Thread: basic_string<char> and basic_string<unsigned char>

I'm using vc++ 2003, and on my compiler, if a character in a string sequence is larger than 0x80, it shows some pretty weird behavior when used with cout.

For example

Code:

char ct[] = {0xa1, 0xc1, 0xbf, 0xff, 0x00};
string s(ct);
string::iterator its;
cout.setf(ios_base::hex, ios_base::basefield);
cout.setf(ios_base::uppercase);
for(its = s.begin(); its != s.end(); its++)
	cout << setw(3) << (int) *its;

This will print out

Code:

FFFFFFA1FFFFFFC1FFFFFFBFFFFFFFFF

However, if I define the string instead as basic_string<unsigned char>:

Code:

unsigned char ct[] = {0xa1, 0xc1, 0xbf, 0xff, 0x00};
basic_string<unsigned char> s(ct);
basic_string<unsigned char>::iterator its;
cout.setf(ios_base::hex, ios_base::basefield);
cout.setf(ios_base::uppercase);
for(its = s.begin(); its != s.end(); its++)
	cout << setw(3) << (int) *its;

This prints out:

Code:

 A1 C1 BF FF

Why is cout giving me that extra garbage in the upper bits (i.e. +8) if I designate it a basic_string as a plain char sequence? If you expand signed characters, do they become signed integers?