Question

Like this:

> (my-append (list 1 2) 3)
'(1 2 3)

I know append in racket is actually to concatenate two list.
And cons just add an element to the head of a list instead of tail

Does anyone have ideas about this?

Answer 1

In Pyton, the append() method modifies the list in-place:

lst = [1, 2]
lst.append(3)

lst
=> [1, 2, 3]

Racket’s lists are immutable by default, the closest thing to an in-place append() requires you to use mutable lists, and you have to pack the element to be appended in a list of its own:

(require scheme/mpair)

(define lst (mlist 1 2))
(mappend! lst (mlist 3))

lst
=> (mlist 1 2 3)

Notice that using immutable lists with the append procedure will produce a new list, leaving the original list untouched – so it wouldn’t be “like Python”:

(define lst (list 1 2))
(append lst (list 3))   ; this returns the new list '(1 2 3)

lst
=> (list 1 2)

In fact, Scheme’s append procedure behaves just like Python’s + operation between lists:

lst = [1, 2]
lst + [3]    # this returns the new list [1, 2, 3]

lst
=> [1, 2]

Come to think of it, Python’s append() might have a misleading name – in most functional programming languages, the append operation is always defined between two lists, whereas in Python it’s an operation between a list and an element. Maybe add() would have been a better name, like Java’s add() operation of the List interface.