Why signed char can hold bigger values than 127? - c++

Question

int main()
{
char MCU = 0b00000000;
char al_av = 0b10100000;
// Before bit operation
cout << "MCU = " << int(MCU) << endl;
MCU = MCU | al_av;
// After the bit operation
cout << "MCU = " << int(MCU) << endl; // Expected 160, got -96
char temp = 160;
cout << temp; // got the a with apostrophe
return 0;
}
I expected the output of char temp to be a negative number (or a warning / error) because 160 exceeds the [-127,127] interval, but instead, the result was the one in the ASCII table (a with apostrophe)
On cpp reference:
char - type for character representation which can be most efficiently processed on the target system (has the same representation and alignment as either signed char or unsigned char, but is always a distinct type)
I don't understand what is written in italic (also I'm not sure it helps a lot for this question). Is there any implicit conversion ?

Answer 1

Why signed char can hold bigger values than 127?
It cannot.
char x = 231;
here, there is an (implicit) integer conversion: 231 is a prvalue of type int and takes value -25 before it is converted to char (which is signed on your system). You can ask your compiler to warn you about it with -Wconstant-conversion.
char - type for character representation which can be most efficiently processed on the target system (has the same representation and alignment as either signed char or unsigned char, but is always a distinct type)
I don't understand what is written in italic
This isn't related to what the type can hold, it only ensures that the three types char, signed char and unsigned char have common properties.

Answer 2

From C++14 char, if signed, must be a 2's complement type. That means that it has the range of at least -128 to +127. It's important to know that the range could be larger than this so it's incorrect to assume that a number greater than 127 cannot be stored in a char if signed. Use
std::numeric_limits<char>::max()
to get the real upper limit on your platform.
If you do assign a value larger than this to a char and char is signed then the behaviour of your code is implementation defined. Typically that means wrap-around to a negative which is practically universal behaviour for a signed char type.
Note also that ASCII is a 7 bit encoding, so it's wrong to say that any character outside the range 0 - 127 is ASCII. Note also that ASCII is not the only encoding supported by C++. There are others.
Finally, the distinct types: Even if char is signed, it is a different type from signed char. This means that the code
int main() {
char c;
signed char d;
std::swap(c, d);
}
will always result in a compile error.

Answer 3

char temp = 160;
It is actually negative. The point is cout supports non-ASCII characters, so it interprets it as non-negative. cout is probably casting it to unsigned char (or any unsigned integral type) before using it.
If you use printf and tell it to interpret it as an integer you will see that it is a negative value.
printf("%d\n", temp); // prints -96