Making Sense of Equality in Ruby

When I discovered programming, I was bamboozled by the incredible amout of times you type the = key. Among other things, I couldn’t help but wonder: Why on earth was I supposed to use == instead of the more ‘natural’ = to express equality?

Turns out, it’s actually possible to make sense of this = madness! Here’s a cheat sheet to gain a better understanding of equality and the equal sign = in Ruby.

The = Madness

Here’s a short program:

def some_method(a)
  if a == 5
    a = 3
  elsif (1..10) === a
    a += 10
    [1, 2, 3][2] = 4

p some_method(5)  # => 3
p some_method(2)  # => 12
p some_method(15) # => 4

In the above code, we have no less than 5 different forms of =:

  • = in a = 3
  • += in a += 10
  • []= in [1, 2, 3][2] = 4
  • == in if a == 5
  • === in elsif (1..10) === a

Granted, the last situation is actually quite uncommon, but this profusion of = signs is still confusing. Let’s dive in!

What about =?

General Case: the Assignment Operator

In most cases, the = character refers to the assignment operator, ie. the operator used to assign a new value to a variable.

Therefore, a = 3 is Ruby’s way of informing the program that from now on, the a variable will now reference the Integer 3.

+= and the combined assignment operators

Ruby comes with 6 extremely handy “combined assignment operators”, that both operate on a variable and assign the result back to the variable.

  a = 2
  a += 10   # => 12
  a         # => 12

They are: +=, -=, *=, /=, %= (modulus), **= (exponent).

As with regular assignment, these operators have nothing to do with testing for equality.

Element Assignment #[]=

Another common Ruby construct is element assignment:

str = "hello"   # => "hello"
str[0] = "Y"    # => "Y"
str             # => "Yello"

Element assignment is actually a method (#[]=), available to Strings, Arrays and Hashes. It replaces the element referenced by its index. Again, it doesn’t check for equality.

Setter Methods #method=

Finally, you’ll encounter methods with a final =, which (spoiler alert) have nothing to do with equality. These methods are setter methods: they’re used to set a value to an instance variable.

class Cat
  def name

  def name=(name)   # setter method
    @name = name

kitty = = "Pixel"
p  # => "Pixel"

To sum up, = on its own is never used to check whether two elements are equal.

Instead, the == method will 99% of the times do the job 🌈

The == method and Value Equality

Examples of == usage

== usually checks whether the two objects on either side have the same value:

3 == 3  # => true

str = "hello"
str ==  "hello" # => true
str == "world"  # => false

{ a: 1, b: 2 } == { b: 2, a: 1 }    # => true

[1, 2, 3] == [1, 2, 3]              # => true
[1, 2, 3] == [3, 2, 1]              # => false

From these examples, we can deduce the following:

  • 2 identical integers have equal value (which sounds pretty obvious)
  • A String of value "hello" referenced by a variable has the same value than another String of value "hello"
  • Two String objects of different values ("hello" and "world") are unequal
  • Two Hashes are equal if their key-value pairs are identical (even if the order is different)
  • Arrays must also have an identical order to be considered equal

The == method

How does Ruby know how to compare two Hashes? Two Arrays? Two custom objects? The answer lies in the fact that == is not an operator but an instance method. You could actually write 2.==(2) instead of 2 == 2.

Additionally, == is a BasicObject instance method (ie. the class that contains every other object in Ruby), which means it’s available to any object.

However, we almost never use the default BasicObject#== method. Instead, each subclass is supposed to override the == method with a custom definition that suits its own needs.

For instance, the Hash#== method does not account for key-value pairs order. Here’s the corresponding entry in the Ruby documentation:

hsh == other_hash → true or false

Equality—Two hashes are equal if they each contain the same number of keys and if each key-value pair is equal to (according to Object#==) the corresponding elements in the other hash. The orders of each hashes are not compared.

Implications for Object Oriented Programming

Therefore, if you create a custom class and want to compare for “value equality” between two instances, you’ll need to implement your own custom == instance method.

Otherwise, == will still work, but probably not how you intended it:

class Cat
  def initialize(name)
    @name = name

kitty ="teacup")
kitten ="teacup")

p kitty == kitten # => false

Those two Cat objects have exactly the same value but kitty == kitten returns false - why?

The answer is that Ruby applied the BasicObject#== method, which has the same implementation than BasicObject#equal? (see next section), and basically checks whether two objects are equal (not whether they have the same value).

Indeed, how could Ruby know when 2 Cat objects are equal? It doesn’t know how to compare these two custom objects. We need to explicitly tell Ruby how to compare them, by including an instance method Cat#==:

class Cat
  attr_reader :name

  def initialize(name)
    @name = name

  def ==(other_cat)
    @name ==

kitty ="teacup")
kitten ="teacup")

p kitty == kitten # => true

Here, we chose to consider two cats identical if they have the same name value. It’s completely up to us: we could also have chosen their size or color. And it works! kitty == kitten now returns true 🍾

The equal? method

If we want to know whether two objects are actually the same object, we can use the object_id method and compare their IDs:

1.object_id   # => 3 

a = 1
a.object_id   # => 3 

1.object_id == a.object_id  # => true

Here, both 1 and a share the same ID 3, which means they are both pointing to the same object.

However, there’s an easier way to compare for object equality, using the BasicObject#equal? method:

a.equal?(1)   # => true 

equal? goes deeper than ==: it checks whether two objects have the same value and whether the two variables point to the same object.

str1 = "hi there"
str2 = "hi there"
str3 = str1

str1 == str2      # => true 
str1.equal? str2  # => false 
str1.equal? str3  # => true

str1, str2 and str3 all have the same value "hi there".

However, only str1 and str3 point to the same object (as expected, given that str3 is a copy of str1). Therefore, str1.equal? str3 returns true, while str1.equal? str2 returns false: str1 and str2 are different objects which happen to have the same value "hi there".

It’s pretty rare to use the equal? method, but it’s still useful to understand its behavior, especially when implementing custom classes.

What about ===?

If we know how to check for value equality (==) and for object equality (equal?), what is === good for? It’s actually used to check for range inclusion.

The === method compares two objects by checking whether the first element (the group) includes the second element. It’s rarely used explicitly.

(1..10) === 3       # true: 3 is included in the range from 1 to 10
(1..10) === 12      # false
('a'..'d') === 'c'  # true: 'c' is included in the range from a to d

=== is much more often used implicitly though, as this is the method used “under the hood” by case statements:

char = 'd'

case char
when ("a".."e")

# => "yay"

The above case statement is equivalent to the following code:

"yay" if ("a".."e") === char
# => "yay"

The eql? method

Here’s a last method to consider when thinking about equality. This one is pretty obscure, and there’s a good chance you’ll never even have to use it, but I’m aiming for exhaustivity here 👩‍🏫

Ruby Documentation

According to the Ruby documentation:

eql?(other) → true or false

The eql? method returns true if obj and other refer to the same hash key. This is used by Hash to test members for equality. For objects of class Object, eql? is synonymous with ==. Subclasses normally continue this tradition by aliasing eql? to their overridden == method, but there are exceptions. Numeric types, for example, perform type conversion across ==, but not across eql?.

Therefore, two Numeric objects could be equal (with ==), but different with eql?:

2 == 2.0        #=> true
2.eql?(2.0)     #=> false
2.0.eql?(2.0)   #=> true

But most of the time, two objects that are equal through == should return true when eql? is called.

When should you use the eql? method?

The answer is: probably never. However, it’s routinely used by Hash objects to check for hash key equality. As you know, hash keys must be unique, and this is enforced through eql?.

To check for this uniqueness, Hash objects use the eql? method to convert the keys through the hash method. This method returns a hash code for each key.

# Two different String objects with a value of "a" have the same hash:
'a'.hash	# => -3382060362238486793
'a'.hash	# => -3382060362238486793

If the two hash return values are equal (like for "a"), then only one element can serve as a Hash key.

hash = { 'a' => 1, 'a' => 2 }
# warning: key "a" is duplicated and overwritten
# => {"a"=>2} 

In this case, it’s interesting to note that both "a" elements share the same hash, but not the same object_id:

# Different object IDs
'a'.object_id	# => 33313480
'a'.object_id	# => 33412160

'a'.eql?('a')	# => true
'a'.equal?('a') # => false

Therefore, eql? returns true (as it compares the two hash) and equal? returns false (because their object_id are different).

Implications for Object Oriented Programming

If you happen to override eql? in a custom class, you’ll also need to override the hash method. If you don’t, instances of this class won’t work as keys in a Hash.

Indeed, objects used as keys must have a method named hash that returns a numeric hashcode for the key.

Key Take-Aways

To sum-up:

  • = on its own is not about equality: it’s an assignment operator.
  • == checks whether two objects have the same value.
  • The == method should be overriden when creating a custom class, in order to tell Ruby how to compare two custom objects.
  • The equal? method checks whether two variables point to the same object.
  • === checks for range inclusion, and is mostly used by case statements.
  • The eql? method is almost exclusively used by Hash objects to check the uniqueness of their hash keys.