CareerCup

I may be misunderstanding something here, and if I am please disregard, but it seems like this solution is almost correct. Let's keep with the question in the sense that they hypothetically want to toggle bits 11-15, it appears that this solution will only toggle bits 11-14.

I think

((1 << last) - 1)

should be slightly altered to

((1 << (last + 1)) - 1)

As an example, let's take the number: 85499444
[represented in binary as 0000 0101 0001 1000 1001 1110 0011 0100]

Assuming

last = 15

and

first = 11

.

With the original solution of:

size_t mask = ((1 << last) - 1) ^ ((1 << first) - 1);

We get a mask of

0000 0000 0000 0000 0111 1000 0000 0000

As we can see this only has an effect on bits 11-14.

However, with the slight alteration of:

size_t mask = ((1 << (last + 1)) - 1) ^ ((1 << first) - 1);

We get a mask of

0000 0000 0000 0000 1111 1000 0000 0000

Which we can see has an effect on bits 11-15.


That said

0000 0101 0001 1000 1001 1110 0011 0100 ^
0000 0000 0000 0000 1111 1000 0000 0000
==================================
0000 0101 0001 1000 0110 0110 0011 0100

This effectively toggles bits 11-15 of the number 85499444.

TL;DR
Unless I misunderstood, I think the correct solution is actually:

size_t toggleBits(size_t num, size_t first, size_t last) {
	size_t mask = ((1 << (last + 1)) - 1) ^ ((1 << first) - 1);
	return num ^ mask;

}

@LinhHA05, @datta016

Nice idea.