How to use error in a Haskell function?

Question

I am using GHCI 7.10.3 and I'm getting error in a simple fatorial code.

I would like to do something like this:

fatorial n
    | n == 0  = 1
    | n > 0 = n * fatorial(n-1)
    | otherwise = error "My error"

But when fatorial -4 was called, the output was:

:21:1: Non type-variable argument in the constraint: Num (a -> a) (Use FlexibleContexts to permit this) When checking that ‘it’ has the inferred type it :: forall a. (Num a, Num (a -> a), Ord a) => a -> a

My code works fine without the last line. So how can I use error message in haskell?

That is because Haskell has interpreted it as factorial - 4 (so a subtraction), you should use factorial (-4). — Willem Van Onsem
I don't really believe "My code works fine without the last line". What makes you think that? — amalloy

Willem Van Onsem Willem Van Onsem · Accepted Answer · 2018-08-07T20:14:12

Well the error is a type error, so that means that Haskell thinks that what you wrote makes no sense (from a type system point of view).

It interprets the - as the "binary minus operator", like:

--         v operator
factorial  -  4
-- ^  operand ^

So Haskell thinks you want to subtract 4 from factorial, but it does not see how factorial is a Number, hence the error. Strictly speaking, one can make functions Numbers, as long as one implements the Num typeclass (as well as the Eq and Show typeclasses) we are fine.

If you want to use negative number literals in such function call, you need to use brackets, like:

factorial (-4)

This then produces:

Prelude> fatorial (-4)
*** Exception: My error
CallStack (from HasCallStack):
  error, called at <interactive>:5:19 in interactive:Ghci1

So now it raises your error "My error" (see the first output line).

How to use error in a Haskell function?

1 Answers