I'm trying to understand this behaviour but it seems I don't. Please see this code:
#include <iostream>
using namespace std;
class Base
{
public:
void operator=(const Base& rf)
{
cout << "base operator=" << endl;
this->y = rf.y;
}
int y;
Base() : y(100) { }
};
class Derived : public Base
{
public:
int x;
Derived() : x(100) { }
};
int main()
{
Derived test;
Derived test2;
test2.x = 0;
test2.y = 0;
test.operator=(test2); // operator auto-generated for derived class but...
cout << test.x << endl << test.y << endl;
cin.ignore();
return 0;
}
PROGRAM OUTPUT:
> base operator=
> 0
> 0
Now where I'm confused is:
The rule says that a derived class never inherits the assigment operator, instead it creates its own operator= however in this example base's operator= gets invoked on the derived class.
Second I was able to explicitly invoke an assigment operator on a derived class, which isn't in turn explicitly defined in the derived class.
Now if I understand it correctly, this means that any user defined base's operator always gets invoked on the derived class?
The generated ones automatically call the base-class assignment operator.
// generated version looks basically like this
Derived& operator=(Derived const& other){
Base::operator=(static_cast<Base const&>(other));
x = other.x;
return *this;
}
The cast is there to avoid an accidential call to a templated Base::operator= like this one:
template<class Other>
Base& operator=(Other const& other); // accepts everything
Or a strange one like this:
// forward-declare 'Derived' outside of 'Base'
Base& operator=(Derived const& other); // accepts derived class (for whatever reason)
Second I was able to explicitly invoke an assigment operator on a derived class, which isn't in turn explicitly defined in the derived class.
The compiler automatically declares an assignment operator if you do not and your class allows it (i.e., no reference members and some other arcane rules) and additionally defines it if you actually use it somewhere.
The compiler-generated assignment operator calls the assignment operator of each subobject. That includes base classes and non-static member variables.
The standard says (section 12.8 [class.copy]):
If the class definition does not explicitly declare a copy assignment operator, one is declared implicitly. If the class definition declares a move constructor or move assignment operator, the implicitly declared copy assignment operator is defined as deleted; otherwise, it is defined as defaulted (8.4). The latter case is deprecated if the class has a user-declared copy constructor or a user-declared destructor. The implicitly-declared copy assignment operator for a class X will have the form
X& X::operator=(const X&)
if
each direct base class B of X has a copy assignment operator whose parameter is of type const B&,
const volatile B& or B, and
for all the non-static data members of X that are of a class type M (or array thereof), each such class type has a copy assignment operator whose parameter is of type const M&, const volatile M& or M.
Otherwise, the implicitly-declared copy assignment operator will have the form
X& X::operator=(X&)
and
The implicitly-defined copy/move assignment operator for a non-union class X performs memberwise copy/move assignment of its subobjects. The direct base classes of X are assigned first, in the order of their declaration in the base-specifier-list, and then the immediate non-static data members of X are assigned, in the order in which they were declared in the class definition. Let x be either the parameter of the function
or, for the move operator, an xvalue referring to the parameter. Each subobject is assigned in the manner appropriate to its type:
if the subobject is of class type, as if by a call to operator= with the subobject as the object expression and the corresponding subobject of x as a single function argument (as if by explicit qualification; that is, ignoring any possible virtual overriding functions in more derived classes);
if the subobject is an array, each element is assigned, in the manner appropriate to the element type;
if the subobject is of scalar type, the built-in assignment operator is used.
It is unspecified whether subobjects representing virtual base classes are assigned more than once by the implicitly-defined copy assignment operator.
That's because the implicitly defined operator = calls the base classes operator =. See the FAQ lite:
I'm creating a derived class; should my assignment operator call my base class's assignment operator?
Yes (if you need to define an assignment operator in the first place).
If you define your own assignment operator, the compiler will not automatically call your base class's assignment operator for you. Unless your base class's assignment operator itself is broken, you should call it explicitly from your derived class's assignment operator (again, assuming you create one in the first place).
However if you do not create your own assignment operator, the one that the compiler creates for you will automatically call your base class's assignment operator.
4 things never get inherited
Constructor
Copy-constructor
Assignment operator
Destructor
Even you have not written the assignment operator your code will be Woking fine.
Whenever you use assignment operator compiler will make sure that assignment operator for each member object and base class get called.
Related
In C++ primer 5 ed. it is said:
"The fact that a base class needs a virtual destructor has an important indirect impact on the definition of base and derived classes: If a class defines a destructor—even if it uses = default to use the synthesized version—the compiler will not synthesize a move operator for that class (§13.6.2, p. 537)."
So for more understanding I've tried this:
class A
{
public:
A(){cout << "A::A()\n";}
virtual ~A(){cout << "A::~A()\n";}
A& operator = (const A&) {cout << "A::=\n"; return *this;};
A& operator = (A&&) = default;
};
int main()
{
A a;
A a2;
a2 = std::move(a); // why copy-assignment operator is not called here. Normally a will be "moved" through copy-ctor (moving through working copy-ctor)
std::cout << std::endl;
}
So according to the text above in the book class A has a virtual destructor thus the compiler won't synthesize a move-ctor/assignment operations But in main I've tried to move an object of type class A so normally move here uses copy-assignment instead of move-assignment but copy-assignment is not called here?!
So when a vairtual dtor causes a default operation be not synthesized? Thank you!
Your textbook is right, the compiler won't automatically synthesize a move operator. In the code sample you force the compiler to generate one by setting it to default, this can only fail if the move operator is implicitly deleted. Which is something else than it being implicitly declared.
Criteria for implicit deletion:
A has a non-static data member that is const
A has a non-static data member of a reference type
A has a non-static data member that cannot be move-assigned (has deleted, inaccessible, or ambiguous move assignment operator)
A has direct or virtual base class that cannot be move-assigned (has deleted, inaccessible, or ambiguous move assignment operator)
A has a non-static data member or a direct or virtual base without a move assignment operator that is not trivially copyable
A has a direct or indirect virtual base class
Criteria for implicit declaration:
there are no user-declared copy constructors;
there are no user-declared move constructors;
there are no user-declared copy assignment operators;
there are no user-declared destructors;
the implicitly-declared move assignment operator would not be defined as deleted
source
The cppreference says:
Because the copy assignment operator is always declared for any class, the base class assignment operator is always hidden. If a using-declaration is used to bring in the assignment operator from the base class, and its argument type could be the same as the argument type of the implicit assignment operator of the derived class, the using-declaration is also hidden by the implicit declaration.
From my understanding, the following code should not compile. Because
B::operator=(const B&) is implicitly declared.
both A::operator=(const A&) and using-declaration are hidden.
#include <iostream>
using namespace std;
class A {
public:
A& operator=(const A& A) {
cout << "A::opreator=" << endl;
}
};
class B : A {
public:
using A::operator=;
};
int main() {
A a1;
B b1;
b1 = a1;
}
However, it compiles successfully and prints "A::operator=", why?
From C++11 Standards#12.8 [emphasis added]:
24 Because a copy/move assignment operator is implicitly declared for a class if not declared by the user, a base class copy/move assignment operator is always hidden by the corresponding assignment operator of a derived class (13.5.3). A using-declaration(7.3.3) that brings in from a base class an assignment operator with a parameter type that could be that of a copy/move assignment operator for the derived class is not considered an explicit declaration of such an operator and does not suppress the implicit declaration of the derived class operator; the operator introduced by the using-declaration is hidden by the implicitly-declared operator in the derived class.
The implicit declaration of class B assignment operation will be like this:
B& B::operator=(const B&)
The parameter type of using-declaration assignment operator in class B is different from implicitly declared assignment operator. Hence, it suppress the implicit declaration of the derived class B operator.
For understanding on 1 & 2 w.r.t. to the code you have posted:
No, the implicit declaration of assignment operator is suppressed in class B.
No, they will not be hidden.
You can't hide the copy assignment operator of B because both operators you've mentioned take different parameters.
I think the reference you mentioned should be broken into 2 parts that suite your 2 question:
Yes, the copy assignments for base class (A) and derived class(B) are implicitly declared by default. You just overriden the implicit declaration on class A by output a message to stream.
the second part of the reference means that if the instance of class B is assigned to an instance of class A and there is a using-declaration to use class A assignment, then the assignment of class A will be used. Which means, if an instance of class B is assigned to an other instance of class B, the default implicit assignment will be used. Therefore, nothing is hidden due to the difference in argument type and the code will call the using-declaration (a.k.a the base class assignment) => the message is output to stream.
I don't see any conflict in this code with standard.
b1 = a1;
This assignment is done because you used using-declaration of base class.
and also " implicit assignment operator of the derived class" is provided by compiler because if you want to assign two objects of derived class it's possible.
I have a bunch of questions related to this, I couldn't find exact answers.
Class A is the main class, and B is a subclass
If A::operator=(const A & other) was defined, does the default implementation of B::operator= copies members of B then calls A::operator= ?
If A copy constructor was defined, does the default implementation of the copy constructor for B copy-constructs members of B, then calls A copy constructor?
Do I need to define above functions virtual in A to get this behavior? (I suppose yes for operator= and no for the copy constructor, as virtual constructors is a nonsense?)
Can I forbid overloading the assignment operator or the copy constructor for subclasses of A, to force the use of default implementations?
The idea behind this, being to offer a plugin API for my users. I need the API to be C++ as scripts are too slow (I'll try JIT compilation one day), but it should be very simple.
If the default copy-assignment operator of class B is not deleted, [class.copy]/28
The implicitly-defined copy/move assignment operator for a non-union class X performs memberwise copy-/move assignment of its subobjects. The direct base classes of X are assigned first, in the order of their declaration in the base-specifier-list [i.e. in the order in which they're listed after class X : /*here*/ {/*...*/};], and then the immediate non-static data members of X are assigned, in the order in which they were declared in the class definition.
Similarly, [class.copy]/15
The implicitly-defined copy/move constructor for a non-union class X performs a memberwise copy/move of its bases and members.
The order is: first the base class(es) (base class subobjects), then the direct non-static data members, similarly to 1).
For the behaviour described in 1) and 2), no. Virtual assignment operators are hardly ever useful. Constructors may not be virtual at all (doesn't make sense).
In order for a virtual function in a derived class B to override a virtual function in a base class A, it must have the same parameter types. That is, you could have a virtual A& operator=(A const&); in the base class A, but the override in class B had to look like virtual B& operator=(A const&);, which is not a copy-assignment operator for B, because of the parameter type.
Not without "hacks". And you're not actually overloading it, but hiding all base class assignment operators. Otherwise, this would be legal:
class A {};
class B { int i; };
A a;
B b = a; // using A::operator=(A const&)
My problem is very easily explained by this example :
http://pastebin.com/VDBE3miY
class Vector3
{
float _x;
float _y;
float _z;
public :
/// constructors and stuff
};
class Point : public Vector3
{
// some BS
Point(float _x):Vector3(float _x)
{}
};
main()
{
Point aPoint(3);
Point anotherPoint(4);
// WHY DOES THIS WORK and copy _x,_y & _z properly
aPoint = anotherPoint;
}
Basically, I am at a loss to understand why the = for the derived class can copy _x, _y and _z, even though it shouldn't have access to them since they are private.
aPoint = anotherPoint;
This line triggers a call of Point::operator= (the assignment operator), which exists because the compiler generates a default implementation. This default implementation performs assignment operations for all members of the class, as well as calling Vector3::operator=, the assignment operator of the base class. This, in turn, is a member function of Vector3 and therefore has access to all private members, which it makes copies of.
(EDIT) A quote from the C++11 Standard to back this answer:
(§12.8/28) The implicitly-defined copy/move assignment operator for a non-union class X performs memberwise copy-/move assignment of its subobjects. The direct base classes of X are assigned first, in the order of their declaration in the base-specifier-list, and then the immediate non-static data members of X are assigned, in the order in which they were declared in the class definition. Let x be either the parameter of the function or, for the move operator, an xvalue referring to the parameter. Each subobject is assigned in the manner appropriate to its type:
— if the subobject is of class type, as if by a call to operator= with the subobject as the object expression
and the corresponding subobject of x as a single function argument (as if by explicit qualification; that is, ignoring any possible virtual overriding functions in more derived classes);
— if the subobject is an array, each element is assigned, in the manner appropriate to the element type;
— if the subobject is of scalar type, the built-in assignment operator is used.
Some of the other (now partly deleted) answers mentioned the idea of a bitwise copy performed by the assignment operation. There is some truth in this: If your class or struct defines a POD (plain old data) type, it is for all practical purposes identical to a C struct. In that case, it can be copied by performing memcpy, therefore you can think of the assignment operation as being basically equivalent to a bitwise copy. But the reason why this is a valid way of thinking about it is §12.8/28 above, and that applies to non-POD types as well.
Also note that from your code it is not necessarily clear that your data type is POD. You mentioned constructors and stuff in the base class: If this involves non-trivial copy constructors, assignment operators or possibly virtual functions, then your data type is no longer POD.
About the question in the comment: In order to call the base-class assignment operator from within the derived-class implementation, just call it:
struct Base
{
};
struct Derived : Base
{
Derived &operator=(const Derived &other)
{ Base::operator=(other); return *this; }
};
Because the default copy operator of Vector3 is invoked (shallow copy).
Because the compiler generate assignment operator(s)
Point& operator=(Point const& rhs)
{
Vector3::operator=(rhs);
return *this;
}
Vector3& operator=(Vector3 const& rhs)
{
// A class is a friend of irself.
// So an object of Type A can read any other object of type A
_x = rhs._x;
_y = rhs._y;
_z = rhs._z;
return *this;
}
I have a small confusion regarding the situations where the implementation (compiler) will not supply the copy constructor and the copy assignment operator.
When we declare the copy ctor and/or copy assignment operator in our class.
Some says when we derive from a class which has a private copy ctor and/or copy assignment operator.
I am a little confused about the second situation, is the second situation is precisely.
a) The implementation will not declare them for you, so you will get a compile time error.
OR
b) The implementation will declare and define them, but when the compiler defined implementation tries to find the base class' method, we will get a compile time error.
I had an interview yesterday, I said its (b) that is happening but the interviewer disagrees, he says its (a).
I tried to compile the following code in both Microsoft C/C++ 14.00 and gcc 4.4.5
struct A
{
private:
A& operator = ( const A& );
};
struct B : A
{
};
int main()
{
B b1;
B b2;
b1 = b2;
return 0;
}
Microsoft compiler output
ctor01.cpp(9) : error C2248: 'A::operator =' : cannot access private member declared in class 'A'
ctor01.cpp(4) : see declaration of 'A::operator ='
ctor01.cpp(2) : see declaration of 'A'
This diagnostic occurred in the compiler generated function 'B &B::operator =(const B &)'
gcc compiler output
Ctor01.cpp: In member function ‘B& B::operator=(const B&)’:
Ctor01.cpp:4: error: ‘A& A::operator=(const A&)’ is private
Ctor01.cpp:8: error: within this context
Ctor01.cpp: In function ‘int main()’:
Ctor01.cpp:15: note: synthesized method ‘B& B::operator=(const B&)’ first required here
So I think, the implementation will declare and define it, but when the compiler defined implementation tries to find the base class method, we will get a compile time error. Correct me if I am wrong.
Regarding the copy constructor, this is what the standard says (12.8/7) :
A program is illformed if the class
for which a copy constructor is
implicitly defined has:
a nonstatic data member of class type (or array thereof) with an
inaccessible or ambiguous copy
constructor, or
a base class with an inaccessible or ambiguous copy
constructor.
Regarding the copy assignment operator (12.8/12) :
A program is illformed if the class
for which a copy assignment operator
is implicitly defined has:
a nonstatic data member of const type, or
a nonstatic data member of reference type, or
a nonstatic data member of class type (or array thereof) with an
inaccessible copy assignment operator,
or
a base class with an inaccessible copy assignment operator.
How the compiler reports the error, or how it actually falls into it, is pretty much irrelevant from my point of view.
However, I do believe that answer (b) is probably more correct : the base class copy assignment is declared, and it's inaccessible. The derived class has an implicitly declared copy assignment which the compiler will try to define if used, thus making the program ill-formed.
A class will have a copy constructor and a copy assignment operator implicitly declared if there is no user declared version of either. This always happens.
Simplistically, the implementation will implicitly define these only if they are actually used. If, when the implementation tries to define them, the implicit definition would be ill-formed (e.g. for copy-assignment the class contains a reference member or const member or for the copy constructor a base or member has private copy constructor) then the program is ill-formed.
A program can still be valid if it contains classes which have implicitly declared copy constructors and copy assignment operators which cannot be implicitly defined so long as it does not cause these to actually be defined by using them or causing them to be used.
Your case (b) is more accurate.
C++03 Standard 12.8p10
If the class definition does not explicitly declare a copy assignment operator, one is declared implicitly.
And 12.8p12
An implicitly-declared copy assignment operator is implicitly defined when an object of its class type is assigned a value of its class type or a value of a class type derived from its class type. A program is ill-formed if the class for which a copy assignment operator is implicitly defined has:
a nonstatic data member of const type, or
a nonstatic data member of reference type, or
a nonstatic data member of class type (or array thereof) with inaccessible copy assignment operator, or
a base class with an inaccessible copy assignment operator.
The corresponding requirements for implicitly defined copy constructors, default constructors, and destructors have similar wordings.
Specifying that the methods exist even though their definitions will be illegal clarifies some things about overload resolution. For example,
class A {
private:
A& operator=(const A&);
};
class B : public A {
public:
operator int() const;
B& operator=(int);
};
void f(B& b1, const B& b2)
{ b1 = b2; }
is illegal because the implicitly-declared B::operator=(const B&) is the better overload but the implicit definition is ill-formed. Without that declaration, you might think the compiler should implicitly convert b2 to int and then assign that to b1.
I think the distinction between the two depends on the details of your specific implementation (and doesn't make much difference). For what it's worth, Comeau gives this:
"ComeauTest.c", line 7: error: "A &A::operator=(const A &)" (declared at line 4) is
inaccessible
struct B : A
^
detected during implicit generation of "B &B::operator=(const B &)"
at line 16
1 error detected in the compilation of "ComeauTest.c".
So on that compiler, it detects the error "during" the implicit generation of B's assignment operator. In other words, it tries to generate it, and finds that it can't. Whether it detects it as it's writing it out, or by looking at A directly, doesn't really matter.
This is what happens :
struct A
{
private:
A& operator = ( const A& );
};
struct B : A
{
B& operator = ( const B& other )
{
A::operator=( other );
return *this;
}
};
int main()
{
B b1;
B b2;
b1 = b2;
return 0;
}
The default operator= tries to call A::operator=, which is private.
The standard seems to agree with you. Quoting from the current draft:
§12.8/8:
If the class definition does not
explicitly declare a copy constructor
and there is no user-declared move
constructor, a copy constructor is
implicitly declared as defaulted
(8.4).
§12.8/12:
A defaulted copy/move constructor for
a class X is defined as deleted
(8.4.3) if X has: […]
a direct or virtual base class B that cannot be copied/moved because
overload resolution (13.3), as applied
to B’s corresponding constructor,
results in an ambiguity or a function
that is deleted or inaccessible from
the defaulted constructor […]
So the synthesized copy constructor is declared and defined, but defined as deleted.