How do `Hash#reject!` and `Hash#reject` differ from `Hash#delete_if`?

Question

How do reject! and reject differ from delete_if for a Hash in Ruby? Can anyone explain the differences between them with simple code snippets?

Answer 1

Since the other answers are referring to Array#delete_if and not Hash#delete_if, which seems to be what you are asking, I thought I should clarify.

As others have pointed out, reject and reject! differ in that reject! version modifies the hash in-place, while reject creates a new hash. Meanwhile delete_if is almost the same as reject!.

In fact, for an Array, reject! and delete_if are exactly the same.

However, for a Hash, they are slightly different. reject! returns nil if no changes were made, or the hash if changes were made. delete_if always returns the hash.

hash = {a: 1, b: 2, c: 3}

return_value = hash.delete_if {|k, v| v > 100}
# hash is unchanged, return_value is {a: 1, b: 2, c: 3}

return_value = hash.reject! {|k, v| v > 100}
# hash is unchanged, return_value is nil

So if you wanted to check whether changes were made to the hash after deleting the elements, you could use reject! and check the return value.

Answer 2

If you read the docs it tells you that reject! is "Equivalent to Array#delete_if"

reject and reject! differ in that the bang (reject!) causes the changes to happen directly on the array you're working with, whereas reject will leave the array you're working with untouched, but will return a new array.

a = [ "a", "b", "c" ]
b = a.reject {|x| x >= "b" }   #=> a is untouched, but b is ["a"]
a.reject! {|x| x >= "b" }   #=> a is now modified and is ["a"]