Safe and elegant recursion in Elm

This package provides tools to use recursion safely in elm without sacrificing the elegance of general recursion.

Example

Imagine we have a generic tree type that we want to write a map function for:

type Tree a
    = Leaf a
    | Node (Tree a) (Tree a)

The safe solution to implement a map function using this library would look like the following:

mapTree : (a -> b) -> Tree a -> Tree b
mapTree f initTree = 
    runRecursion (\tree -> 
        case tree of
            Leaf a -> 
                base <| Leaf (f a)

            Node l r -> 
                recurseThen l (\newL -> recurseThen r (\newR -> base (Node newL newR)))
    ) initTree

For more on the types and functions involved and details on how this works check out the Recursion module.

This module pairs extremely well with the elm-review rule NoUnoptimizedRecursion.

Can't I just...

Refactor my recursive code by hand to be Tail Call Optimized?

Sure! But for non-trivial recursions like the mapTree example above or recursive data structures involving lists you'll find yourself needing to write parallel data structures just to keep track of where you are in a step of the recursion. You quickly lose the elegance of recursion in all the bookkeeping.

Use elm-tail-recursion?

Sure! But again for non-trivial recursions you will require custom data structures and are only in a slightly better situation than writing tail call optimized code by hand.

Use trampoline?

Sure! But it's the same situation as above.

Additional Information

This library is based on the excellent paper "Stackless Scala With Free Monads" by Bjarnason. Unfortunately, we don't have existentially quantified types in Elm so we can not directly implement the Trampoline monad from the paper as a dataype.

We instead had to choose a different representation that unfortunately does not allow us to "disallow the construction of deeply nested left binds". To mitigate the risk of a user accidentally creating a stack of left binds, we provide safe implementations of folds and traversals for common data structures in Recursion.Fold and Recursion.Traverse.