« Back to All Posts

Learning Haskell, Part 2 - More on Types

I'm learning Haskell by following Learn You a Haskell for Great Good. This is a series about learning something new.

NOTE: This and any other related posts are going to be rife with grammar and spelling errors as well as bad advice and misconceptions, so don't take them as a guide.

Type Information & Typeclasses

In GHCi there's a command (:t) to show the type signature of anything, values, variables, functions. Here's some examples of seeing value types:

ghci> :t 12 12 :: Num p => p

In the above it's saying that 12 is of type p. What's type p? Well p implements Num. What's Num? Num is a typeclass! From what I can gather, Haskell typeclasses are something like an interface in Java or a trait in Rust. Members of the Num typeclass act like numbers. Num is actually a subclass of Eq (since all numbers can test for equality) and Show (which lets it be printed as a string using the show function).

So the above says that 12 is of type p which is a member of Num.

ghci> :t 12.3 12.3 :: Fractional p => p

The above says 12.3 is of type p which is a member of Fractional which is a subclass of Num. Fractionals are non-integer numbers, including floating point numbers. Integers are all members of the Integral typeclass.

ghci> :t True True :: Bool ghci> :t 3 == 4 3 == 4 :: Bool

So True and False are a bit different. I think that, since they are reserved words, they directly implement their typeclass, Bool. So their definition doesn't abstract anything (meaning True :: Bool p => p isn't necessary). And since 3 == 4 evaluates to False, its type is also Bool.

ghci> :t 'Z' 'Z' :: Char

Pretty straight forward, 'Z' is a character, so it's a member of the Char typeclass.

ghci> :t "Hello, there" "Hello, there" :: [Char]

A little more interesting. Strings are actually just a list of characters, so instead of a String typeclass, strings are actually a list where each item is a member of the Char class.

ghci> :t [1,2,3] [1,2,3] :: Num a => [a]

Other lists are defined ina similar way. In the above, we have a list where each member is a member of Num.

Function Constraints

Using it, you can also see the constraints (input requirements and output types) of a function:

ghci> :t head head :: [a] -> a

The above says "head is a function that takes in a list whose members are of type a and returns a value of type a". What is type a? In this case it can be anything. Functions can also be typeclass-specific:

ghci> :t (+) (+) :: Num a => a -> a -> a

This has a few differences:

  1. When we use an infix function like + with :t we need to wrap it in parentheses.
  2. Similar to doing :t 12 in the previous section, we have that Num a =>. This is saying a is a member of Num.
  3. To the right of the => that we see a -> a -> a. This is the function definition (like [a] -> a in the head example).

I think this is difficult to read (at least at first). In Haskell the final -> a means that the function returns a value of type a. Everything before that is a parameter to the function. So in the above this means that + takes 2 parameters, both of type a and that it returns a value of type a.

So the full description of + is that it is a function that takes two parameters of type a and returns a value of type a where a is a member of Num.

There are a bunch of typeclasses and the basic ones are provided in the Prelude. As of the 2010 version of Haskell, this is the documentation for the predefined types and classes.