TITLE: the scope of function and class names

(Newsgroups: comp.lang.c++.moderated, 8 Oct 97)

GORMAN: Art Gorman <gorman@roguewave.com>

>Does the function signature constitute the type of the function?

CLAMAGE: stephen.clamage@eng.Sun.COM (Steve Clamage)

The signature of a function is its name plus parameter types. The type
of a function is its return type plus its parameter types. (That is,
the function's signature includes its name but not its return type;
the function's type includes its return type but not its name.)
But these distinctions are not the source of your problem.

GORMAN:

>If I have a routine that defines a class and a function with identical
>names in the same scope

CLAMAGE:

Valid, but not a good idea. It is better to use different names for
different things when you have the option. (C++ originally did not
allow such name duplication, but the rule was relaxed for class names
because of C compatibility issues. C does not allow any other sort of
name duplication, and neither does C++.)

GORMAN:

>   The problem arises
>when you turn the "struct a" into a "template <class T> struct a" ...

CLAMAGE:

That violates a rule. A template name cannot be the same as the name
of any other entity declared in the same scope.