Is it possible to "zip" an array without using 2 subarrays? Normally the result of zipping 2 arrays would look something like this
a1 = [1,2,3]
a2 = [4,5,6]
a3 = zip(a1, a2)
a3 == [1,4,2,5,3,6]
However, I don't have 2 subarrays and all I am given is an input (lets say creating arrays is not a capability at this moment), and I want to zip the array in place. e.g.
a1 = [1,2,3,4,5,6]
zip(a1)
a1 == [1,4,2,5,3,6]
My current algorithm doesn't do this properly (it results in a sequence where the first half is zipped properly, but the second half is zipped in reverse)
array = [1,2,3,4,5,6,7,8]
a = array[0]
b = array[1]
i = 0, j = array.length / 2
while (i + 2 < j)
array[i] = a
array[i + 1] = b
swap(array, i + 1, i + 2)
swap(array, i + 1, j)
i += 2, j += 1
a = b
b = array[i + 1]
array == [1,5,2,6,4,8,3,7]
Complete answers aren't necessary - a guide into the general direction is preferred.
The straightforward way is by opening room for each element on swap, by shifting to the right:
void swap(int *array, int left, int right) {
int tmp = array[right];
while (right > left)
array[right] = array[--right];
array[right] = tmp;
}
Now you just have to adjust your for loop indexes:
void zip(int *array, int sz) {
int i, j;
for (i = 1, j = sz/2; j < sz; i += 2, ++j)
swap(array, i, j);
}