Maybe

Erlang introduced maybe expressions to deal with excessively nested case statements.

Take this nested case statement:

case foobar():
  #('ok, v):
    case barfoo(v):
      #('ok, v2):
        case final_foo(v2):
          #('ok, v3): v3
          #('error, _err) = e: e
      #('error, _err) = e: e
  #('error, _err) = e: e

This may seem like a corner case but it is more common than you think since the ok/error tuple pattern is very common in Erlang.

Let's see how we can refactor that with a maybe expression.

maybe:
  #('ok, v) ?= foobar()
  #('ok, v2) ?= barfoo(v)
  #('ok, v3) ?= final_foo(v2)
  v3
else:
  #('error, _err) = e: e

We can see that this is a lot easier to read. Not the ?= which is called the conditional match operator. If the result of the function call matches the ok tuple then it continues. If not, it short circuits to the else and goes from there.