A tweak on this question that I've run into. Consider:
class A {};
class B : private A {
static void foo();
};
void B::foo(){
B* bPtr1 = new B;
A* aPtr1 = dynamic_cast<A*>(bPtr1); // gives pointer
B* bPtr2 = dynamic_cast<B*>(aPtr1); // gives NULL
}
Since aPtr1
is, in fact, of type B*
, and since we've got full access to B
and its inheritance from A
, I'd expected both casts to work. But they don't; why? Is there another way to achieve this cast?
Note that:
- If
foo()
were not a member of B, both casts would fail.
- If
B
inherits from A
publicly, both casts would work.
5.2.7 (ISO/IEC 14882, 12/29/2003) is pretty explicit on this point:
[about the expression dynamic_cast<T>(v)
]
If T
is “pointer to cv1 B
” and v
has type “pointer to cv2
D
” such that B
is a base class of D
, the result is a
pointer to the unique B
sub-object of the D
object pointed to by v
.
[... bla bla about cv1 and cv2 ...] and B shall be an accessible unambiguous base class
of D (emphasis mine)
(recall 11.2 "A base class is said to be accessible if an invented public member of the base class is accessible.").
This explains why the first cast works. Now, for the second:
[...]
Otherwise, a run-time check is applied to see if the object pointed or
referred to by v
can be converted to the type pointed or referred to
by T
.
The run-time check logically executes as follows:
- If, in the most derived object pointed (referred) to by
v
, v
points
(refers) to a public
base class subobject of a T
object, and if only
one object of type T is derived from the sub-object pointed (referred)
to by v
, the result is a pointer (an lvalue referring) to that T
object.
- Otherwise, if
v
points (refers) to a public
base class sub-object
of the most derived object, and the type of the most derived object
has a base class, of type T
, that is unambiguous and public
, the
result is a pointer (an lvalue referring) to the T
sub-object of the
most derived object.
- Otherwise, the run-time check fails.
The value of a failed cast to pointer type is the null pointer value
of the required result type. A failed cast to reference type throws
bad_cast (18.5.2).
So it seems that the behavior you observe is due to the private
inheritance: even if the base class is accessible, it is not public, and the standard requires public, not accessible.
Annoying, isn't it ? I don't have the C++0x draft handy, perhaps someone can edit my answer with quotes from it, in case things have changed.
Is there another way to achieve this cast?
It depends on what you want to do. Basically, private inheritance is just another device for performing composition. If you really are to return a pointer to the private derived instance, then either make the inheritance public, or return a member.
Anyways, you'll happy to know that static_cast
doesn't seem to have this limitation:
5.2.9. [about static_cast<T>(v)
] [...]
An rvalue of type “pointer to cv1 B”, where B is a class type, can be converted to an rvalue of type “pointer
to cv2 D”, where D is a class derived (clause 10) from B, if a valid standard conversion from “pointer to D”
to “pointer to B” exists (4.10), cv2 is the same cv-qualification as, or greater cv-qualification than, cv1, and
B is not a virtual base class of D. The null pointer value (4.10) is converted to the null pointer value of the
destination type. If the rvalue of type “pointer to cv1 B” points to a B that is actually a sub-object of an
object of type D, the resulting pointer points to the enclosing object of type D. Otherwise, the result of the
cast is undefined.
so if you know for sure what the actual dynamic type of the pointer is, you are allowed to static_cast
inside foo
.
I'd be interested in any additional information about why this inconsistency exists.
They don't work because there's no virtual functions in A. When you do the downcast then that's trivial- the compiler probably doesn't even bother making a check. When you make an upcast then the compiler must check, but it's only defined to work when you have virtual functions. When you don't, then the compiler will fail to make the check and the result is NULL
.
The inheritance protection level and other accessibility issues are orthogonal to the problem, and they only exist at compile-time, if the program compiles then they work fine.
Note that:
If foo() were not a member of B, both casts would fail.
If B inherits
from A publicly, both casts would work.
That's just not true. foo()
has absolutely no bearing on RTTI functionality- it's not virtual and it's not even an instance member. If B inherits from A publicly- then A still has no virtual functions and it still won't work.