Fold and foldLeft method difference

Question

I am not sure what is the difference between fold and foldLeft in Scala.

The question Difference between fold and foldLeft or foldRight? has an

Answer 1

General difference

Here are the prototypes of the methods

fold[A1 >: A](z: A1)(op: (A1, A1) ⇒ A1): A1
foldLeft[B](z: B)(f: (B, A) ⇒ B): B

So, for fold the result is of type A1 >: Ainstead of any B. Moreover, as specified in the doc, for fold the order is not

About your error

When typing scala> Array("1","2","3").fold(0)(_ + _.toInt) you assume that 0, an int is a subtype of String. This is why the compiler throws an error.

About the weird z in fold

Here we have to see the implementation of fold to understand what happens. Here is what we get:

def fold[A1 >: A](z: A1)(op: (A1, A1) => A1): A1 = foldLeft(z)(op)

So basically, fold is an implementation of foldleft with a constraint on the output type. We can now see that z will in practice be used the same way as in foldleft. So we can just conclude that this comment was made because nothing assures that behavior in future implementations. We can already see it now, with parallels:

def fold[U >: T](z: U)(op: (U, U) => U): U = {
  executeAndWaitResult(new Fold(z, op, splitter))
}