smallest unsigned int that can contain the positive values of a given signed integral?

Question

Can the smallest unsigned int that can contain the positive values of a given signed integral type be determined at compile time?

So the smallest unsigned integral type that is large enough to hold the max value of a given signed integral type.

Answer 1

A good starting point is that 2 is the base-number and we need to take into account the 2-based logarithm. General rules:

1. If an unsigned integer has a sizeof n bytes, then the largest positive number is (2^(8 * n)) - 1

2. If a signed integer has a sizeof n bytes, then the largest positive number is (2^(8 * n - 1)) - 1

As a result, you need to determine the sizeof a variable of the given type and see if it has a sign.