What's the diference between use:
static Foo foo;
// ...
foo.func();
And:
Foo& GetFoo(void)
{
static Foo foo;
return foo;
}
// ...
GetFoo().func();
Which is better?
-
With the global variable you risk naming conflicts.VoidStar– VoidStar2012-08-30 00:44:14 +00:00Commented Aug 30, 2012 at 0:44
-
See stackoverflow.com/questions/12186857/….Pete Becker– Pete Becker2012-08-30 00:54:37 +00:00Commented Aug 30, 2012 at 0:54
-
@VoidStar: With the global function you risk naming conflicts.Mike Seymour– Mike Seymour2012-08-30 06:59:44 +00:00Commented Aug 30, 2012 at 6:59
Add a comment
|
2 Answers
The principal difference is when construction occurs. In the first case, it occurs sometime before main() begins. In the second case, it occurs during the first call to GetFoo().
It is possible, in the first case, for code to (illegally) use foo prior to its initialization. That is not possible in the second case.
1 Comment
David Rodríguez - dribeas
While the intent is correct, with enough effort you can use the function static (in C++03) before initialization through recursive calls to the function.
A GetFoo is generally used when you don't want copies of your class/object.
For example:
class Foo
{
private:
Foo(){};
~Foo();
public:
static Foo* GetFoo(void)
{
static Foo foo;
return &foo;
}
int singleobject;
};
You can externally access singleobject via Foo::GetFoo()->sinlgeobject. The private constructors and destructors avoid your class getting copies created.
For the use of static Foo foo, you must have public constructors declared which means you are always accessing your class by it, but your class will also be able to get copies.
