Jafrog's dev blog

The edge cases of array slicing in Ruby

Last Thursday at our monthly ScotRUG meetup we did Ruby Koans. To my surprise it was very educational, probably because of Ruby being so inconsistent and unpredictable comprehensive and feature rich. Ceri Shaw (who reads C code much better than I do) and I paired over some unexpected array slicing behaviour and found out how it’s wired under the hood.

Array slicing

We investigated the slice method that takes two arguments: start and length.

ar = [1,2,3]
# => [2,3]

Now let’s take a look at edge cases. Namely at out of range slicing.

ar = [1,2,3]

# => [3]

# => []

# => nil

Wait, what? The last index of ar is 2, so why ar[3,1] and ar[4,1] return different results if they both are out of range?

The logic behind this, as James Bell put it, is that slices start between the elements. For ar[3,1] slice starts right before the 3rd element (or after 2nd element) and takes 1 element. Because there are no elements after the index 2 it return an empty array. And ar[4,1] starts right after the 3rd element, which is out of range and therefore returns nil.

Under the hood

Let’s take a look on the implementation. Below is source code of the slice method.

rb_ary_aref(int argc, const VALUE *argv, VALUE ary)
    VALUE arc;
    long beg, len;

    if (argc == 2) {
	beg = NUM2LONG(argv[0]);
	len = NUM2LONG(argv[1]);
	if (beg < 0) {
	    beg += RARRAY_LEN(ary);
	return rb_ary_subseq(ary, beg, len);

If 2 arguments (starting index and the length) were passed it calls rb_ary_subseq:

rb_ary_subseq(VALUE ary, long beg, long len)
    VALUE klass;
    long alen = RARRAY_LEN(ary);

    if (beg > alen) return Qnil;
    if (beg < 0 || len < 0) return Qnil;

    if (alen < len || alen < beg + len) {
	len = alen - beg;
    klass = rb_obj_class(ary);
    if (len == 0) return ary_new(klass, 0);

    return ary_make_partial(ary, klass, beg, len);

Let’s break down what is going on there:

  • If starting index is larger than array length return nil. This is the condition that executes for ar[4,1] as the length of ar = [1,2,3] is 3 and 4 > 3.

  • If either starting index or length are less than 0 return nil. The caller function rb_ary_aref takes care of negative starting indexes transforming them to indexes “from the end” with beg += RARRAY_LEN(ary);

  • If the length of requested slice is larger than the length of the array or if the last element of the requested slice falls out of range take only the elements from the starting index to the end of the array;

  • Finally if the length of the requested slice is 0 return an empty array. And from the previous line it’s clear that for ar[3,1] when ar = [1,2,3] len would be equal to 0.

So here it is, the logic and the implementation of Ruby array slicing.

comments powered by Disqus