\rSec0[class.access]{Member access control}%

\indextext{access control|(}

\indextext{protection|see{access control}}

\indextext{\idxcode{private}|see{access control, \tcode{private}}}

\indextext{\idxcode{protected}|see{access control, \tcode{protected}}}

\indextext{\idxcode{public}|see{access control, \tcode{public}}}

A member of a class can be

\begin{itemize}

\indextext{access control!\idxcode{private}}%

\tcode{private};

that is, its name can be used only by members and friends

of the class in which it is declared.

\indextext{access control!\idxcode{protected}}%

\tcode{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~\ref{class.protected}).

\indextext{access control!\idxcode{public}}%

\tcode{public};

that is, its name can be used anywhere without access restriction.

\end{itemize}

A member of a class can also access all the names to which the class has access.

A local class of a member function may access

the same names that the member function itself may access.\footnote{Access

permissions are thus transitive and cumulative to nested

and local classes.}

\indextext{access control!member name}%

\indextext{default access control|see{access control, default}}%

\indextext{access control!default}%

Members of a class defined with the keyword

\tcode{class}

are

\tcode{private}

by default.

Members of a class defined with the keywords

\tcode{struct}

or

\tcode{union}

are

\tcode{public}

by default.

\begin{example}

\begin{codeblock}

class X {

int a; // \tcode{X::a} is private by default

};

struct S {

int a; // \tcode{S::a} is public by default

};

\end{codeblock}

\end{example}

Access control is applied uniformly to all names, whether the names are

referred to from declarations or expressions.

\begin{note}

Access control applies to names nominated by

\tcode{friend}

declarations~(\ref{class.friend}) and

\grammarterm{using-declaration}{s}~(\ref{namespace.udecl}).

\end{note}

In the case of overloaded function names, access control is applied to

the function selected by overload resolution.

\begin{note}

Because access control applies to names, if access control is applied to a

typedef name, only the accessibility of the typedef name itself is considered.

The accessibility of the entity referred to by the typedef is not considered.

For example,

\begin{codeblock}

class A {

class B { };

public:

typedef B BB;

};

void f() {

A::BB x; // OK, typedef name \tcode{A::BB} is public

A::B y; // access error, \tcode{A::B} is private

}

\end{codeblock}

\end{note}

It should be noted that it is

\term{access}

to members and base classes that is controlled, not their

\term{visibility}.

Names of members are still visible, and implicit conversions to base

classes are still considered, when those members and base classes are

inaccessible.

The interpretation of a given construct is

established without regard to access control.

If the interpretation

established makes use of inaccessible member names or base classes,

the construct is ill-formed.

All access controls in Clause~\ref{class.access} affect the ability to access a class member

name from the declaration of a particular

entity, including parts of the declaration preceding the name of the entity

being declared and, if the entity is a class, the definitions of members of

the class appearing outside the class's \grammarterm{member-specification}{.}

\begin{note} This access also applies to implicit references to constructors,

conversion functions, and destructors. \end{note}

\begin{example}

\begin{codeblock}

class A {

typedef int I; // private member

I f();

friend I g(I);

static I x;

template<int> struct Q;

template<int> friend struct R;

protected:

struct B { };

};

A::I A::f() { return 0; }

A::I g(A::I p = A::x);

A::I g(A::I p) { return 0; }

A::I A::x = 0;

template<A::I> struct A::Q { };

template<A::I> struct R { };

struct D: A::B, A { };

\end{codeblock}

Here, all the uses of

\tcode{A::I}

are well-formed because

\tcode{A::f},

\tcode{A::x}, and \tcode{A::Q}

are members of class

\tcode{A}

and

\tcode{g}

and \tcode{R} are friends of class

\tcode{A}.

This implies, for example, that access checking on the first use of

\tcode{A::I}

must be deferred until it is determined that this use of

\tcode{A::I}

is as the return type of a member of class

\tcode{A}.

Similarly, the use of \tcode{A::B} as a

\grammarterm{base-specifier} is well-formed because \tcode{D}

is derived from \tcode{A}, so checking of \grammarterm{base-specifier}{s}

must be deferred until the entire \grammarterm{base-specifier-list} has been seen.

\end{example}

\indextext{argument!access checking and default}%

\indextext{access control!default argument}%

The names in a default argument~(\ref{dcl.fct.default}) are

bound at the point of declaration, and access is checked at that

point rather than at any points of use of the default argument.

Access checking for default arguments in function templates and in

member functions of class templates is performed as described in~\ref{temp.inst}.

The names in a default \grammarterm{template-argument}~(\ref{temp.param})

have their access checked in the context in which they appear rather than at any

points of use of the default \grammarterm{template-argument}. \begin{example}

\begin{codeblock}

class B { };

template <class T> class C {

protected:

typedef T TT;

};

template <class U, class V = typename U::TT>

class D : public U { };

D <C<B> >* d; // access error, \tcode{C::TT} is protected

\end{codeblock}

\end{example}

\rSec1[class.access.spec]{Access specifiers}%

\indextext{access specifier}

Member declarations can be labeled by an

\grammarterm{access-specifier}

(Clause~\ref{class.derived}):

\begin{ncbnftab}

access-specifier \terminal{:} member-specification\opt

\end{ncbnftab}

An

\grammarterm{access-specifier}

specifies the access rules for members following it

until the end of the class or until another

\grammarterm{access-specifier}

is encountered.

\begin{example}

\begin{codeblock}

class X {

int a; // \tcode{X::a} is private by default: \tcode{class} used

public:

int b; // \tcode{X::b} is public

int c; // \tcode{X::c} is public

};

\end{codeblock}

\end{example}

Any number of access specifiers is allowed and no particular order is required.

\begin{example}

\begin{codeblock}

struct S {

int a; // \tcode{S::a} is public by default: \tcode{struct} used

protected:

int b; // \tcode{S::b} is protected

private:

int c; // \tcode{S::c} is private

public:

int d; // \tcode{S::d} is public

};

\end{codeblock}

\end{example}

\begin{note} The effect of access control on the order of allocation

of data members is described in~\ref{class.mem}.\end{note}

When a member is redeclared within its class definition,

the access specified at its redeclaration shall

be the same as at its initial declaration.

\begin{example}

\begin{codeblock}

struct S {

class A;

enum E : int;

private:

class A { }; // error: cannot change access

enum E: int { e0 }; // error: cannot change access

};

\end{codeblock}

\end{example}

\begin{note}

In a derived class, the lookup of a base class name will find the

injected-class-name instead of the name of the base class in the scope

in which it was declared. The injected-class-name might be less accessible

than the name of the base class in the scope in which it was declared.

\end{note}

\begin{example}

\begin{codeblock}

class A { };

class B : private A { };

class C : public B {

A* p; // error: injected-class-name \tcode{A} is inaccessible

::A* q; // OK

};

\end{codeblock}

\end{example}

\rSec1[class.access.base]{Accessibility of base classes and base class members}%

\indextext{access control!base class}%

\indextext{access specifier}%

\indextext{base class!\idxcode{private}}%

\indextext{base class!\idxcode{protected}}%

\indextext{base class!\idxcode{public}}

If a class is declared to be a base class (Clause~\ref{class.derived}) for another class using the

\tcode{public}

access specifier, the

\tcode{public}

members of the base class are accessible as

\tcode{public}

members of the derived class and

\tcode{protected}

members of the base class are accessible as

\tcode{protected}

members of the derived class.

If a class is declared to be a base class for another class using the

\tcode{protected}

access specifier, the

\tcode{public}

and

\tcode{protected}

members of the base class are accessible as

\tcode{protected}

members of the derived class.

If a class is declared to be a base class for another class using the

\tcode{private}

access specifier, the

\tcode{public}

and

\tcode{protected}

members of the base class are accessible as

\tcode{private}

members of the derived class\footnote{As specified previously in Clause~\ref{class.access},

private members of a base class remain inaccessible even to derived classes

unless

\tcode{friend}

declarations within the base class definition are used to grant access explicitly.}.

In the absence of an

\grammarterm{access-specifier}

for a base class,

\tcode{public}

is assumed when the derived class is

defined with the \grammarterm{class-key}

\tcode{struct}

and

\tcode{private}

is assumed when the class is

defined with the \grammarterm{class-key}

\tcode{class}.

\begin{example}

\begin{codeblock}

class B { @\commentellip@ };

class D1 : private B { @\commentellip@ };

class D2 : public B { @\commentellip@ };

class D3 : B { @\commentellip@ }; // \tcode{B} private by default

struct D4 : public B { @\commentellip@ };

struct D5 : private B { @\commentellip@ };

struct D6 : B { @\commentellip@ }; // \tcode{B} public by default

class D7 : protected B { @\commentellip@ };

struct D8 : protected B { @\commentellip@ };

\end{codeblock}

Here

\tcode{B}

is a public base of

\tcode{D2},

\tcode{D4},

and

\tcode{D6},

a private base of

\tcode{D1},

\tcode{D3},

and

\tcode{D5},

and a protected base of

\tcode{D7}

and

\tcode{D8}.

\end{example}

\begin{note}

A member of a private base class might be inaccessible as an inherited

member name, but accessible directly.

Because of the rules on pointer conversions~(\ref{conv.ptr}) and explicit casts~(\ref{expr.cast}), a conversion from a pointer to a derived class to a pointer

to an inaccessible base class might be ill-formed if an implicit conversion

is used, but well-formed if an explicit cast is used.

For example,

\begin{codeblock}

class B {

public:

int mi; // non-static member

static int si; // static member

};

class D : private B {

};

class DD : public D {

void f();

};

void DD::f() {

mi = 3; // error: \tcode{mi} is private in \tcode{D}

si = 3; // error: \tcode{si} is private in \tcode{D}

::B b;

b.mi = 3; // OK (\tcode{b.mi} is different from \tcode{this->mi})

b.si = 3; // OK (\tcode{b.si} is different from \tcode{this->si})

::B::si = 3; // OK

::B* bp1 = this; // error: \tcode{B} is a private base class

::B* bp2 = (::B*)this; // OK with cast

bp2->mi = 3; // OK: access through a pointer to \tcode{B}.

}

\end{codeblock}

\end{note}

A base class

\tcode{B}

of

\tcode{N}

is

\term{accessible}

at

\term{R},

if

\begin{itemize}

an invented public member of

\tcode{B}

would be a public member of

\tcode{N}, or

\term{R}

occurs in a member or friend of class

\tcode{N},

and an invented public member of

\tcode{B}

would be a private or protected member of

\tcode{N}, or

\term{R}

occurs in a member or friend of a class

\tcode{P}

derived from

\tcode{N},

and an invented public member of

\tcode{B}

would be a private or protected member of

\tcode{P}, or

there exists a class

\tcode{S}

such that

\tcode{B}

is a base class of

\tcode{S}

accessible at

\term{R}

and

\tcode{S}

is a base class of

\tcode{N}

accessible at

\term{R}.

\end{itemize}

\begin{example}

\begin{codeblock}

class B {

public:

int m;

};

class S: private B {

friend class N;

};

class N: private S {

void f() {

B* p = this; // OK because class \tcode{S} satisfies the fourth condition above: \tcode{B} is a base class of \tcode{N}

// accessible in \tcode{f()} because \tcode{B} is an accessible base class of \tcode{S} and \tcode{S} is an accessible

// base class of \tcode{N}.

}

};

\end{codeblock}

\end{example}

If a base class is accessible, one can implicitly convert a pointer to

a derived class to a pointer to that base class~(\ref{conv.ptr}, \ref{conv.mem}).

\begin{note}

It follows that

members and friends of a class

\tcode{X}

can implicitly convert an

\tcode{X*}

to a pointer to a private or protected immediate base class of

\tcode{X}.

\end{note}

The access to a member is affected by the class in which the member is

named.

This naming class is the class in which the member name was looked

up and found.

\begin{note}

This class can be explicit, e.g., when a

\grammarterm{qualified-id}

is used, or implicit, e.g., when a class member access operator~(\ref{expr.ref}) is used (including cases where an implicit

is

added).

If both a class member access operator and a

\grammarterm{qualified-id}

are used to name the member (as in

\tcode{p->T::m}),

the class naming the member is the class denoted by the

\grammarterm{nested-name-specifier}

of the

\grammarterm{qualified-id}

(that is,

\tcode{T}).

\end{note}

A member

\tcode{m}

is accessible at the point

\term{R}

when named in class

\tcode{N}

if

\begin{itemize}

\tcode{m}

as a member of

\tcode{N}

is public, or

\tcode{m}

as a member of

\tcode{N}

is private, and

\term{R}

occurs in a member or friend of class

\tcode{N},

or

\tcode{m}

as a member of

\tcode{N}

is protected, and

\term{R}

occurs in a member or friend of class

\tcode{N},

or in a member of a class

\tcode{P}

derived from

\tcode{N},

where

\tcode{m}

as a member of

\tcode{P}

is public, private, or protected, or

there exists a base class

\tcode{B}

of

\tcode{N}

that is accessible at

\term{R},

and

\tcode{m}

is accessible at

\term{R}

when named in class

\tcode{B}.

\begin{example}

\begin{codeblock}

class B;

class A {

private:

int i;

friend void f(B*);

};

class B : public A { };

void f(B* p) {

p->i = 1; // OK: \tcode{B*} can be implicitly converted to \tcode{A*}, and \tcode{f} has access to \tcode{i} in \tcode{A}

}

\end{codeblock}

\end{example}

\end{itemize}

If a class member access operator, including an implicit

``\tcode{this->}'',

is used to access a non-static data member or non-static

member function, the reference is ill-formed if the

left operand (considered as a pointer in the

operator case) cannot be implicitly converted to a

pointer to the naming class of the right operand.

\begin{note}

This requirement is in addition to the requirement that

the member be accessible as named.

\end{note}

\rSec1[class.friend]{Friends}%

\indextext{friend function!access and}%

\indextext{access control!friend function}

A friend of a class is a function or class that is

given permission to use the private and protected member names from the class.

A class specifies its friends, if any, by way of friend declarations.

Such declarations give special access rights to the friends, but they

do not make the nominated friends members of the befriending class.

\begin{example}

The following example illustrates the differences between

members and friends:

\indextext{friend function!member function and}%

\indextext{example!friend function}%

\indextext{example!member function}%

\begin{codeblock}

class X {

int a;

friend void friend_set(X*, int);

public:

void member_set(int);

};

void friend_set(X* p, int i) { p->a = i; }

void X::member_set(int i) { a = i; }

void f() {

X obj;

friend_set(&obj,10);

obj.member_set(10);

}

\end{codeblock}

\end{example}

\indextext{friend!class access and}%

Declaring a class to be a friend implies that the names of private and

protected members from the class granting friendship can be accessed in the

\grammarterm{base-specifier}{s} and member declarations of the befriended

class. \begin{example}

\begin{codeblock}

class A {

class B { };

friend class X;

};

struct X : A::B { // OK: \tcode{A::B} accessible to friend

A::B mx; // OK: \tcode{A::B} accessible to member of friend

class Y {

A::B my; // OK: \tcode{A::B} accessible to nested member of friend

};

};

\end{codeblock}

\end{example}

\begin{example}

\begin{codeblock}

class X {

enum { a=100 };

friend class Y;

};

class Y {

int v[X::a]; // OK, \tcode{Y} is a friend of \tcode{X}

};

class Z {

int v[X::a]; // error: \tcode{X::a} is private

};

\end{codeblock}

\end{example}

A class shall not be defined in a friend declaration. \begin{example}

\begin{codeblock}

class A {

friend class B { }; // error: cannot define class in friend declaration

};

\end{codeblock}

\end{example}

A \tcode{friend} declaration that does not declare a function

shall have one of the following forms:

\begin{ncsimplebnf}

\terminal{friend} elaborated-type-specifier \terminal{;}\br

\terminal{friend} simple-type-specifier \terminal{;}\br

\terminal{friend} typename-specifier \terminal{;}

\end{ncsimplebnf}

\begin{note} A \tcode{friend} declaration may be the

\term{declaration} in a \grammarterm{template-declaration}

(Clause~\ref{temp}, \ref{temp.friend}).\end{note} If the

type specifier in a \tcode{friend} declaration designates a (possibly

cv-qualified) class type, that class is declared as a friend; otherwise, the

\tcode{friend} declaration is ignored. \begin{example}

\begin{codeblock}

class C;

typedef C Ct;

class X1 {

friend C; // OK: \tcode{class C} is a friend

};

class X2 {

friend Ct; // OK: \tcode{class C} is a friend

friend D; // error: no type-name \tcode{D} in scope

friend class D; // OK: elaborated-type-specifier declares new class

};

template <typename T> class R {

friend T;

};

R<C> rc; // \tcode{class C} is a friend of \tcode{R<C>}

R<int> Ri; // OK: \tcode{"friend int;"} is ignored

\end{codeblock}

\end{example}

\indextext{friend function!linkage of}%

A function first declared in a friend declaration

has the linkage of the namespace of which it is a member~(\ref{basic.link}).

Otherwise, the function retains its previous linkage~(\ref{dcl.stc}).

\indextext{declaration!overloaded name and \tcode{friend}}%

When a

\tcode{friend}

declaration refers to an overloaded name or operator, only the function specified

by the parameter types becomes a friend.

A member function of a class

\tcode{X}

can be a friend of

a class

\tcode{Y}.

\indextext{member function!friend}%

\begin{example}

\begin{codeblock}

class Y {

friend char* X::foo(int);

friend X::X(char); // constructors can be friends

friend X::~X(); // destructors can be friends

};

\end{codeblock}

\end{example}

\indextext{friend function!inline}%

A function can be defined in a friend declaration of a class if and only if the

class is a non-local class~(\ref{class.local}), the function name is unqualified,

and the function has namespace scope.

\begin{example}

\begin{codeblock}

class M {

friend void f() { } // definition of global \tcode{f}, a friend of \tcode{M},

// not the definition of a member function

};

\end{codeblock}

\end{example}

Such a function is implicitly an inline function~(\ref{dcl.inline}).

A

\tcode{friend}

function defined in a class is in the (lexical) scope of the class in which it is defined.

A friend function defined outside the class is not~(\ref{basic.lookup.unqual}).

No

\grammarterm{storage-class-specifier}

shall appear in the

\grammarterm{decl-specifier-seq}

of a friend declaration.

\indextext{friend!access specifier and}%

A name nominated by a friend declaration shall be accessible in the scope of the

class containing the friend declaration.

The meaning of the friend declaration is the same whether the friend declaration

appears in the

\tcode{private},

\tcode{protected}

or

\tcode{public}~(\ref{class.mem})

portion of the class

\grammarterm{member-specification}.

\indextext{friend!inheritance and}%

Friendship is neither inherited nor transitive.

\begin{example}

\begin{codeblock}

class A {

friend class B;

int a;

};

class B {

friend class C;

};

class C {

void f(A* p) {

p->a++; // error: \tcode{C} is not a friend of \tcode{A} despite being a friend of a friend

}

};

class D : public B {

void f(A* p) {

p->a++; // error: \tcode{D} is not a friend of \tcode{A} despite being derived from a friend

}

};

\end{codeblock}

\end{example}

\indextext{local class!friend}%

\indextext{friend!local class and}%

If a friend declaration appears in a local class~(\ref{class.local}) and the

name specified is an unqualified name, a prior declaration is looked

up without considering scopes that are outside the innermost enclosing

non-class scope.

For a friend function declaration, if there is no

prior declaration, the program is ill-formed.

For a friend class

declaration, if there is no prior declaration, the class that is

specified belongs to the innermost enclosing non-class scope, but if it is

subsequently referenced, its name is not found by name lookup

until a matching declaration is provided in the innermost enclosing

non-class scope.

\begin{example}

\begin{codeblock}

class X;

void a();

void f() {

class Y;

extern void b();

class A {

friend class X; // OK, but \tcode{X} is a local class, not \tcode{::X}

friend class Y; // OK

friend class Z; // OK, introduces local class \tcode{Z}

friend void a(); // error, \tcode{::a} is not considered

friend void b(); // OK

friend void c(); // error

};

X* px; // OK, but \tcode{::X} is found

Z* pz; // error, no \tcode{Z} is found

}

\end{codeblock}

\end{example}

\rSec1[class.protected]{Protected member access}

\indextext{access control!\idxcode{protected}}%

An additional access check beyond those described earlier in Clause~\ref{class.access}

is applied when a non-static data member or non-static member function is a

protected member of its naming class~(\ref{class.access.base})\footnote{This

additional check does not apply to other members,

e.g., static data members or enumerator member constants.}

As described earlier, access to a protected member is granted because the

reference occurs in a friend or member of some class \tcode{C}. If the access is

to form a pointer to member~(\ref{expr.unary.op}), the

\grammarterm{nested-name-specifier} shall denote \tcode{C} or a class derived from

\tcode{C}. All other accesses involve a (possibly implicit) object

expression~(\ref{expr.ref}). In this case, the class of the object expression shall be

\tcode{C} or a class derived from \tcode{C}.

\begin{example}

\begin{codeblock}

class B {

protected:

int i;

static int j;

};

class D1 : public B {

};

class D2 : public B {

friend void fr(B*,D1*,D2*);

void mem(B*,D1*);

};

void fr(B* pb, D1* p1, D2* p2) {

pb->i = 1; // ill-formed

p1->i = 2; // ill-formed

p2->i = 3; // OK (access through a \tcode{D2})

p2->B::i = 4; // OK (access through a \tcode{D2}, even though naming class is \tcode{B})

int B::* pmi_B = &B::i; // ill-formed

int B::* pmi_B2 = &D2::i; // OK (type of \tcode{\&D2::i} is \tcode{int B::*})

B::j = 5; // ill-formed (not a friend of naming class \tcode{B})

D2::j = 6; // OK (because refers to static member)

}

void D2::mem(B* pb, D1* p1) {

pb->i = 1; // ill-formed

p1->i = 2; // ill-formed

i = 3; // OK (access through \tcode{this})

B::i = 4; // OK (access through \tcode{this}, qualification ignored)

int B::* pmi_B = &B::i; // ill-formed

int B::* pmi_B2 = &D2::i; // OK

j = 5; // OK (because \tcode{j} refers to static member)

B::j = 6; // OK (because \tcode{B::j} refers to static member)

}

void g(B* pb, D1* p1, D2* p2) {

pb->i = 1; // ill-formed

p1->i = 2; // ill-formed

p2->i = 3; // ill-formed

}

\end{codeblock}

\end{example}

\rSec1[class.access.virt]{Access to virtual functions}%

\indextext{access control!virtual function}

The access rules (Clause~\ref{class.access}) for a virtual function are determined by its declaration

and are not affected by the rules for a function that later overrides it.

\begin{example}

\begin{codeblock}

class B {

public:

virtual int f();

};

class D : public B {

private:

int f();

};

void f() {

D d;

B* pb = &d;

D* pd = &d;

pb->f(); // OK: \tcode{B::f()} is public, \tcode{D::f()} is invoked

pd->f(); // error: \tcode{D::f()} is private

}

\end{codeblock}

\end{example}

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

(\tcode{B*}

in the example above).

The access of the member function in the class in which it was defined

(\tcode{D}

in the example above) is in general not known.

\rSec1[class.paths]{Multiple access}%

\indextext{access control!multiple access}

If a name can be reached by several paths through a multiple inheritance

graph, the access is that of the path that gives most access.

\begin{example}

\begin{codeblock}

class W { public: void f(); };

class A : private virtual W { };

class B : public virtual W { };

class C : public A, public B {

void f() { W::f(); } // OK

};

\end{codeblock}

Since

\tcode{W::f()}

is available to

\tcode{C::f()}

along the public path through

\tcode{B},

access is allowed.

\end{example}

\rSec1[class.access.nest]{Nested classes}%

\indextext{access control!nested class}%

\indextext{member function!nested class}

A nested class is a member and as such has the same access rights as any other member.

The members of an enclosing class have no special access to members of a nested

class; the usual access rules (Clause~\ref{class.access}) shall be obeyed.

\begin{example}

\indextext{example!nested class definition}%

\begin{codeblock}

class E {

int x;

class B { };

class I {

B b; // OK: \tcode{E::I} can access \tcode{E::B}

int y;

void f(E* p, int i) {

p->x = i; // OK: \tcode{E::I} can access \tcode{E::x}