Understanding Continuation Passing Style

[shonfeder]

I suspect the implementation in the OCaml Pro post has a typo, and that it should be

(* Tail-recursive CPS implementation of map *)
let rec map_cps f l k =
  match l with
  | [] -> k []
  | x :: r ->  let fx = f x in map_cps f r (fun r -> k (fx :: r))
;;

With the continuation applied to the cons in the closure.

EDIT: re:

@fdw

My functional background is mostly in Erlang where the idiomatic approach would be to have map construct the list in reverse order (which is trivially tail recursive) and then at the end reverse the list (which is a built in and also trivially tail recursive).

This is also very common in OCaml, and AFAIK, it is the more idiomatic approach if your aim is just to get a tail-call position version of these kinds of functions. That approach an accumulator argument rather than a continuation argument. See What is the use of Continuation Passing Style (CPS)? for related discussion.

(I also removed a link to a likely LLM spam source. Sorry about missing that initially!)