog
#=> #<Dog:0x00007fb7483dd9a0 @name="Tomi">

What is attr_accessor, attr_writer, and attr_reader?

Now that we understand the basics of how data is stored and accessed in a Ruby class, let’s revisit the original question. What are these attr methods? Well... they're ✨magic✨!

Getting and setting an attribute of a class is a pretty basic operation. If we want a Dog class that has name, breed, age, and any number of attributes, we need a getter and setter method for each of these attributes. Writing methods for something very simple can quickly get very repetitive.

You might've heard of Ruby as a very developer-friendly language. Aside from being very human-readable, Ruby aims to make our lives easier via metaprogramming. In other words, as a programming language, Ruby takes care of a lot of things so that developers don't have to! attr methods are one of those things.

attr_accessor provides a shorthand way to define getter and setter methods for specified attributes.

Desired attributes are passed in to attr methods as a symbol:

  class Dog

    attr_accessor :name, :breed, :age

    def initialize(name, breed, age)
      @name = name
      @breed = breed
      @age = age
    end

  end

With a single attr_accessor :name, :breed, :age line, we were able to define six methods!

attr_accessor combines attr_reader and attr_writer. If you only want to shorthand a getter method, you can use attr_reader. If you want to shorthand a writer method, you can use attr_writer.

• attr_accessor - creates a getter and setter method
• attr_reader - creates a getter method only
• attr_writer - creates a setter method only

It might make sense to use attr_reader or attr_writer in situations where we want to customize one method, but not the other.

Example: We want dog.age to return @age in a certain format, but don't want to change how @age is stored.

  class Dog

    attr_accessor :name, :breed
    attr_writer :age # Default setter method to set @age

    def initialize(name, breed, age)
      @name = name
      @breed = breed
      @age = age
    end

    # Defining our own getter method to return a formatted age
    def age
      "Age #{@age}"
    end

  end

dog = Dog.new('Tomi', 'pug', 2)
#=> #<Dog:0x00007fb72debcdf0 @age=2, @breed="pug", @name="Tomi">
dog.age
#=> "Age 2"
dog
#=> #<Dog:0x00007fb72debcdf0 @age=2, @breed="pug", @name="Tomi">

Example: We want @age to always be set in a certain format.

  class Dog

    attr_accessor :name, :breed
    attr_reader :age # Default getter method to read @age

    def initialize(name, breed, age)
      @name = name
      @breed = breed
      @age = "Age #{age}" # Set @age to age in desired format.
    end

    # Defining our own setter method so @age is always set in desired format
    def age=(age)
      @age = "Age #{age}"
    end

  end

dog = Dog.new('Tomi', 'pug', 2)
#=> #<Dog:0x00007fb73383efe0 @age="Age 2", @breed="pug", @name="Tomi">
dog.age = 3
#=> 3
dog
#=> #<Dog:0x00007fb73383efe0 @age="Age 3", @breed="pug", @name="Tomi">

Conclusion

• Instance variables hold data in a Ruby class instance.
• Instance variables can only be directly accessed within a Ruby class.
• To access an instance variable outside of a Ruby class, we must define methods.
• attr_reader, attr_writer, and attr_accessor are convenient shorthands for writing getter and setter methods.