Functions as data structures

So far, we have learned how to use functions to work with or transform other functions to process data structures or to create data types. Now, we'll finish this chapter by showing you how a function can actually implement a data type by itself, becoming a sort of container of its own. In fact, this is a basic theoretical point of the lambda calculus (and if you want to learn more, look up Church Encoding and Scott Encoding), so we may very well say that we have come back to where we began this book, at the origins of FP! We will start with a detour that considers binary trees in a different functional language, Haskell, and then move on to implementing trees as functions, but in JavaScript; this experience will help you work out how to deal with other data structures.