Found this strange compilation behavior, checked on VS2012, VS2017 and https://www.onlinegdb.com/online_c++_compiler)
Basically for private nested classes you can call public functions outside, but not public constructors.
3 questions:
what is the reasoning behind compiler letting me call func()?
if compiler lets me call func(), why I cannot call ctor?
if I cannot call ctor, how come emplace_back is able to do it?
class Outer {
struct PrivateInner {
PrivateInner() {}
void func() {}
};
public:
PrivateInner inner;
std::vector<PrivateInner> innerVect;
};
void f1()
{
Outer c;
c.inner.func(); // COMPILING, but why?
}
void f2()
{
Outer c;
c.innerVect.push_back(Outer::PrivateInner()); // NOT COMPILING, why no access to ctor if there is access to func()?
c.innerVect.emplace_back(); // COMPILING, but why? Doesn't it call Outer::PrivateInner inside?
}
As I see I still can create a (static) function createObject():
class Outer {
struct PrivateInner {
PrivateInner() {}
static PrivateInner createObject() { return PrivateInner(); }
void func() {}
};
.....
};
and then call it.
createObject() may be non-static if calling static from instances is not pure standard thing.
c.innerVect.push_back(c.inner.createObject()); // COMPILING
to "hack" compilation
Note that the nested struct PrivateInner is declared as private, so only Outer::PrivateInner is private, you can't use this name to declare variable like Outer::PrivateInner pi; without sufficient access right, but you can write it like decltype(Outer::inner) pi;.
On the other hand, its constructor and member function are public, so they can be called.
c.inner.func(); // COMPILING, but why?
func() is public, so if you have got an instance of type Outer::PrivateInner you can call func on it.
c.innerVect.push_back(Outer::PrivateInner()); // NOT COMPILING, why no access to ctor if there is access to func()?
It has nothing to do with the constructor, you just can't use the name Outer::PrivateInner here.
c.innerVect.emplace_back(); // COMPILING, but why? Doesn't it call Outer::PrivateInner inside?
The constructor is public, then it could be used to construct the object. std::vector doesn't use the name like Outer::PrivateInner directly; it uses the name specified as the template argument.
BTW: For the same reason,
c.innerVect.push_back(c.inner.createObject()); // COMPILING
but c.innerVect.push_back(Outer::PrivateInner::createObject()); won't compile.
The member is public so of course you can call its member functions. If its name is visible, then so are its own public members.
Likewise the inner class is private, so of course you can't refer to its name from outside the class. These are the basic definitions of the access control rules.
emplace_back can call the constructor because std::vector received the type as a template parameter from someone who was allowed to refer to it. The access check occurred when the template was instantiated. At that point its name was accessible.
You can call the constructor outside of the class anywhere using decltype:
Outer c;
auto this_works_too = decltype(c.inner){};
This works because you don't have to refer to it by an inaccessible name.
The access control is applied to names. That means only the name of the struct PrivateInner is restricted. The members of the struct itself have own access control. So all members of PrivateInner are public.
11 Member access control [class.access]
1 A member of a class can be
private; that is, its name can be used only by members and friends of the class in which it is declared.
...
4 Access control is applied uniformly to all names, whether the names
are referred to from declarations or expressions. ...
As you already found out, you can use the constructor (and the struct in general) if you're not using its name:
c.innerVect.push_back({});
c.innerVect.emplace_back();
even
template <typename T>
T createPrivateInner()
{
return T();
}
...
c.innerVect.push_back(createPrivateInner<decltype(Outer::inner)>());
The member "createObject()" is private. So, of course, you cannot access it. You should add some member function in public field to implement this private member.
Related
I have a class that looks something like this:
class A
{
public:
void foo(int arg) { foo(arg, false); }
private:
void foo(int arg, bool flag) {}
};
It is built this way because I want foo's flag argument to only be false when called from outside A. I want to inherit it privately, but allow calling foo:
class B : private A
{
public:
using A::foo;
};
However, this fails because the using declaraion attempts to bring all the overloads of foo into scope, including the private one, which the the compiler rightly rejects.
This isn't hard to fix, I can either:
Change the accessibility of A::foo(int, bool) to protected or public
Inherit A publicly; only public overloads of foo will be inherited
Change the name of A::foo(int, bool) so that the using declaration does not attempt to bring it into scope
This is a small, private project, and besides, that overload is only called inside A. So fixing the problem is a non-issue here. (I'm just going to rename the private overload.)
But it doesn't feel like it should be necessary. Why does the using declaration attempt to bring the non-accessible methods into scope? Is this particular case just not covered by the standard? Is there a way to fix this besides the methods I listed?
You can also redefine the overload you want and have it forward its argument to the function in A:
class B : private A
{
public:
void foo(int arg) { A::foo(arg); }
};
The using declaration is just too blunt a tool in this case. It brings function names into the derived class scope. And when the name refers to something private, it chokes. It can't distinguish overloads. The standard requires the names introduced by a using declaration to be accessible:
[namespace.udecl]/17
In a using-declarator that does not name a constructor, all members of
the set of introduced declarations shall be accessible. In a
using-declarator that names a constructor, no access check is
performed. In particular, if a derived class uses a using-declarator
to access a member of a base class, the member name shall be
accessible. If the name is that of an overloaded member function, then
all functions named shall be accessible. The base class members
mentioned by a using-declarator shall be visible in the scope of at
least one of the direct base classes of the class where the
using-declarator is specified.
The forwarding function can also be templated. So one won't need to redefine each function they want to expose individually.
class B : private A
{
public:
template<typename ...Args>
void foo(Args ...args) { A::foo(args...); }
};
It's "catch-all" like the using declaration, except the access specifier is checked only on template instantiation, i.e. when the function is called. So the template will be ill-formed based on its scope and whether or not the member in A is accessible there.
I found the following extract from Scott Meyer's Effective C++ which is related to your predicament (with emphasis added):
Item 33: Avoid hiding inherited names.
...
This means that if you inherit from a base class with overloaded functions
and you want to redefine or override only some of them, you need
to include a using declaration for each name you’d otherwise be hiding.
If you don’t, some of the names you’d like to inherit will be hidden.
...
It’s conceivable that you sometimes won’t want to inherit all the functions
from your base classes. Under public inheritance, this should
never be the case, because, again, it violates public inheritance’s is-a
relationship between base and derived classes. (That’s why the using
declarations above are in the public part of the derived class: names
that are public in a base class should also be public in a publicly
derived class.)
Under private inheritance, however, it can
make sense. For example, suppose Derived privately inherits from
Base, and the only version of the function that Derived wants to inherit is the
one taking no parameters. A using declaration won’t do the trick here,
because a using declaration makes all inherited functions with a given
name visible in the derived class.
No, this is a case for a different technique, namely, a simple forwarding function:
class Base {
public:
virtual void mf1() = 0;
virtual void mf1(int);
... // as before
};
class Derived: private Base {
public:
virtual void mf1() // forwarding function; implicitly
{
Base::mf1(); } // inline
};
}
If I will define a private static field in class. Given that it's a private field, can't I initialize it outside the class?
class X{
static int private_static_field;
public:
static void setPrivateStaticField(int val);
};
void X::setPrivateStaticField(int val) {
private_static_field = val;
}
int X::private_static_field(0); // something like that it's ok? if yes, I must write this line? why? can I write it in main?
It's look that it's ok (according to the compiler), but if so, I don't understand the concept of private - How it's ok if it's outside the class?
In addition, given the class above, and the next code:
int main() {
X x1();
x1.setPrivateStaticField(3);
return 0;
}
What is the meaning of x1.setPrivateStaticField(3); , after all, this function is static and hence it's not related to some object.
Hence, I don't understand how it's ok to call setPrivateStaticField with object (x1) ?
(I thought that just X::setPrivateStaticField(3); will be ok and that x1.setPrivateStaticField(3); will be error)
I don't understand the concept of private - How it's ok if it's outside the class?
There is no contradiction here. Prior to C++ 17 static member variables required a definition that is placed separately from the class declaration.
Despite the fact that the text of the definition is placed outside the class, the member variable remains part of the class where it is declared, and retains its accessibility according to its declaration inside the class. This includes private accessibility.
What is the meaning of x1.setPrivateStaticField(3); , after all, this function is static and hence it's not related to some object.
Although C++ compiler lets you call static member functions on the object, it is cleaner to call them using scope resolution operator :: and the class name:
X::setPrivateStaticField(3);
Allowing or disallowing class method calls on an instance is up to the designers of the language. C++ designers decided to allow it; designers of other programming languages disallow it, or require compilers to issue a warning.
Within a class definition static data members are declared but not defined. So they even can have an incomplete type.
So this record
int X::private_static_field(0);
is a definition of the static data member declared in the class definition like
class X{
static int private_static_field;
// ...
This record
x1.setPrivateStaticField(3);
means an access to a class member. Static members are also class members. Using this access method the compiler will know to search the name setPrivateStaticField in the class definition because the name x1 defines an object of the class X.
Another way to access a static member is to use the following record
X::setPrivateStaticField
using namespace std;
#include <cstdio>
#include <iostream>
class One{
private:
virtual void func(){
cout<<"bark!"<<endl;
}
};
class Two: public One{
public:
void func(){
cout<<"two!"<<endl;
}
};
int main(){
One *o = new Two();
o->func();
}
Why is there an error on o->func()?
I don't know the mechanism behind it... In my opinion, o->func() should call the func() in the derived class, which is public, so there wouldn't be problems, but it says:
error: ‘virtual void One::func()’ is private
Accessibility check is performed based on the static type of the object. The type of o is One*. This means that if One::func() is private, then o->func() won't compile.
On the other hand, which virtual member function will be called (i.e. dynamic dispatch) happens at run-time, based on the dynamic type of the object. So if One::func() is public, o->func() will call Two::func(), because o is pointing to an object of type Two.
For your sample code and use case, making One::func() private is just meaningless. But note that there's a famous idiom called Non-Virtual Interface, which makes use of private virtual member functions of base class.
Other suggestions:
Don't forget to delete o;
Add a virtual destructor in the base class One. Otherwise delete o; will lead to undefined behavior; e.g. the destructor of Two might not be invoked.
class One {
public:
virtual ~One() {}
// ...
};
A subclass can't ease inheritance restriction,
even though func is virtual, it is still the inheritance restrictions remain.
please see this answer for compliated view of inheritance restrictions :
Difference between private, public, and protected inheritance
Please check Access specifiers and virtual functions.
From standard :
§11.5 [class.access.virt] The access rules (Clause 11) for a virtual
function are determined by its declaration and are not affected by the
rules for a function that later overrides it.
Access is checked at the call point using the type of the expression
used to denote the object for which the member function is called. The
access of the member function in the class in which it was defined is
in general not known.
If name lookup determines a viable function to be a virtual function, the access specifier of the virtual function is checked in the scope of the static type of the object expression used to name the function. At run time, the actual function to be called could be defined in the derived class with a completely different access specifier. This is because 'access specifiers' are a compile time phenomenon.
Since access specifier of function func() is checked in the scope of One *o, and it is private in class One, it produces error.
If Onedeclares func() as public, and Two declares it private, there won't be any errors. See this Private function invoked and it works. Could any of you reason it please
My code:-
#include<iostream>
using namespace std;
class a{
private:
int x;
public:
a(int data)
{
x=data;
}
friend void printPrivateMember(a);
};
void printPrivateMember(a obj)
{
cout<<obj.x; //I can access private data member by an object inside this function.
}
int main()
{
a obj1(5);
printPrivateMember(obj1);
cout<<obj1.x; //this gives error
return 0;
}
I wanted to know as to how can I access a PRIVATE data type by an object in the friend function but cannot do so in main.
When I read about access specifier . It was specified that private can be accessed by only member functions (I don't have a problem with the friend function) and not by the object of that class. I wanted to know as to what difference is there because of which I can access private member by an object in one case and cannot do so in another. The same is applicable for copy constructor.
That's exactly what friend functions do: any friend function of a class can access it's private members. Since your printPrivateMember is declared as a friend of a, it can access it's private x member. Since main is not a friend function, it can't.
Forestalling a question about declaring main as friend, this question covers it.
Because friends could do that.
$11/1 Member access control [class.access]
(emphasis mine)
1 A member of a class can be
(1.1) — private; that is, its name can be
used only by members and friends of the class in which it is
declared.
(1.2) — protected; that is, its name can be used only by
members and friends of the class in which it is declared, by classes
derived from that class, and by their friends (see 11.4).
(1.3) —
public; that is, its name can be used anywhere without access
restriction.
As you correctly observed, only member functions (including constructors and destructors) and friend functions and classes may access you're privates. That's the purpose of friends: they provide an exception (not std::exception) to the encapsulation mechanism.
Now you may think about whether this breaks encapsulation or actually stabilizes it.
if you want to access private member, you'd better use a public function like:
class a {
private:
int m;
public:
int getM() {
return m;
}
};
Your use of the phrase not by the object of that class makes me think that you are unclear on the access rules. The access rules don't apply to the objects but who can access member variables and member functions of the objects.
A member variable of a class can be accessed in a function -- which can be a member function of the class, a member function of another class, or a global function.
It can also be accessed in the global space, e.g. to initialize a global variable.
A friend declaration in a class changes the default access rules that are in place by use of private, protected, and public access specifiers.
A function declared a friend of a class can access all the members of all instances of the class.
The answer by songyuanyao cites the section of the standard that provides more details on the subject.
This function should be public, so that you can access it through main().
void print(){
/**print or return your private variable here**/
}
We make a non-member function a friend of a class when we want it to access that class's private members. This gives it the same access rights as a static member function would have. Both alternatives would give you a function that is not associated with any instance of that class.
When must we use a friend function? When must we use a static function? If both are viable options to solve a problem, how do we weigh up their suitability? Is there one that should be preferred by default?
For example, when implementing a factory that creates instances of class foo which only has a private constructor, should that factory function be a static member of foo (you would call foo::create()) or should it be a friend function (you would call create_foo())?
Section 11.5 "The C++ Programming Language" by Bjarne Stroustrup states that ordinary member functions get 3 things:
access to internals of class
are in the scope of the class
must be invoked on an instance
friends get only 1.
static functions get 1 and 2.
The question seems to address the situation where the programmer needs to introduce a function that does not work on any instance of a class (hence the possibility of choosing a static member function). Therefore, I will limit this answer to the following design situation, where the choice is between a static function f() and a friend free function f():
struct A
{
static void f(); // Better this...
private:
friend void f(); // ...or this?
static int x;
};
int A::x = 0;
void A::f() // Defines static function
{
cout << x;
}
void f() // Defines friend free function
{
cout << A::x;
}
int main()
{
A::f(); // Invokes static function
f(); // Invokes friend free function
}
Without knowing anything in advance about the semantics of f() and A (I'll come back to this later), this limited scenario has an easy answer: the static function is preferable. I see two reasons for this.
GENERIC ALGORITHMS:
The main reason is that a template such as the following can be written:
template<typename T> void g() { T::f(); }
If we had two or more classes that have a static function f() on their interface, this would allow us writing one single function that invokes f() generically on any such class.
There is no way to write an equivalent generic function if we make f() a free, non-member function. Although it is true that we could put f() into a namespace, so that the N::f() syntax could be used to mimic the A::f() syntax, it would still be impossible to write a template function such as g<>() above, because namespace names are not valid template arguments.
REDUNDANT DECLARATIONS:
The second reason is that if we were to put the free function f() in a namespace, we would not be allowed to inline its definition directly in the class definition without introducing any other declaration for f():
struct A
{
static void f() { cout << x; } // OK
private:
friend void N::f() { cout << x; } // ERROR
static int x;
};
In order to fix the above, we would to preceed the definition of class A with the following declaration:
namespace N
{
void f(); // Declaration of f() inside namespace N
}
struct A
{
...
private:
friend void N::f() { cout << x; } // OK
...
};
This, however, defeats our intention of having f() declared and defined in just one place.
Moreover, if we wanted to declare and define f() separately while keeping f() in a namespace, we would still have to introduce a declaration for f() before the class definition for A: failing to do so would cause the compiler to complain about the fact that f() had to be declared inside namespace N before the name N::f could be used legally.
Thus, we would now have f() mentioned in three separate places rather than just two (declaration and definition):
The declaration inside namespace N before A's definition;
The friend declaration inside A's definition;
The definition of f() inside namespace N.
The reason why the declaration and definition of f() inside N cannot be joined (in general) is that f() is supposed to access the internals of A and, therefore, A's definition must be seen when f() is defined. Yet, as previously said, f()'s declaration inside N must be seen before the corresponding friend declaration inside of A is made. This effectively forces us to split the declaration and the definition of f().
SEMANTIC CONSIDERATIONS:
While the above two points are universally valid, there are reasons why one might prefer declaring f() as static over making it a friend of A or vice versa which are driven by the universe of discourse.
To clarify, it is important to stress the fact that a member function of a class, whether it is static or non-static, is logically part of that class. It contributes to its definition and thus provides a conceptual characterization of it.
On the other hand, a friend function, in spite of being granted access to the internal members of the class it is friend of, is still an algorithm which is logically external to the definition of the class.
A function can be friend of more than one class, but it can be member of just one.
Thus, in a particular application domain, the designer may want to keep into consideration the semantics of both the function and the class when deciding whether to make the former a friend or a member of the latter (this applies not only to static functions, but to non-static functions as well, where other language constraints may intervene).
Does the function logically contribute to characterize a class and/or its behavior, or is it rather an external algorithm? This question can't be answered without knowledge of the particular application domain.
TASTE:
I believe that any argument other the ones just given stems purely from a matter of taste: both the free friend and the static member approach, in fact, allow to clearly state what the interface of a class is into one single spot (the class's definition), so design-wise they are equivalent (modulo the above observations, of course).
The remaining differences are stylistic: whether we want to write the static keyword or the friend keyword when declaring a function, and whether we want to write the A:: class scope qualifier when defining the class rather than the N:: namespace scope qualifier. Thus, I will not comment further on this.
The difference is clearly expressing the intent of the relationship between the class and the function.
You use friend when you want to intentionally indicate a strong coupling and special relationship between two unrelated classes or between a class and a function.
You use static member function when the function is logically a part of the class to which it is a member.
Friend functions (and classes) can access the private and protected members of your class.
There's rarely a good case for using a friend function or class. Avoid them in general.
Static functions may only access static data (that is, class-scoped data). They may be called without creating an instance of your class. Static functions are great for circumstances you want all of the instances of your class to behave the same way. You can use them:
as callback functions
to manipulate class-scoped members
to retrieve constant data that you don't want to enumerate in your header file
Static functions are used when you want a function that is the same for every instance of a class. Such functions do not have access to "this" pointer and thus cannot access any non static fields. They are used often when you want a function that can be used without instantiating the class.
Friend functions are functions which are not in the class and you want to give them access to private members of your class.
And this(static vs. friend) is not a matter of using one vs the other since they are not opposites.
The standard requires that operator = () [] and -> must be members, and class-specific
operators new, new[], delete and delete[] must be static members. If the situation
arises where we don't need the object of the class to invoke a function, then make
the function static. For all other functions:
if a function requires the operators = () [] and -> for stream I/O,
or if it needs type conversions on its leftmost argument,
or if it can be implemented using the class' public interface alone,
make it nonmember ( and friend if needed in the first two cases)
if it needs to behave virtually,
add a virtual member function to provide the virtual behaviour
and implement in terms of that
else
make it a member.
Static function can only access members of one class. Friend function has access to several classes, as explained by the following code:
class B;
class A { int a; friend void f(A &a, B &b); };
class B { int b; friend void f(A &a, B &b); };
void f(A &a, B &b) { std::cout << a.a << b.b; }
f() can access data of both A and B class.
One reason to prefer a friend over static member is when the function needs to be written in assembly (or some other language).
For instance, we can always have an extern "C" friend function declared in our .cpp file
class Thread;
extern "C" int ContextSwitch(Thread & a, Thread & b);
class Thread
{
public:
friend int ContextSwitch(Thread & a, Thread & b);
static int StContextSwitch(Thread & a, Thread & b);
};
And later defined in assembly:
.global ContextSwitch
ContextSwitch: // ...
retq
Technically speaking, we could use a static member function to do this, but defining it in assembly won't be easy due to name mangling (http://en.wikipedia.org/wiki/Name_mangling)
Another situation is when you need to overload operators. Overloading operators can be done only through friends or non-static members. If the first argument of the operator is not an instance of the same class, then non-static member would also not work; friend would be the only option:
class Matrix
{
friend Matrix operator * (double scaleFactor, Matrix & m);
// We can't use static member or non-static member to do this
};
A static function is a function that does not have access to this.
A friend function is a function that can access private members of the class.
You would use a static function if the function has no need to read or modify the state of a specific instance of the class (meaning you don't need to modify the object in memory), or if you need to use a function pointer to a member function of a class. In this second instance, if you need to modify the state of the resident object, you would need to pass this in and use the local copy. In the first instance, such a situation may happen where the logic to perform a certain task is not reliant on an object's state, yet your logical grouping and encapsulation would have it be a member of a specific class.
You use a friend function or class when you have created code that is not a member of your class and should not be a member of your class, yet has a legitimate purpose for circumventing the private/protected encapsulation mechanisms. One purpose of this may be that you have two classes that have need of some common data yet to code the logic twice would be bad. Really, I have only used this functionality in maybe 1% of the classes I've ever coded. It is rarely needed.
A friend function can not be inherited while a static function can be. So when an aim can be achieved with both static function and friend function, think that whether you want to inherit it or not.
Static function can be used in many different ways.
For example as simple factory function:
class Abstract {
private:
// no explicit construction allowed
Abstract();
~Abstract();
public:
static Abstract* Construct() { return new Abstract; }
static void Destroy(Abstract* a) { delete a; }
};
...
A* a_instance = A::Conctruct();
...
A::Destroy(a_instance);
This is very simplified example but I hope it explains what I meant.
Or as thread function working with Your class:
class A {
public:
static void worker(void* p) {
A* a = dynamic_cast<A*>(p);
do something wit a;
}
}
A a_instance;
pthread_start(&thread_id, &A::worker, &a_instance);
....
Friend is completely different story and they usage is exactly as described by thebretness
Friend functions can access the private and protected members of other classes.
Means they can be used to access all the data weather it is private or public.
So friend functions are used to access that data which static methods can not.
Those methods are made static which are called so many times that declaring a different location inside every object, for them becomes too costly(In terms of memory).
This can be made clear with the help of example:
Let the class's name is fact and its data member is n(which represents integer whose factorial is concern)
then in this case declaring find_factorial() as static would be wise decision!!
They are used as callback functions
to manipulate class-scoped members
to retrieve constant data that you don't want to enumerate in your header file
Now we are clear with following questions..
When a friend function is used? When a static function is used?
Now If both are viable options to solve a problem,
We can weight up their suitability in terms of accessibility(accessibility of Private data) and memory efficiency.
By default no one can be preferred as there are many situation when we need better memory management and sometimes we are are concerned with the scope of data.
For example:
foo::create() will be preferred over create_foo() when we have to call create() method after every small instance of time and we are not interested on scope of data(Private data)
And if we are interested to get the private information of more than one class(s) then create_foo() will be preferred over foo::create().
I hope this would help you!!
Here is what I think it is:
Friend function- when you need access to a different class member, but the classes are not related. Static function- when you no not need access to the 'this' pointer. But, I have a feeling there is more to it....
Static data members always share the memory.
only static function can used static data members.
static member function can be called with class name.
They must be defined outside of the class when we create a object of static member or member function in the class. It will automatically initialize the value.
It always used keyword static.
Static members can share by all the objects.
Type and scope of data members and member function is outside of the class.
A static member variable must be defined outside of the class.