🚀 go-pugleaf

In article <1992Dec21.225604.18749@netcom.com>, rfg@netcom.com (Ronald F. Guilmette) writes:
|> The ANSI C standard says (in 3.5.2.2):
|> 
|> 	"Each enumerated type shall be compatible with an integer type;
|> 	the choice of type is implementation-defined."
|> 
|> I am having a polite disagreement with one implementor regarding the
|> proper interpretation of this statement.
|> 
|> I interpret it to mean that there *must* exist some integer type `T'
|> such that:
|> 
|> 		enum E { red, green, blue };
|> 
|> 		enum E object;
|> 		T object;
|> 
|> 		enum E *ptr;
|> 		T *ptr;
|> 
|> ... is a valid translation unit.

It requires more for

	enum E object ;
	int object ;

to be legal than that E and int be compatible. 3.1.2.6 
says that the "type of the identifier becomes the composite type".
It also gives the rules for determining a composite type and
says nothing about combining enums and integer types.  Granted,
this isn't a constraint section, but if there isn't a composite
type then I think the compiler is justified in treating the above
as an error.

[Maybe I'm stretching. I feel strongly that the above ought
to be an error and I went looking for a way to make it so. I'm
assuming that the original x3j11 committee didn't intend to force
us to accept code like the above, but I wasn't a member.]

    -- Jerry Schwarz





  -- Jerry Schwarz