Yeah, that index threw me for quite a while. I think a more accurate replacement is:

@v= sort {
    length($a) <=> length($b)  ||  $a cmp $b
} @v;
[download]

however (sort first by length and sort strings of the same length lexigraphically).

Basically, the use of index doesn't end up changing anything since $a can only be a substring of $b if it is either shorter than $b or it is exactly equal to $b.

        - tye (but my friends call me "Tye")