Thread: On the sign of char

I quite didn't catch yet how char is represented regarding its sign. What I got so far is that char, much like any other integral, can be sign or unsigned. But unlike other integrals the decision of which is the default representation is left to the compiler. So far so good.

However, I read repeatedly that there's a third type of char; a plain char. And that this type cannot be represented. However I wonder why it is mentioned and what does it actually mean?

signed char a;
This is signed

unsigned char b;
This is unsigned

char c;
This is implementation specific as to whether it is signed or unsigned.
For representing the printable ASCII characters, it doesn't matter.

For anything else, it is best to be specific.

So the 'plain char' term is only used to identify an implementation specific char? Sounds strange considering that the same treatment is not given to other integrals... I suppose because they are not implementation defined?

This is the quote from C++ Primer:

Unlike the other integral types, there are three distinct types for char: plain char, signed char, and unsigned char. Although there are three distinct types, there are only two ways a char can be represented. The char type is respresented using either the signed char or unsigned char version. Which representation is used for char varies by compiler.

char is always equivalent to "unsigned char" or "signed char".
The implementation specific behaviour is which one you get.

gcc for example has compiler switchs
-fsigned-char
-funsigned-char
to specifically request which version you end up with.

Ok. I think I got it through my thick skull.

A plain char is regarded as a separate type from unsigned int and signed int, only in the context of it being implementation defined. The compiler understands three types; unsigned char, signed char and plain char, being that a plain char will be one of unsigned or signed.