\rSec0[algorithms]{Algorithms library}

\rSec1[algorithms.general]{General}

This Clause describes components that \Cpp{} programs may use to perform

algorithmic operations on containers\iref{containers} and other sequences.

The following subclauses describe components for

non-modifying sequence operations,

mutating sequence operations,

sorting and related operations,

and algorithms from the ISO C library,

as summarized in \tref{algorithms.summary}.

\begin{libsumtab}{Algorithms library summary}{algorithms.summary}

\ref{algorithms.requirements} & Algorithms requirements & \\

\ref{algorithms.parallel} & Parallel algorithms & \\ \rowsep

\ref{alg.nonmodifying} & Non-modifying sequence operations & \tcode{<algorithm>} \\

\ref{alg.modifying.operations} & Mutating sequence operations & \\

\ref{alg.sorting} & Sorting and related operations & \\ \rowsep

\ref{numeric.ops} & Generalized numeric operations & \tcode{<numeric>} \\ \rowsep

\ref{alg.c.library} & C library algorithms & \tcode{<cstdlib>} \\

\end{libsumtab}

\rSec1[algorithms.requirements]{Algorithms requirements}

All of the algorithms

are separated from the particular implementations of data structures and

are parameterized by iterator types.

Because of this, they can work with program-defined data structures,

as long as these data structures have iterator types

satisfying the assumptions on the algorithms.

The entities defined in the \tcode{std::ranges} namespace in this Clause

are not found by argument-dependent name lookup\iref{basic.lookup.argdep}.

When found by unqualified\iref{basic.lookup.unqual} name lookup

for the \grammarterm{postfix-expression} in a function call\iref{expr.call},

they inhibit argument-dependent name lookup.

\begin{example}

\begin{codeblock}

void foo() {

using namespace std::ranges;

std::vector<int> vec{1,2,3};

find(begin(vec), end(vec), 2); // \#1

}

\end{codeblock}

The function call expression at \tcode{\#1} invokes \tcode{std::ranges::find},

not \tcode{std::find}, despite that

(a) the iterator type returned from \tcode{begin(vec)} and \tcode{end(vec)}

may be associated with namespace \tcode{std} and

(b) \tcode{std::find} is more specialized\iref{temp.func.order} than

\tcode{std::ranges::find} since the former requires

its first two parameters to have the same type.

\end{example}

For purposes of determining the existence of data races,

algorithms shall not modify objects referenced through an iterator argument

unless the specification requires such modification.

Throughout this Clause, where the template parameters are not constrained,

the names of template parameters are used to express type requirements.

\begin{itemize}

If an algorithm's template parameter is named

\tcode{InputIterator},

\tcode{InputIterator1}, or

\tcode{Input\-Iterator2},

the template argument shall meet the

\oldconcept{InputIterator} requirements\iref{input.iterators}.

If an algorithm's template parameter is named

\tcode{OutputIterator},

\tcode{OutputIterator1}, or

\tcode{Output\-Iterator2},

the template argument shall meet the

\oldconcept{OutputIterator} requirements\iref{output.iterators}.

If an algorithm's template parameter is named

\tcode{ForwardIterator},

\tcode{ForwardIterator1}, or

\tcode{Forward\-Iterator2},

the template argument shall meet the

\oldconcept{ForwardIterator} requirements\iref{forward.iterators}.

If an algorithm's template parameter is named

\tcode{BidirectionalIterator},

\tcode{Bidirectional\-Iterator1}, or

\tcode{BidirectionalIterator2},

the template argument shall meet the

\oldconcept{BidirectionalIterator} requirements\iref{bidirectional.iterators}.

If an algorithm's template parameter is named

\tcode{RandomAccessIterator},

\tcode{Random\-AccessIterator1}, or

\tcode{RandomAccessIterator2},

the template argument shall meet the

\oldconcept{RandomAccessIterator} requirements\iref{random.access.iterators}.

\end{itemize}

If an algorithm's \effects element specifies

that a value pointed to by any iterator passed as an argument is modified,

then that algorithm has an additional type requirement:

The type of that argument shall meet

the requirements of a mutable iterator\iref{iterator.requirements}.

\begin{note}

This requirement does not affect arguments that are named

\tcode{OutputIterator}, \tcode{OutputIterator1}, or \tcode{OutputIterator2},

because output iterators must always be mutable, nor

does it affect arguments that are constrained,

for which mutability requirements are expressed explicitly.

\end{note}

Both in-place and copying versions are provided for certain algorithms.%

\footnote{The decision whether to include a copying version was

usually based on complexity considerations.

When the cost of doing the operation dominates the cost of copy,

the copying version is not included.

For example, \tcode{sort_copy} is not included

because the cost of sorting is much more significant,

and users might as well do \tcode{copy} followed by \tcode{sort}.}

When such a version is provided for \textit{algorithm} it is called

\textit{algorithm\tcode{_copy}}.

Algorithms that take predicates end with the suffix \tcode{_if}

(which follows the suffix \tcode{_copy}).

When not otherwise constrained, the \tcode{Predicate} parameter is used

whenever an algorithm expects a function object\iref{function.objects} that,

when applied to the result of dereferencing the corresponding iterator,

returns a value testable as \tcode{true}.

In other words,

if an algorithm takes \tcode{Predicate pred} as its argument and

\tcode{first} as its iterator argument with value type \tcode{T},

it should work correctly in the construct

\tcode{pred(*first)} contextually converted to \tcode{bool}\iref{conv}.

The function object \tcode{pred} shall not apply any non-constant function

through the dereferenced iterator.

Given a glvalue \tcode{u} of type (possibly \tcode{const}) \tcode{T}

that designates the same object as \tcode{*first},

\tcode{pred(u)} shall be a valid expression

that is equal to \tcode{pred(*first)}.

When not otherwise constrained, the \tcode{BinaryPredicate} parameter is used

whenever an algorithm expects a function object that when applied

to the result of dereferencing two corresponding iterators or

to dereferencing an iterator and type \tcode{T}

when \tcode{T} is part of the signature

returns a value testable as \tcode{true}.

In other words,

if an algorithm takes \tcode{BinaryPredicate binary_pred} as its argument and

\tcode{first1} and \tcode{first2} as its iterator arguments

with respective value types \tcode{T1} and \tcode{T2},

it should work correctly in the construct

\tcode{binary_pred(*first1, *first2)} contextually converted to \tcode{bool}\iref{conv}.

Unless otherwise specified,

\tcode{BinaryPredicate} always takes the first iterator's \tcode{value_type}

as its first argument, that is, in those cases when \tcode{T value}

is part of the signature, it should work correctly in the construct

\tcode{binary_pred(*first1, value)} contextually converted to \tcode{bool}\iref{conv}.

\tcode{binary_pred} shall not apply any non-constant function

through the dereferenced iterators.

Given a glvalue \tcode{u} of type (possibly \tcode{const}) \tcode{T1}

that designates the same object as \tcode{*first1}, and

a glvalue \tcode{v} of type (possibly \tcode{const}) \tcode{T2}

that designates the same object as \tcode{*first2},

\tcode{binary_pred(u, *first2)},

\tcode{binary_pred(*first1, v)}, and

\tcode{binary_pred(u, v)}

shall each be a valid expression that is equal to

\tcode{binary_pred(*first1, *first2)}, and

\tcode{binary_pred(u, value)}

shall be a valid expression that is equal to

\tcode{binary_pred(*first1, value)}.

The parameters

\tcode{UnaryOperation},

\tcode{BinaryOperation},

\tcode{BinaryOperation1},

and \tcode{BinaryOperation2}

are used

whenever an algorithm expects a function object\iref{function.objects}.

\begin{note}

Unless otherwise specified, algorithms that take function objects as arguments

are permitted to copy those function objects freely.

Programmers for whom object identity is important should consider

using a wrapper class that points to a noncopied implementation object

such as \tcode{reference_wrapper<T>}\iref{refwrap}, or some equivalent solution.

\end{note}

When the description of an algorithm gives an expression such as

\tcode{*first == value} for a condition, the expression

shall evaluate to either \tcode{true} or \tcode{false} in boolean contexts.

In the description of the algorithms, operator \tcode{+}

is used for some of the iterator categories

for which it does not have to be defined.

In these cases the semantics of \tcode{a + n} are the same as those of

\begin{codeblock}

auto tmp = a;

for (; n < 0; ++n) --tmp;

for (; n > 0; --n) ++tmp;

return tmp;

\end{codeblock}

Similarly, operator \tcode{-} is used

for some combinations of iterators and sentinel types

for which it does not have to be defined.

If \range{a}{b} denotes a range,

the semantics of \tcode{b - a} in these cases are the same as those of

\begin{codeblock}

iter_difference_t<remove_reference_t<decltype(a)>> n = 0;

for (auto tmp = a; tmp != b; ++tmp) ++n;

return n;

\end{codeblock}

and if \range{b}{a} denotes a range, the same as those of

\begin{codeblock}

iter_difference_t<remove_reference_t<decltype(b)>> n = 0;

for (auto tmp = b; tmp != a; ++tmp) --n;

return n;

\end{codeblock}

In the description of algorithm return values,

a sentinel value \tcode{s} denoting the end of a range \range{i}{s}

is sometimes returned where an iterator is expected.

In these cases,

the semantics are as if the sentinel is converted into an iterator using

\tcode{ranges::next(i, s)}.

Overloads of algorithms that take \libconcept{Range} arguments\iref{range.range}

behave as if they are implemented by calling \tcode{ranges::begin} and

\tcode{ranges::end} on the \libconcept{Range}(s) and

dispatching to the overload in namespace \tcode{ranges}

that takes separate iterator and sentinel arguments.

The number and order of deducible template parameters for algorithm declarations

are unspecified, except where explicitly stated otherwise.

\begin{note}

Consequently, the algorithms may not

be called with explicitly-specified template argument lists.

\end{note}

The class templates

\tcode{binary_transform_result},

\tcode{for_each_result},

\tcode{minmax_result},

\tcode{mismatch_result},

\tcode{copy_result}, and

\tcode{partition_copy_result}

have the template parameters, data members, and special members specified below.

They have no base classes or members other than those specified.

\rSec1[algorithms.parallel]{Parallel algorithms}

This subclause describes components that \Cpp{} programs may use

to perform operations on containers and other sequences in parallel.

\rSec2[algorithms.parallel.defns]{Terms and definitions}

A \defn{parallel algorithm} is a function template listed in this document

with a template parameter named \tcode{ExecutionPolicy}.

Parallel algorithms access objects indirectly accessible via their arguments

by invoking the following functions:

\begin{itemize}

All operations of the categories of the iterators

that the algorithm is instantiated with.

Operations on those sequence elements that are required by its specification.

User-provided function objects

to be applied during the execution of the algorithm,

if required by the specification.

Operations on those function objects required by the specification.

\begin{note}

See~\ref{algorithms.requirements}.

\end{note}

\end{itemize}

These functions are herein called \defn{element access functions}.

\begin{example}

The \tcode{sort} function may invoke the following element access functions:

\begin{itemize}

Operations of the random-access iterator of the actual template argument

(as per \ref{random.access.iterators}),

as implied by the name of the template parameter \tcode{RandomAccessIterator}.

The \tcode{swap} function on the elements of the sequence

(as per the preconditions specified in \ref{sort}).

The user-provided \tcode{Compare} function object.

\end{itemize}

\end{example}

A standard library function is \defn{vectorization-unsafe}

if it is specified to synchronize with another function invocation, or

another function invocation is specified to synchronize with it,

and if it is not a memory allocation or deallocation function.

\begin{note}

Implementations must ensure that internal synchronization

inside standard library functions does not prevent forward progress

when those functions are executed by threads of execution

with weakly parallel forward progress guarantees.

\end{note}

\begin{example}

\begin{codeblock}

int x = 0;

std::mutex m;

void f() {

int a[] = {1,2};

std::for_each(std::execution::par_unseq, std::begin(a), std::end(a), [&](int) {

std::lock_guard<mutex> guard(m); // incorrect: \tcode{lock_guard} constructor calls \tcode{m.lock()}

++x;

});

}

\end{codeblock}

The above program may result in two consecutive calls to \tcode{m.lock()}

on the same thread of execution (which may deadlock),

because the applications of the function object are not guaranteed

to run on different threads of execution.

\end{example}

\rSec2[algorithms.parallel.user]{Requirements on user-provided function objects}

Unless otherwise specified,

function objects passed into parallel algorithms as objects of type

\tcode{Predicate},

\tcode{BinaryPredicate},

\tcode{Compare},

\tcode{UnaryOperation},

\tcode{BinaryOperation},

\tcode{BinaryOperation1},

\tcode{BinaryOperation2}, and

the operators used by the analogous overloads to these parallel algorithms

that could be formed by the invocation

with the specified default predicate or operation (where applicable)

shall not directly or indirectly modify objects via their arguments,

nor shall they rely on the identity of the provided objects.

\rSec2[algorithms.parallel.exec]{Effect of execution policies on algorithm execution}

Parallel algorithms have

template parameters named \tcode{ExecutionPolicy}\iref{execpol}

which describe the manner in which the execution of these algorithms may be

parallelized and the manner in which they apply the element access functions.

If an object is modified by an element access function,

the algorithm will perform no other unsynchronized accesses to that object.

The modifying element access functions are those

which are specified as modifying the object.

\begin{note}

For example,

\tcode{swap},

\tcode{++},

\tcode{--},

\tcode{@=}, and

assignments

modify the object.

For the assignment and \tcode{@=} operators, only the left argument is modified.

\end{note}

Unless otherwise stated, implementations may make arbitrary copies of elements

(with type \tcode{T}) from sequences

where \tcode{is_trivially_copy_constructible_v<T>}

and \tcode{is_trivially_destructible_v<T>} are \tcode{true}.

\begin{note}

This implies that user-supplied function objects should not rely on

object identity of arguments for such input sequences.

Users for whom the object identity of the arguments to these function objects

is important should consider using a wrapping iterator

that returns a non-copied implementation object

such as \tcode{reference_wrapper<T>}\iref{refwrap}

or some equivalent solution.

\end{note}

The invocations of element access functions in parallel algorithms invoked with

an execution policy object of type \tcode{execution::sequenced_policy} all occur

in the calling thread of execution.

\begin{note}

The invocations are not interleaved; see~\ref{intro.execution}.

\end{note}

The invocations of element access functions in parallel algorithms invoked with

an execution policy object of type \tcode{execution::unsequenced_policy}

are permitted to execute in an unordered fashion

in the calling thread of execution,

unsequenced with respect to one another in the calling thread of execution.

\begin{note}

This means that multiple function object invocations

may be interleaved on a single thread of execution,

which overrides the usual guarantee from \ref{intro.execution}

that function executions do not overlap with one another.

\end{note}

The behavior of a program is undefined if

it invokes a vectorization-unsafe standard library function

from user code

called from a \tcode{execution::unsequenced_policy} algorithm.

\begin{note}

Because \tcode{execution::unsequenced_policy} allows

the execution of element access functions

to be interleaved on a single thread of execution,

blocking synchronization, including the use of mutexes, risks deadlock.

\end{note}

The invocations of element access functions in parallel algorithms invoked with

an execution policy object of type \tcode{execution::parallel_policy}

are permitted to execute either

in the invoking thread of execution or

in a thread of execution implicitly created by the library

to support parallel algorithm execution.

If the threads of execution created by \tcode{thread}\iref{thread.thread.class}

provide concurrent forward progress guarantees\iref{intro.progress},

then a thread of execution implicitly created by the library will provide

parallel forward progress guarantees;

otherwise, the provided forward progress guarantee is

\impldef{forward progress guarantees for

implicit threads of parallel algorithms (if not defined for \tcode{thread})}.

Any such invocations executing in the same thread of execution

are indeterminately sequenced with respect to each other.

\begin{note}

It is the caller's responsibility to ensure

that the invocation does not introduce data races or deadlocks.

\end{note}

\begin{example}

\begin{codeblock}

int a[] = {0,1};

std::vector<int> v;

std::for_each(std::execution::par, std::begin(a), std::end(a), [&](int i) {

v.push_back(i*2+1); // incorrect: data race

});

\end{codeblock}

The program above has a data race because of the unsynchronized access to the

container \tcode{v}.

\end{example}

\begin{example}

\begin{codeblock}

std::atomic<int> x{0};

int a[] = {1,2};

std::for_each(std::execution::par, std::begin(a), std::end(a), [&](int) {

x.fetch_add(1, std::memory_order::relaxed);

// spin wait for another iteration to change the value of \tcode{x}

while (x.load(std::memory_order::relaxed) == 1) { } // incorrect: assumes execution order

});

\end{codeblock}

The above example depends on the order of execution of the iterations, and

will not terminate if both iterations are executed sequentially

on the same thread of execution.

\end{example}

\begin{example}

\begin{codeblock}

int x = 0;

std::mutex m;

int a[] = {1,2};

std::for_each(std::execution::par, std::begin(a), std::end(a), [&](int) {

std::lock_guard<mutex> guard(m);

++x;

});

\end{codeblock}

The above example synchronizes access to object \tcode{x}

ensuring that it is incremented correctly.

\end{example}

The invocations of element access functions in parallel algorithms invoked with

an execution policy object of type \tcode{execution::parallel_unsequenced_policy} are

permitted to execute

in an unordered fashion in unspecified threads of execution, and

unsequenced with respect to one another within each thread of execution.

These threads of execution are

either the invoking thread of execution

or threads of execution implicitly created by the library;

the latter will provide weakly parallel forward progress guarantees.

\begin{note}

This means that multiple function object invocations may be interleaved

on a single thread of execution,

which overrides the usual guarantee from \ref{intro.execution}

that function executions do not overlap with one another.

\end{note}

The behavior of a program is undefined if

it invokes a vectorization-unsafe standard library function

from user code

called from a \tcode{execution::parallel_unsequenced_policy} algorithm.

\begin{note}

Because \tcode{execution::parallel_unsequenced_policy} allows

the execution of element access functions

to be interleaved on a single thread of execution,

blocking synchronization, including the use of mutexes, risks deadlock.

\end{note}

\begin{note}

The semantics of invocation with

\tcode{execution::unsequenced_policy},

\tcode{execution::parallel_policy}, or

\tcode{execution::parallel_unsequenced_policy}

allow the implementation to fall back to sequential execution

if the system cannot parallelize an algorithm invocation,

e.g., due to lack of resources.

\end{note}

If an invocation of a parallel algorithm uses threads of execution

implicitly created by the library,

then the invoking thread of execution will either

\begin{itemize}

temporarily block

with forward progress guarantee delegation\iref{intro.progress}

on the completion of these library-managed threads of execution, or

eventually execute an element access function;

\end{itemize}

the thread of execution will continue to do so until the algorithm is finished.

\begin{note}

In blocking with forward progress guarantee delegation in this context,

a thread of execution created by the library

is considered to have finished execution

as soon as it has finished the execution

of the particular element access function

that the invoking thread of execution logically depends on.

\end{note}

The semantics of parallel algorithms invoked with an execution policy object of

\impldef{additional execution policies supported by parallel algorithms} type

are \impldef{semantics of parallel algorithms invoked with

imple\-men\-tation-defined execution policies}.

\rSec2[algorithms.parallel.exceptions]{Parallel algorithm exceptions}

During the execution of a parallel algorithm,

if temporary memory resources are required for parallelization

and none are available, the algorithm throws a \tcode{bad_alloc} exception.

During the execution of a parallel algorithm,

if the invocation of an element access function exits via an uncaught exception,

the behavior is determined by the \tcode{ExecutionPolicy}.

\rSec2[algorithms.parallel.overloads]{\tcode{ExecutionPolicy} algorithm overloads}

Parallel algorithms are algorithm overloads.

Each parallel algorithm overload

has an additional template type parameter named \tcode{ExecutionPolicy},

which is the first template parameter.

Additionally, each parallel algorithm overload

has an additional function parameter of type \tcode{ExecutionPolicy\&\&},

which is the first function parameter.

\begin{note}

Not all algorithms have parallel algorithm overloads.

\end{note}

Unless otherwise specified,

the semantics of \tcode{ExecutionPolicy} algorithm overloads

are identical to their overloads without.

Unless otherwise specified,

the complexity requirements of \tcode{ExecutionPolicy} algorithm overloads

are relaxed from the complexity requirements of the overloads without

as follows:

when the guarantee says ``at most \placeholder{expr}'' or

and does not specify the number of assignments or swaps, and

\placeholder{expr} is not already expressed with \bigoh{} notation,

the complexity of the algorithm shall be \bigoh{\placeholder{expr}}.

Parallel algorithms shall not participate in overload resolution unless

\tcode{is_execution_policy_v<remove_cvref_t<ExecutionPolicy>>} is \tcode{true}.

\rSec1[algorithm.syn]{Header \tcode{<algorithm>} synopsis}

\indexhdr{algorithm}%

\begin{codeblock}

#include <initializer_list>

namespace std {

// \ref{alg.nonmodifying}, non-modifying sequence operations

// \ref{alg.all.of}, all of

template<class InputIterator, class Predicate>

constexpr bool all_of(InputIterator first, InputIterator last, Predicate pred);

template<class ExecutionPolicy, class ForwardIterator, class Predicate>

bool all_of(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last, Predicate pred);

namespace ranges {

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryPredicate<projected<I, Proj>> Pred>

constexpr bool all_of(I first, S last, Pred pred, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryPredicate<projected<iterator_t<R>, Proj>> Pred>

constexpr bool all_of(R&& r, Pred pred, Proj proj = {});

}

// \ref{alg.any.of}, any of

template<class InputIterator, class Predicate>

constexpr bool any_of(InputIterator first, InputIterator last, Predicate pred);

template<class ExecutionPolicy, class ForwardIterator, class Predicate>

bool any_of(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last, Predicate pred);

namespace ranges {

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryPredicate<projected<I, Proj>> Pred>

constexpr bool any_of(I first, S last, Pred pred, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryPredicate<projected<iterator_t<R>, Proj>> Pred>

constexpr bool any_of(R&& r, Pred pred, Proj proj = {});

}

// \ref{alg.none.of}, none of

template<class InputIterator, class Predicate>

constexpr bool none_of(InputIterator first, InputIterator last, Predicate pred);

template<class ExecutionPolicy, class ForwardIterator, class Predicate>

bool none_of(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last, Predicate pred);

namespace ranges {

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryPredicate<projected<I, Proj>> Pred>

constexpr bool none_of(I first, S last, Pred pred, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryPredicate<projected<iterator_t<R>, Proj>> Pred>

constexpr bool none_of(R&& r, Pred pred, Proj proj = {});

}

// \ref{alg.foreach}, for each

template<class InputIterator, class Function>

constexpr Function for_each(InputIterator first, InputIterator last, Function f);

template<class ExecutionPolicy, class ForwardIterator, class Function>

void for_each(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last, Function f);

namespace ranges {

template<class I, class F>

struct for_each_result {

[[no_unique_address]] I in;

[[no_unique_address]] F fun;

template<class I2, class F2>

requires ConvertibleTo<const I&, I2> && ConvertibleTo<const F&, F2>

operator for_each_result<I2, F2>() const & {

return {in, fun};

}

template<class I2, class F2>

requires ConvertibleTo<I, I2> && ConvertibleTo<F, F2>

operator for_each_result<I2, F2>() && {

return {std::move(in), std::move(fun)};

}

};

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryInvocable<projected<I, Proj>> Fun>

constexpr for_each_result<I, Fun>

for_each(I first, S last, Fun f, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryInvocable<projected<iterator_t<R>, Proj>> Fun>

constexpr for_each_result<safe_iterator_t<R>, Fun>

for_each(R&& r, Fun f, Proj proj = {});

}

template<class InputIterator, class Size, class Function>

constexpr InputIterator for_each_n(InputIterator first, Size n, Function f);

template<class ExecutionPolicy, class ForwardIterator, class Size, class Function>

ForwardIterator for_each_n(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, Size n, Function f);

// \ref{alg.find}, find

template<class InputIterator, class T>

constexpr InputIterator find(InputIterator first, InputIterator last,

const T& value);

template<class ExecutionPolicy, class ForwardIterator, class T>

ForwardIterator find(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last,

const T& value);

template<class InputIterator, class Predicate>

constexpr InputIterator find_if(InputIterator first, InputIterator last,

Predicate pred);

template<class ExecutionPolicy, class ForwardIterator, class Predicate>

ForwardIterator find_if(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last,

Predicate pred);

template<class InputIterator, class Predicate>

constexpr InputIterator find_if_not(InputIterator first, InputIterator last,

Predicate pred);

template<class ExecutionPolicy, class ForwardIterator, class Predicate>

ForwardIterator find_if_not(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last,

Predicate pred);

namespace ranges {

template<InputIterator I, Sentinel<I> S, class T, class Proj = identity>

requires IndirectRelation<ranges::equal_to, projected<I, Proj>, const T*>

constexpr I find(I first, S last, const T& value, Proj proj = {});

template<InputRange R, class T, class Proj = identity>

requires IndirectRelation<ranges::equal_to, projected<iterator_t<R>, Proj>, const T*>

constexpr safe_iterator_t<R>

find(R&& r, const T& value, Proj proj = {});

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryPredicate<projected<I, Proj>> Pred>

constexpr I find_if(I first, S last, Pred pred, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryPredicate<projected<iterator_t<R>, Proj>> Pred>

constexpr safe_iterator_t<R>

find_if(R&& r, Pred pred, Proj proj = {});

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryPredicate<projected<I, Proj>> Pred>

constexpr I find_if_not(I first, S last, Pred pred, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryPredicate<projected<iterator_t<R>, Proj>> Pred>

constexpr safe_iterator_t<R>

find_if_not(R&& r, Pred pred, Proj proj = {});

}

// \ref{alg.find.end}, find end

template<class ForwardIterator1, class ForwardIterator2>

constexpr ForwardIterator1

find_end(ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2);

template<class ForwardIterator1, class ForwardIterator2, class BinaryPredicate>

constexpr ForwardIterator1

find_end(ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2,

BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2>

ForwardIterator1

find_end(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2);

template<class ExecutionPolicy, class ForwardIterator1,

class ForwardIterator2, class BinaryPredicate>

ForwardIterator1

find_end(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2,

BinaryPredicate pred);

namespace ranges {

template<ForwardIterator I1, Sentinel<I1> S1, ForwardIterator I2, Sentinel<I2> S2,

class Pred = ranges::equal_to, class Proj1 = identity, class Proj2 = identity>

requires IndirectlyComparable<I1, I2, Pred, Proj1, Proj2>

constexpr subrange<I1>

find_end(I1 first1, S1 last1, I2 first2, S2 last2, Pred pred = {},

Proj1 proj1 = {}, Proj2 proj2 = {});

template<ForwardRange R1, ForwardRange R2,

class Pred = ranges::equal_to, class Proj1 = identity, class Proj2 = identity>

requires IndirectlyComparable<iterator_t<R1>, iterator_t<R2>, Pred, Proj1, Proj2>

constexpr safe_subrange_t<R1>

find_end(R1&& r1, R2&& r2, Pred pred = {},

Proj1 proj1 = {}, Proj2 proj2 = {});

}

// \ref{alg.find.first.of}, find first

template<class InputIterator, class ForwardIterator>

constexpr InputIterator

find_first_of(InputIterator first1, InputIterator last1,

ForwardIterator first2, ForwardIterator last2);

template<class InputIterator, class ForwardIterator, class BinaryPredicate>

constexpr InputIterator

find_first_of(InputIterator first1, InputIterator last1,

ForwardIterator first2, ForwardIterator last2,

BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2>

ForwardIterator1

find_first_of(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2);

template<class ExecutionPolicy, class ForwardIterator1,

class ForwardIterator2, class BinaryPredicate>

ForwardIterator1

find_first_of(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2,

BinaryPredicate pred);

namespace ranges {

template<InputIterator I1, Sentinel<I1> S1, ForwardIterator I2, Sentinel<I2> S2,

class Proj1 = identity, class Proj2 = identity,

IndirectRelation<projected<I1, Proj1>,

projected<I2, Proj2>> Pred = ranges::equal_to>

constexpr I1 find_first_of(I1 first1, S1 last1, I2 first2, S2 last2,

Pred pred = {},

Proj1 proj1 = {}, Proj2 proj2 = {});

template<InputRange R1, ForwardRange R2, class Proj1 = identity, class Proj2 = identity,

IndirectRelation<projected<iterator_t<R1>, Proj1>,

projected<iterator_t<R2>, Proj2>> Pred = ranges::equal_to>

constexpr safe_iterator_t<R1>

find_first_of(R1&& r1, R2&& r2,

Pred pred = {},

Proj1 proj1 = {}, Proj2 proj2 = {});

}

// \ref{alg.adjacent.find}, adjacent find

template<class ForwardIterator>

constexpr ForwardIterator

adjacent_find(ForwardIterator first, ForwardIterator last);

template<class ForwardIterator, class BinaryPredicate>

constexpr ForwardIterator

adjacent_find(ForwardIterator first, ForwardIterator last,

BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator>

ForwardIterator

adjacent_find(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last);

template<class ExecutionPolicy, class ForwardIterator, class BinaryPredicate>

ForwardIterator

adjacent_find(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last,

BinaryPredicate pred);

namespace ranges {

template<ForwardIterator I, Sentinel<I> S, class Proj = identity,

IndirectRelation<projected<I, Proj>> Pred = ranges::equal_to>

constexpr I adjacent_find(I first, S last, Pred pred = {},

Proj proj = {});

template<ForwardRange R, class Proj = identity,

IndirectRelation<projected<iterator_t<R>, Proj>> Pred = ranges::equal_to>

constexpr safe_iterator_t<R>

adjacent_find(R&& r, Pred pred = {}, Proj proj = {});

}

// \ref{alg.count}, count

template<class InputIterator, class T>

constexpr typename iterator_traits<InputIterator>::difference_type

count(InputIterator first, InputIterator last, const T& value);

template<class ExecutionPolicy, class ForwardIterator, class T>

typename iterator_traits<ForwardIterator>::difference_type

count(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last, const T& value);

template<class InputIterator, class Predicate>

constexpr typename iterator_traits<InputIterator>::difference_type

count_if(InputIterator first, InputIterator last, Predicate pred);

template<class ExecutionPolicy, class ForwardIterator, class Predicate>

typename iterator_traits<ForwardIterator>::difference_type

count_if(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator first, ForwardIterator last, Predicate pred);

namespace ranges {

template<InputIterator I, Sentinel<I> S, class T, class Proj = identity>

requires IndirectRelation<ranges::equal_to, projected<I, Proj>, const T*>

constexpr iter_difference_t<I>

count(I first, S last, const T& value, Proj proj = {});

template<InputRange R, class T, class Proj = identity>

requires IndirectRelation<ranges::equal_to, projected<iterator_t<R>, Proj>, const T*>

constexpr iter_difference_t<iterator_t<R>>

count(R&& r, const T& value, Proj proj = {});

template<InputIterator I, Sentinel<I> S, class Proj = identity,

IndirectUnaryPredicate<projected<I, Proj>> Pred>

constexpr iter_difference_t<I>

count_if(I first, S last, Pred pred, Proj proj = {});

template<InputRange R, class Proj = identity,

IndirectUnaryPredicate<projected<iterator_t<R>, Proj>> Pred>

constexpr iter_difference_t<iterator_t<R>>

count_if(R&& r, Pred pred, Proj proj = {});

}

// \ref{mismatch}, mismatch

template<class InputIterator1, class InputIterator2>

constexpr pair<InputIterator1, InputIterator2>

mismatch(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2);

template<class InputIterator1, class InputIterator2, class BinaryPredicate>

constexpr pair<InputIterator1, InputIterator2>

mismatch(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2, BinaryPredicate pred);

template<class InputIterator1, class InputIterator2>

constexpr pair<InputIterator1, InputIterator2>

mismatch(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2, InputIterator2 last2);

template<class InputIterator1, class InputIterator2, class BinaryPredicate>

constexpr pair<InputIterator1, InputIterator2>

mismatch(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2, InputIterator2 last2,

BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2>

pair<ForwardIterator1, ForwardIterator2>

mismatch(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2,

class BinaryPredicate>

pair<ForwardIterator1, ForwardIterator2>

mismatch(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2>

pair<ForwardIterator1, ForwardIterator2>

mismatch(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2,

class BinaryPredicate>

pair<ForwardIterator1, ForwardIterator2>

mismatch(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2,

BinaryPredicate pred);

namespace ranges {

template<class I1, class I2>

struct mismatch_result {

[[no_unique_address]] I1 in1;

[[no_unique_address]] I2 in2;

template<class II1, class II2>

requires ConvertibleTo<const I1&, II1> && ConvertibleTo<const I2&, II2>

operator mismatch_result<II1, II2>() const & {

return {in1, in2};

}

template<class II1, class II2>

requires ConvertibleTo<I1, II1> && ConvertibleTo<I2, II2>

operator mismatch_result<II1, II2>() && {

return {std::move(in1), std::move(in2)};

}

};

template<InputIterator I1, Sentinel<I1> S1, InputIterator I2, Sentinel<I2> S2,

class Proj1 = identity, class Proj2 = identity,

IndirectRelation<projected<I1, Proj1>,

projected<I2, Proj2>> Pred = ranges::equal_to>

constexpr mismatch_result<I1, I2>

mismatch(I1 first1, S1 last1, I2 first2, S2 last2, Pred pred = {},

Proj1 proj1 = {}, Proj2 proj2 = {});

template<InputRange R1, InputRange R2,

class Proj1 = identity, class Proj2 = identity,

IndirectRelation<projected<iterator_t<R1>, Proj1>,

projected<iterator_t<R2>, Proj2>> Pred = ranges::equal_to>

constexpr mismatch_result<safe_iterator_t<R1>, safe_iterator_t<R2>>

mismatch(R1&& r1, R2&& r2, Pred pred = {},

Proj1 proj1 = {}, Proj2 proj2 = {});

}

// \ref{alg.equal}, equal

template<class InputIterator1, class InputIterator2>

constexpr bool equal(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2);

template<class InputIterator1, class InputIterator2, class BinaryPredicate>

constexpr bool equal(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2, BinaryPredicate pred);

template<class InputIterator1, class InputIterator2>

constexpr bool equal(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2, InputIterator2 last2);

template<class InputIterator1, class InputIterator2, class BinaryPredicate>

constexpr bool equal(InputIterator1 first1, InputIterator1 last1,

InputIterator2 first2, InputIterator2 last2,

BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2>

bool equal(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2,

class BinaryPredicate>

bool equal(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, BinaryPredicate pred);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2>

bool equal(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2);

template<class ExecutionPolicy, class ForwardIterator1, class ForwardIterator2,

class BinaryPredicate>

bool equal(ExecutionPolicy&& exec, // see \ref{algorithms.parallel.overloads}

ForwardIterator1 first1, ForwardIterator1 last1,

ForwardIterator2 first2, ForwardIterator2 last2,

BinaryPredicate pred);