FIX a bug: integer type resolution

[MustBastani]

Hi,

I noticed a bug in the integer type resolver when determining the bit width of signed integers.

For example, for a range like -140..=-44, it assigns 8 bits, even though an i8 can only store values in the range -128..=127.

The problematic line is primarily in:

asnfuzzgen/asnfuzzgen/src/resolver/asn/types/base/integer.rs

Line 39 in 0d82d8e

std::cmp::max(bits_needed_min, bits_needed_max) + if min < 0 && max > 0 { 1 } else { 0 }

However, my pull request includes additional changes to address the various integer bit widths.