bubble-sort procedure in scheme

Question

I don't understand why (bubble-sort! (vector 2 1)) returns #(2 2) with my code.

(define (bubble-sort! v)
    (define (helper c orig-v n)
    (cond
        ((< n 0)
            v)
        ((> c n)
            (helper 0 v (- n 1)))
        ((>= (vector-ref orig-v c) (vector-ref orig-v (+ c 1)))
            (begin
                (vector-set! v (+ c 1) (vector-ref orig-v c))
                (vector-set! v c (vector-ref orig-v (+ c 1)))
                (helper (+ c 1) v n)))
        (else
            (helper (+ c 1) v n))))
(helper 0 v (- (vector-length v) 2)))

I've traced through my code myself but have not found the problem.

Answer 1

In your code v and orig-v are the same vector (if you pass a vector as parameter the vector is not copied: this is similar to “call-by-reference” as it is called in other languages).

So inside the function you work on the vector passed as parameter, and the two expressions:

(vector-set! v (+ c 1) (vector-ref orig-v c))
(vector-set! v c (vector-ref orig-v (+ c 1)))

are equivalent to:

(vector-set! orig-v (+ c 1) (vector-ref orig-v c))
(vector-set! orig-v c (vector-ref orig-v (+ c 1)))

so 2 is copied in the second element (which contains 1), and then 2 is copied back in the first element, and the final result is #(2,2).