Why does length return 1 for a tuple with 2 elements, and gives an error for a tuple with more elements?

by

You have encountered a Haskell cause célèbre that has sparked much discussion and gnashing of teeth.

Basically, for the purposes of Foldable (the typeclass that provides length), 2-tuples are not considered a container of two elements, but a container of one element accompanied by some context.

You can extract a list of elements of type a from any Foldable a. Notice that for 2-tuples the type variable of the Foldable is that of the second element of the tuple, and it can be different from the type of the first element.

If you had a ('c',2) :: (Char,Int) tuple, it would be no mystery that you couldn’t extract two Ints in that case! But when the types are equal it becomes confusing.

As for why length (1::Int, 1::Int, 1::Int) fails, 3-tuples don’t have a Foldable instance defined, but perhaps they should have one, for consistency. 3-tuples would also have length 1.

By the way, the Identity functor, that could be considered a kind of 1-tuple, is also Foldable and of course has length 1 as well.

Should the Foldable instance for tuples exist at all? I think the underlying philosophy in favor of yes is one of, shall we call it, “plenitude”. If a type can be made an instance of a typeclass in a well defined, lawful way, it should have that instance. Even if it doesn’t seem very useful and, in some cases, may be confusing.

More Related Contents:

  1. How to run a for loop in Haskell
  2. Getting started with Haskell
  3. How is this fibonacci-function memoized?
  4. runST and function composition
  5. Explain this chunk of haskell code that outputs a stream of primes
  6. How to create a polyvariadic haskell function?
  7. What does (f .) . g mean in Haskell?
  8. foldl versus foldr behavior with infinite lists
  9. What is the purpose of Rank2Types?
  10. Finite comprehension of an infinite list
  11. Are there pronounceable names for common Haskell operators? [closed]
  12. Split a number into its digits with Haskell
  13. What indentation is required for a case statement within a let statement?
  14. Currying subtraction
  15. What is the difference between Int and Integer?
  16. Is Haskell truly pure (is any language that deals with input and output outside the system)?
  17. Euler 43 – is there a monad to help write this list comprehension?
  18. Is there a way to place some impure code inside pure functions?
  19. How to increment a variable in functional programming?
  20. What part of Hindley-Milner do you not understand?
  21. Why is GHC complaining about non-exhaustive patterns?
  22. Why are side-effects modeled as monads in Haskell?
  23. foldl Implementation with Runtime Errors
  24. Exception handling in Haskell
  25. How to make a type with restrictions
  26. Given a Haskell type signature, is it possible to generate the code automatically?
  27. Functional lenses
  28. How to reuse a type variable in an inner type declaration
  29. How does Haskell printf work?
  30. Sampling sequences of random numbers in Haskell

Leave a Comment