Why is the std::bitset<8> variable unable to handl

Question

Why is this program showing the following output ?

#include <bitset>
...

{
    std::bitset<8> b1(01100100); std::cout<<b1<<std::endl;
    std::bitset<8> b2(11111111); std::cout<<b2<<std::endl; //see, this variable
                                                           //has been assigned
                                                           //the value 11111111
                                                           //whereas, during
                                                           //execution, it takes
                                                           //the value 11000111
    std::cout << "b1 & b2: " << (b1 & b2) << '\n';
    std::cout << "b1 | b2: " << (b1 | b2) << '\n';
    std::cout << "b1 ^ b2: " << (b1 ^ b2) << '\n';
}

This is the OUTPUT:

01000000
11000111
b1 & b2: 01000000
b1 | b2: 11000111
b1 ^ b2: 10000111

First, I thought there is something wrong with the header file (I was using MinGW) so I checked using MSVCC. But it too showed the same thing. Please help.

Answer 1

As per NPE's answer, you are constructing the bitset with an unsigned long, and not with bits as you were expecting. An alternative way to construct it, which enables you to specify the bits, is by using the string constructor as follows:

#include <bitset>
#include <cstdio>
#include <iostream>

int main()
{
    std::bitset<8> b1(std::string("01100100")); std::cout<<b1<<std::endl;
    std::bitset<8> b2(std::string("11111111")); std::cout<<b2<<std::endl;
    std::cout << "b1 & b2: " << (b1 & b2) << '\n';
    std::cout << "b1 | b2: " << (b1 | b2) << '\n';
    std::cout << "b1 ^ b2: " << (b1 ^ b2) << '\n';
getchar();
return 0;
}

Click here to view the output.

Answer 2

Despite the appearance, the 11111111 is decimal. The binary representation of 11111111₁₀ is 101010011000101011000111₂. Upon construction, std::bitset<8> takes the eight least significant bits of that: 11000111₂.

The first case is similar except the 01100100 is octal (due to the leading zero). The same number expressed in binary is 1001000000001000000₂.

One way to represent a bitset with a value of 11111111₂ is std::bitset<8> b1(0xff).

Alternatively, you can construct a bitset from a binary string:

std::bitset<8> b1(std::string("01100100"));
std::bitset<8> b2(std::string("11111111"));