\rSec0[stmt.stmt]{Statements}%

\indextext{statement|(}

\indextext{block statement|see{statement, compound}}

Except as indicated, statements are executed in sequence.

\begin{bnf}

\nontermdef{statement}\br

labeled-statement\br

attribute-specifier-seq\opt expression-statement\br

attribute-specifier-seq\opt compound-statement\br

attribute-specifier-seq\opt selection-statement\br

attribute-specifier-seq\opt iteration-statement\br

attribute-specifier-seq\opt jump-statement\br

declaration-statement\br

attribute-specifier-seq\opt try-block

\end{bnf}

The optional \grammarterm{attribute-specifier-seq} appertains to the respective statement.

\rSec1[stmt.label]{Labeled statement}%

\indextext{statement!labeled}

\indextext{statement!labeled}%

\indextext{\idxcode{:}!label~specifier}%

A statement can be labeled.

\begin{bnf}

\nontermdef{labeled-statement}\br

attribute-specifier-seq\opt identifier \terminal{:} statement\br

attribute-specifier-seq\opt \terminal{case} constant-expression \terminal{:} statement\br

attribute-specifier-seq\opt \terminal{default :} statement

\end{bnf}

The optional \grammarterm{attribute-specifier-seq} appertains to the label. An identifier label declares the identifier. The only use of an

identifier label is as the target of a

\indextext{statement!\idxcode{goto}}%

\tcode{goto}.

\indextext{label!scope~of}%

The scope of a label is the function in which it appears. Labels shall

not be redeclared within a function. A label can be used in a

\tcode{goto} statement before its definition.

\indextext{name~space!label}%

Labels have their own name space and do not interfere with other

identifiers.

\indextext{label!\idxcode{case}}%

\indextext{label!\idxcode{default}}%

Case labels and default labels shall occur only in switch statements.

\rSec1[stmt.expr]{Expression statement}%

\indextext{statement!expression}

Expression statements have the form

\begin{bnf}

\nontermdef{expression-statement}\br

expression\opt \terminal{;}

\end{bnf}

The expression is

a discarded-value expression (Clause~\ref{expr}).

All

\indextext{side effects}%

side effects from an expression statement

are completed before the next statement is executed.

\indextext{statement!null}%

\indextext{statement!empty}%

An expression statement with the expression missing is called a null

statement.

Most statements are expression statements --- usually assignments or

function calls. A null statement is useful to carry a label just before

the \tcode{\}} of a compound statement and to supply a null body to an

iteration statement such as a \tcode{while}

statement~(\ref{stmt.while}).

\rSec1[stmt.block]{Compound statement or block}%

\indextext{statement!compound}%

\indextext{\idxcode{\{\}}!block statement}%

So that several statements can be used where one is expected, the

compound statement (also, and equivalently, called ``block'') is

provided.

\begin{bnf}

\nontermdef{compound-statement}\br

\terminal{\{} statement-seq\opt \terminal{\}}

\end{bnf}

\begin{bnf}

\nontermdef{statement-seq}\br

statement\br

statement-seq statement

\end{bnf}

A compound statement defines a block scope~(\ref{basic.scope}).

A declaration is a \grammarterm{}{statement}~(\ref{stmt.dcl}).

\rSec1[stmt.select]{Selection statements}%

\indextext{statement!selection|(}

Selection statements choose one of several flows of control.

\indextext{statement!\idxcode{if}}%

\indextext{statement!\idxcode{switch}}%

\begin{bnf}

\nontermdef{selection-statement}\br

\terminal{if (} condition \terminal{)} statement\br

\terminal{if (} condition \terminal{)} statement \terminal{else} statement\br

\terminal{switch (} condition \terminal{)} statement

\end{bnf}

\begin{bnf}

\nontermdef{condition}\br

expression\br

attribute-specifier-seq\opt decl-specifier-seq declarator \terminal{=} initializer-clause\br

attribute-specifier-seq\opt decl-specifier-seq declarator braced-init-list

\end{bnf}

See~\ref{dcl.meaning} for the optional \grammarterm{attribute-specifier-seq} in a condition.

In Clause~\ref{stmt.stmt}, the term \term{substatement} refers to

the contained \grammarterm{}{statement} or \grammarterm{}{statement}{s} that appear

in the syntax notation.

\indextext{scope!\idxgram{selection-statement}}%

The substatement in a \grammarterm{selection-statement} (each substatement,

in the \tcode{else} form of the \tcode{if} statement) implicitly defines

a block scope~(\ref{basic.scope}). If the substatement in a

selection-statement is a single statement and not a

\grammarterm{compound-statement,} it is as if it was rewritten to be a

compound-statement containing the original substatement.

\begin{codeblock}

if (x)

int i;

\end{codeblock}

can be equivalently rewritten as

\begin{codeblock}

if (x) {

int i;

}

\end{codeblock}

Thus after the \tcode{if} statement, \tcode{i} is no longer in scope.

\indextext{\idxgram{condition}{s}!rules~for}%

The rules for \grammarterm{}{condition}{s} apply both to

\grammarterm{selection-statement}{s} and to the \tcode{for} and \tcode{while}

statements~(\ref{stmt.iter}). The \grammarterm{}{declarator} shall not

specify a function or an array. If the \tcode{auto} \nonterminal{type-specifier} appears in

the \nonterminal{decl-specifier-seq},

the type of the identifier being declared is deduced from the initializer as described in~\ref{dcl.spec.auto}.

\indextext{statement!declaration in \tcode{if}}%

\indextext{statement!declaration in \tcode{switch}}%

A name introduced by a declaration in a \grammarterm{}{condition} (either

introduced by the \grammarterm{decl-specifier-seq} or the

\grammarterm{}{declarator} of the condition) is in scope from its point of

declaration until the end of the substatements controlled by the

condition. If the name is re-declared in the outermost block of a

substatement controlled by the condition, the declaration that

re-declares the name is ill-formed.

\begin{codeblock}

if (int x = f()) {

int x; // ill-formed, redeclaration of \tcode{x}

}

else {

int x; // ill-formed, redeclaration of \tcode{x}

}

\end{codeblock}

The value of a \grammarterm{}{condition} that is an initialized declaration

in a statement other than a \tcode{switch} statement is the value of the

declared variable

contextually converted to \tcode{bool} (Clause~\ref{conv}).

If that

conversion is ill-formed, the program is ill-formed. The value of a

\grammarterm{}{condition} that is an initialized declaration in a

\tcode{switch} statement is the value of the declared variable if it has

integral or enumeration type, or of that variable implicitly converted

to integral or enumeration type otherwise. The value of a

\grammarterm{}{condition} that is an expression is the value of the

expression, contextually converted to \tcode{bool}

for statements other

than \tcode{switch};

if that conversion is ill-formed, the program is

ill-formed. The value of the condition will be referred to as simply

``the condition'' where the usage is unambiguous.

If a \grammarterm{}{condition} can be syntactically resolved as either an

expression or the declaration of a block-scope name, it is interpreted as a

declaration.

In the \grammarterm{decl-specifier-seq} of a \grammarterm{condition}, each

\grammarterm{decl-specifier} shall be either a \grammarterm{type-specifier}

or \tcode{constexpr}.

\rSec2[stmt.if]{The \tcode{if} statement}%

\indextext{statement!\idxcode{if}}

If the condition~(\ref{stmt.select}) yields \tcode{true} the first

substatement is executed. If the \tcode{else} part of the selection

statement is present and the condition yields \tcode{false}, the second

substatement is executed. If the first substatement is reached via a

label, the condition is not evaluated and the second substatement is

not executed. In the second form of \tcode{if} statement

(the one including \tcode{else}), if the first substatement is also an

\tcode{if} statement then that inner \tcode{if} statement shall contain

an \tcode{else} part.\footnote{In other words, the \tcode{else} is associated with the nearest un-elsed

\tcode{if}.}

\rSec2[stmt.switch]{The \tcode{switch} statement}%

\indextext{statement!\idxcode{switch}}

The \tcode{switch} statement causes control to be transferred to one of

several statements depending on the value of a condition.

The condition shall be of integral type, enumeration type, or class

type. If of class type, the

condition is contextually implicitly converted (Clause~\ref{conv}) to

an integral or enumeration type. Integral promotions are performed. Any

statement within the \tcode{switch} statement can be labeled with one or

more case labels as follows:

\begin{ncbnf}

\indextext{label!\idxcode{case}}%

\terminal{case} constant-expression \terminal{:}

\end{ncbnf}

where the \grammarterm{constant-expression} shall be

a converted constant expression~(\ref{expr.const}) of the

promoted type of the switch condition. No two of the case constants in

the same switch shall have the same value after conversion to the

promoted type of the switch condition.

\indextext{label!\idxcode{default}}%

There shall be at most one label of the form

\begin{codeblock}

default :

\end{codeblock}

within a \tcode{switch} statement.

Switch statements can be nested; a \tcode{case} or \tcode{default} label

is associated with the smallest switch enclosing it.

When the \tcode{switch} statement is executed, its condition is

evaluated and compared with each case constant.

\indextext{label!\idxcode{case}}%

If one of the case constants is equal to the value of the condition,

control is passed to the statement following the matched case label. If

no case constant matches the condition, and if there is a

\indextext{label!\idxcode{default}}%

\tcode{default} label, control passes to the statement labeled by the

default label. If no case matches and if there is no \tcode{default}

then none of the statements in the switch is executed.

\tcode{case} and \tcode{default} labels in themselves do not alter the

flow of control, which continues unimpeded across such labels. To exit

from a switch, see \tcode{break},~\ref{stmt.break}.

Usually, the substatement that is the subject of a switch is compound

and \tcode{case} and \tcode{default} labels appear on the top-level

statements contained within the (compound) substatement, but this is not

required.

\indextext{statement!declaration~in \tcode{switch}}%

Declarations can appear in the substatement of a

\grammarterm{switch-statement}.

\indextext{statement!selection|)}

\rSec1[stmt.iter]{Iteration statements}%

\indextext{statement!iteration|(}

Iteration statements specify looping.

\indextext{statement!\idxcode{while}}%

\indextext{statement!\idxcode{do}}%

\indextext{statement!\idxcode{for}}%

\begin{bnf}

\nontermdef{iteration-statement}\br

\terminal{while (} condition \terminal{)} statement\br

\terminal{do} statement \terminal{while (} expression \terminal{) ;}\br

\terminal{for (} for-init-statement condition\opt \terminal{;} expression\opt \terminal{)} statement\br

\terminal{for (} for-range-declaration \terminal{:} for-range-initializer \terminal{)} statement

\end{bnf}

\begin{bnf}

\nontermdef{for-init-statement}\br

expression-statement\br

simple-declaration

\end{bnf}

\begin{bnf}

\nontermdef{for-range-declaration}\br

attribute-specifier-seq\opt decl-specifier-seq declarator

\end{bnf}

\begin{bnf}

\nontermdef{for-range-initializer}\br

expression\br

braced-init-list

\end{bnf}

See~\ref{dcl.meaning} for the optional \grammarterm{attribute-specifier-seq} in a

\grammarterm{for-range-declaration}.

A \grammarterm{for-init-statement} ends with a semicolon.

The substatement in an \grammarterm{iteration-statement} implicitly defines

a block scope~(\ref{basic.scope}) which is entered and exited each time

through the loop.

\indextext{scope!\idxgram{iteration-statement}}%

If the substatement in an iteration-statement is a single statement and

not a \grammarterm{compound-statement,} it is as if it was rewritten to be

a compound-statement containing the original statement.

\begin{codeblock}

while (--x >= 0)

int i;

\end{codeblock}

can be equivalently rewritten as

\begin{codeblock}

while (--x >= 0) {

int i;

}

\end{codeblock}

Thus after the \tcode{while} statement, \tcode{i} is no longer in scope.

The requirements on \grammarterm{}{condition}{s} in iteration statements are

described in~\ref{stmt.select}.

\rSec2[stmt.while]{The \tcode{while} statement}%

\indextext{statement!\idxcode{while}}

In the \tcode{while} statement the substatement is executed repeatedly

until the value of the condition~(\ref{stmt.select}) becomes

\tcode{false}. The test takes place before each execution of the

substatement.

\indextext{statement!declaration~in \tcode{while}}%

When the condition of a \tcode{while} statement is a declaration, the scope of

the variable that is declared extends from its point of

declaration~(\ref{basic.scope.pdecl}) to the end of the \tcode{while}

\grammarterm{}{statement}. A \tcode{while} statement of the form

\begin{codeblock}

while (T t = x) @\grammarterm{statement}@

\end{codeblock}

is equivalent to

\begin{codeblock}

label:

{ // start of condition scope

T t = x;

if (t) {

@\grammarterm{statement}@

goto label;

}

} // end of condition scope

\end{codeblock}

The variable created in a condition is destroyed and created with each

iteration of the loop.

\begin{codeblock}

struct A {

int val;

A(int i) : val(i) { }

~A() { }

operator bool() { return val != 0; }

};

int i = 1;

while (A a = i) {

// ...

i = 0;

}

\end{codeblock}

In the while-loop, the constructor and destructor are each called twice,

once for the condition that succeeds and once for the condition that

fails.

\rSec2[stmt.do]{The \tcode{do} statement}%

\indextext{statement!\idxcode{do}}

The expression is contextually converted to \tcode{bool} (Clause~\ref{conv});

if that conversion is ill-formed, the program is ill-formed.

In the \tcode{do} statement the substatement is executed repeatedly

until the value of the expression becomes \tcode{false}. The test takes

place after each execution of the statement.

\rSec2[stmt.for]{The \tcode{for} statement}%

\indextext{statement!\idxcode{for}}

The \tcode{for} statement

\begin{ncbnf}

\terminal{for (} for-init-statement condition\opt \terminal{;} expression\opt \terminal{)} statement

\end{ncbnf}

is equivalent to

\begin{ncbnftab}

\terminal{\{}\br

\>for-init-statement\br

\>\terminal{while (} condition \terminal{) \{}\br

\>\>statement\br

\>\>expression \terminal{;}\br

\>\terminal{\}}\br

\terminal{\}}

\end{ncbnftab}

except that names declared in the \grammarterm{for-init-statement} are in

the same declarative-region as those declared in the

\grammarterm{}{condition}, and except that a

\indextext{statement!\tcode{continue}~in \tcode{for}}%

\tcode{continue} in \grammarterm{}{statement} (not enclosed in another

iteration statement) will execute \grammarterm{}{expression} before

re-evaluating \grammarterm{}{condition}.

Thus the first statement specifies initialization for the loop; the

condition~(\ref{stmt.select}) specifies a test, made before each

iteration, such that the loop is exited when the condition becomes

\tcode{false}; the expression often specifies incrementing that is done

after each iteration.

Either or both of the condition and the expression can be omitted. A

missing \grammarterm{}{condition} makes the implied \tcode{while} clause

equivalent to \tcode{while(true)}.

\indextext{statement!declaration~in \tcode{for}}%

\indextext{\idxcode{for}!scope~of declaration~in}%

If the \grammarterm{for-init-statement} is a declaration, the scope of the

name(s) declared extends to the end of the \grammarterm{for-statement}.

\begin{codeblock}

int i = 42;

int a[10];

for (int i = 0; i < 10; i++)

a[i] = i;

int j = i; // \tcode{j = 42}

\end{codeblock}

\rSec2[stmt.ranged]{The range-based \tcode{for} statement}%

\indextext{statement!range~based \idxcode{for}}

For a range-based \tcode{for} statement of the form

\begin{ncbnf}

\terminal{for (} for-range-declaration : expression \terminal{)} statement

\end{ncbnf}

let \textit{range-init} be equivalent to the \grammarterm{expression} surrounded

by parentheses\footnote{this ensures that a top-level comma operator cannot be

reinterpreted as a delimiter between \grammarterm{init-declarator}{s} in the

declaration of \tcode{__range}.}

\begin{ncbnf}

\terminal{(} expression \terminal{)}

\end{ncbnf}

and for a range-based \tcode{for} statement of the form

\begin{ncbnf}

\terminal{for} \terminal{(} for-range-declaration \terminal{:} braced-init-list \terminal{)} statement

\end{ncbnf}

let \textit{range-init} be equivalent to the \grammarterm{braced-init-list}. In each case, a

range-based \tcode{for} statement

is equivalent to

\begin{codeblock}

{

auto && __range = range-init;

for ( auto __begin = begin-expr,

__end = end-expr;

__begin != __end;

++__begin ) {

@\textit{for-range-declaration}@ = *__begin;

@\textit{statement}@

}

}

\end{codeblock}

where \tcode{__range}, \tcode{__begin}, and \tcode{__end} are variables defined for

exposition only, and \tcode{_RangeT} is the type of the

\grammarterm{}{expression}, and \textit{begin-expr} and \textit{end-expr} are

determined as follows:

\begin{itemize}

\item if \tcode{_RangeT} is an array type, \textit{begin-expr} and \textit{end-expr} are

\tcode{__range} and \tcode{__range + __bound}, respectively, where \tcode{__bound} is

the array bound. If \tcode{_RangeT} is an array of unknown size or an array of

incomplete type, the program is ill-formed;

\item if \tcode{_RangeT} is a class type, the \grammarterm{unqualified-id}{s}

\tcode{begin} and \tcode{end} are looked up in the scope of class \tcode{\mbox{_RangeT}}

as if by class member access lookup~(\ref{basic.lookup.classref}), and if either

(or both) finds at least one declaration, \grammarterm{begin-expr} and

\grammarterm{end-expr} are \tcode{__range.begin()} and \tcode{__range.end()},

respectively;

\item otherwise, \textit{begin-expr} and \textit{end-expr} are \tcode{begin(__range)}

and \tcode{end(__range)}, respectively, where \tcode{begin} and \tcode{end} are looked

up in the associated namespaces~(\ref{basic.lookup.argdep}).

\enternote Ordinary unqualified lookup~(\ref{basic.lookup.unqual}) is not

performed. \exitnote

\end{itemize}

\begin{codeblock}

int array[5] = { 1, 2, 3, 4, 5 };

for (int& x : array)

x *= 2;

\end{codeblock}

\indextext{statement!iteration|)}

In the \grammarterm{decl-specifier-seq} of a \grammarterm{for-range-declaration},

each \grammarterm{decl-specifier} shall be either a \grammarterm{type-specifier}

or \tcode{constexpr}.

\rSec1[stmt.jump]{Jump statements}%

\indextext{statement!jump}

Jump statements unconditionally transfer control.

\indextext{statement!jump}%

\indextext{statement!\idxcode{break}}%

\indextext{statement!\idxcode{continue}}%

\indextext{return statement@\tcode{return} statement|see{\tcode{return}}}%

\indextext{\idxcode{return}}%

\indextext{statement!\idxcode{goto}}%

\begin{bnf}

\nontermdef{jump-statement}\br

\terminal{break ;}\br

\terminal{continue ;}\br

\terminal{return} expression\opt \terminal{;}\br

\terminal{return} braced-init-list \terminal{;}\br

\terminal{goto} identifier \terminal{;}

\end{bnf}

\indextext{local~variable!destruction~of}%

\indextext{scope!destructor~and exit~from}%

On exit from a scope (however accomplished), objects with automatic storage

duration~(\ref{basic.stc.auto}) that have been constructed in that scope are destroyed

in the reverse order of their construction. \enternote For temporaries,

see~\ref{class.temporary}. \exitnote Transfer out of a loop, out of a block, or back

past

an initialized variable with automatic storage duration involves the

destruction of objects with automatic storage duration that are in

scope at the point transferred from but not at the point transferred to.

(See~\ref{stmt.dcl} for transfers into blocks).

However, the program can be terminated (by calling

\indextext{\idxcode{exit}}%

\indexlibrary{\idxcode{exit}}%

\tcode{std::exit()} or

\indextext{\idxcode{abort}}%

\indexlibrary{\idxcode{abort}}%

\tcode{std::abort()}~(\ref{support.start.term}), for example) without

destroying class objects with automatic storage duration.

\rSec2[stmt.break]{The \tcode{break} statement}%

\indextext{statement!\idxcode{break}}

The \tcode{break} statement shall occur only in an

\indextext{\idxgram{iteration-statement}}%

\indextext{statement!\idxcode{switch}}%

\grammarterm{iteration-statement} or a \tcode{switch} statement and causes

termination of the smallest enclosing \grammarterm{iteration-statement} or

\tcode{switch} statement; control passes to the statement following the

terminated statement, if any.

\rSec2[stmt.cont]{The \tcode{continue} statement}%

\indextext{statement!\idxcode{continue}}

The

\tcode{continue}

statement shall occur only in an

\indextext{\idxgram{iteration-statement}}%

\grammarterm{iteration-statement}

and causes control to pass to the loop-continuation portion of the

smallest enclosing \grammarterm{iteration-statement}, that is, to the end

of the loop. More precisely, in each of the statements

\begin{minipage}{.30\hsize}

\begin{codeblock}

while (foo) {

{

// ...

}

contin: ;

}

\end{codeblock}

\end{minipage}

\begin{minipage}{.30\hsize}

\begin{codeblock}

do {

{

// ...

}

contin: ;

} while (foo);

\end{codeblock}

\end{minipage}

\begin{minipage}{.30\hsize}

\begin{codeblock}

for (;;) {

{

// ...

}

contin: ;

}

\end{codeblock}

\end{minipage}

a \tcode{continue} not contained in an enclosed iteration statement is

equivalent to \tcode{goto} \tcode{contin}.

\rSec2[stmt.return]{The \tcode{return} statement}%

\indextext{\idxcode{return}}%

\indextext{function~return|see{\tcode{return}}}%

A function returns to its caller by the \tcode{return} statement.

A return statement

with neither an \grammarterm{expression} nor a \grammarterm{braced-init-list}

can be used only in functions

that do not return a value, that is, a function with the return type

\cv\ \tcode{void}, a constructor~(\ref{class.ctor}), or a

destructor~(\ref{class.dtor}).

\indextext{\idxcode{return}!constructor~and}%

\indextext{\idxcode{return}!constructor~and}%

A return statement with an expression of non-void type can be used only

in functions returning a value; the value of the expression is returned

to the caller of the function.

\indextext{conversion!return~type}%

The value of the expression is implicitly converted to the return type of the

function in which it appears. A return statement can involve the

construction and copy or move of a temporary object~(\ref{class.temporary}).

A copy or move operation associated with a return statement may be elided or

considered as an rvalue for the purpose of overload resolution in

selecting a constructor~(\ref{class.copy}).

\exitnote A return statement with a \grammarterm{braced-init-list} initializes the object or reference to be returned from the function by copy-list-initialization~(\ref{dcl.init.list}) from the specified initializer list. \enterexample

\begin{codeblock}

std::pair<std::string,int> f(const char* p, int x) {

return {p,x};

}

\end{codeblock}

Flowing off the end of a function is equivalent to a \tcode{return} with

no value; this results in undefined behavior in a value-returning

function.

A return statement with an expression of type \tcode{void}

can be used only in functions with a return type of

\cvqual{cv} \tcode{void}; the expression is evaluated just before the function

returns to its caller.

\rSec2[stmt.goto]{The \tcode{goto} statement}%

\indextext{statement!\idxcode{goto}}

The \tcode{goto} statement unconditionally transfers control to the

statement labeled by the identifier. The identifier shall be a

\indextext{label}%

label~(\ref{stmt.label}) located in the current function.

\rSec1[stmt.dcl]{Declaration statement}%

\indextext{statement!declaration}

A declaration statement introduces one or more new identifiers into a

block; it has the form

\begin{bnf}

\nontermdef{declaration-statement}\br

block-declaration

\end{bnf}

If an identifier introduced by a declaration was previously declared in

an outer block,

\indextext{declaration~hiding|see{name~hiding}}%

\indextext{name~hiding}%

\indextext{block~structure}%

the outer declaration is hidden for the remainder of the block, after

which it resumes its force.

\indextext{block!initialization~in}%

\indextext{initialization!automatic}%

Variables with automatic storage duration~(\ref{basic.stc.auto}) are

initialized each time their \grammarterm{declaration-statement} is executed.

\indextext{local~variable!destruction~of}%

Variables with automatic storage duration declared in the block are

destroyed on exit from the block~(\ref{stmt.jump}).

\indextext{initialization!jump~past}%

\indextext{\idxcode{goto}!initialization~and}%

It is possible to transfer into a block, but not in a way that bypasses

declarations with initialization. A program that jumps\footnote{The transfer from the condition of a \tcode{switch} statement to a

\tcode{case} label is considered a jump in this respect.}

from a point where a variable with automatic storage duration is

not in scope to a point where it is in scope is ill-formed unless the

variable has scalar type, class type with a trivial default constructor and a trivial destructor, a cv-qualified version of one of these types, or an array of one of the preceding types and is declared without an

\grammarterm{}{initializer}~(\ref{dcl.init}).

\begin{codeblock}

void f() {

// ...

goto lx; // ill-formed: jump into scope of \tcode{a}

// ...

ly:

X a = 1;

// ...

lx:

goto ly; // OK, jump implies destructor

// call for \tcode{a} followed by construction

// again immediately following label \tcode{ly}

}

\end{codeblock}

\indextext{initialization!automatic}%

\indextext{initialization!local \tcode{static}}%

The zero-initialization~(\ref{dcl.init}) of all block-scope

variables with

static storage duration~(\ref{basic.stc.static}) or thread storage duration~(\ref{basic.stc.thread}) is performed before any

other initialization takes place. Constant initialization~(\ref{basic.start.init}) of a

block-scope entity with static storage duration, if applicable, is performed before its block is first

entered. An implementation is permitted to perform early initialization

of other block-scope variables with static or thread storage duration under the same

conditions that an implementation is permitted to statically initialize

a variable with static or thread storage duration in namespace

scope~(\ref{basic.start.init}). Otherwise such a variable is initialized

the first time control passes through its declaration; such a variable is

considered initialized upon the completion of its initialization. If the

initialization exits by throwing an exception, the initialization is not

complete, so it will be tried again the next time control enters the

declaration. If control enters the declaration concurrently while the variable is being initialized, the concurrent execution shall wait for completion of the initialization.\footnote{The implementation must not introduce any deadlock around execution of the initializer.} If control re-enters the declaration recursively while

the variable is being initialized, the behavior is undefined.

\begin{codeblock}

int foo(int i) {

static int s = foo(2*i); // recursive call - undefined

return i+1;

}

\end{codeblock}

\indextext{\idxcode{static}!destruction~of local}%

The destructor for a block-scope object with static or thread storage duration will be

executed if and only if it was constructed.

\ref{basic.start.term} describes the order in which block-scope objects with

static and thread storage duration are destroyed.

\rSec1[stmt.ambig]{Ambiguity resolution}%

\indextext{ambiguity!declaration~versus expression}

There is an ambiguity in the grammar involving

\grammarterm{expression-statement}{s} and \grammarterm{}{declaration}{s}: An

\grammarterm{expression-statement} with a function-style explicit type

conversion~(\ref{expr.type.conv}) as its leftmost subexpression can be

indistinguishable from a \grammarterm{}{declaration} where the first

\grammarterm{}{declarator} starts with a \tcode{(}. In those cases the

\grammarterm{}{statement} is a \grammarterm{}{declaration}. \enternote To

disambiguate, the whole \grammarterm{}{statement} might have to be examined

to determine if it is an \grammarterm{expression-statement} or a

\grammarterm{}{declaration}. This disambiguates many examples.

assuming \tcode{T} is a

\grammarterm{simple-type-specifier}~(\ref{dcl.type}),

\begin{codeblock}

T(a)->m = 7; // expression-statement

T(a)++; // expression-statement

T(a,5)<<c; // expression-statement

T(*d)(int); // declaration

T(e)[5]; // declaration

T(f) = { 1, 2 }; // declaration

T(*g)(double(3)); // declaration

\end{codeblock}

In the last example above, \tcode{g}, which is a pointer to \tcode{T},

is initialized to \tcode{double(3)}. This is of course ill-formed for

semantic reasons, but that does not affect the syntactic analysis.

The remaining cases are \grammarterm{}{declaration}{s}.

\begin{codeblock}

class T {

// ...

public:

T();

T(int);

T(int, int);

};

T(a); // declaration

T(*b)(); // declaration

T(c)=7; // declaration

T(d),e,f=3; // declaration

extern int h;

T(g)(h,2); // declaration

\end{codeblock}

The disambiguation is purely syntactic; that is, the meaning of the

names occurring in such a statement, beyond whether they are

\grammarterm{type-name}{s} or not, is not generally used in or changed by the

disambiguation. Class templates are instantiated as necessary to

determine if a qualified name is a \grammarterm{type-name}. Disambiguation

precedes parsing, and a statement disambiguated as a declaration may be

an ill-formed declaration. If, during parsing, a name in a template

parameter is bound differently than it would be bound during a trial

parse, the program is ill-formed. No diagnostic is required.

This can occur only when the name is declared earlier in the

declaration.

\begin{codeblock}

struct T1 {

T1 operator()(int x) { return T1(x); }

int operator=(int x) { return x; }

T1(int) { }

};

struct T2 { T2(int){ } };

int a, (*(*b)(T2))(int), c, d;

void f() {

// disambiguation requires this to be parsed as a declaration:

T1(a) = 3,

T2(4), // \tcode{T2} will be declared as

(*(*b)(T2(c)))(int(d)); // a variable of type \tcode{T1}

// but this will not allow

// the last part of the

// declaration to parse

// properly since it depends

// on \tcode{T2} being a type-name

}

\end{codeblock}

\indextext{statement|)}