Defined in header <algorithm> | ||
---|---|---|
Call signature | ||
template< std::forward_iterator I, std::sentinel_for<I> S, class Proj = std::identity, std::indirect_unary_predicate<std::projected<I, Proj>> Pred > constexpr I partition_point( I first, S last, Pred pred, Proj proj = {} ); | (1) | (since C++20) |
template< ranges::forward_range R, class Proj = std::identity, std::indirect_unary_predicate< std::projected<ranges::iterator_t<R>, Proj>> Pred > constexpr ranges::borrowed_iterator_t<R> partition_point( R&& r, Pred pred, Proj proj = {} ); | (2) | (since C++20) |
Examines the partitioned (as if by ranges::partition
) range [
first
,
last
)
or r
and locates the end of the first partition, that is, the projected element that does not satisfy pred
or last
if all projected elements satisfy pred
.
The function-like entities described on this page are niebloids, that is:
In practice, they may be implemented as function objects, or with special compiler extensions.
first, last | - | iterator-sentinel defining the partially-ordered range to examine |
r | - | the partially-ordered range to examine |
pred | - | predicate to apply to the projected elements |
proj | - | projection to apply to the elements |
The iterator past the end of the first partition within [
first
,
last
)
or the iterator equal to last
if all projected elements satisfy pred
.
Given N = ranges::distance(first, last)
, performs O(log N) applications of the predicate pred
and projection proj
.
However, if sentinels don't model std::sized_sentinel_for<I>
, the number of iterator increments is O(N).
This algorithm is a more general form of ranges::lower_bound
, which can be expressed in terms of ranges::partition_point
with the predicate [&](auto const& e) { return std::invoke(pred, e, value); });
.
#include <algorithm> #include <array> #include <iostream> #include <iterator> auto print_seq = [](auto rem, auto first, auto last) { for (std::cout << rem; first != last; std::cout << *first++ << ' ') {} std::cout << '\n'; }; int main() { std::array v {1, 2, 3, 4, 5, 6, 7, 8, 9}; auto is_even = [](int i) { return i % 2 == 0; }; std::ranges::partition(v, is_even); print_seq("After partitioning, v: ", v.cbegin(), v.cend()); const auto pp = std::ranges::partition_point(v, is_even); const auto i = std::ranges::distance(v.cbegin(), pp); std::cout << "Partition point is at " << i << "; v[" << i << "] = " << *pp << '\n'; print_seq("First partition (all even elements): ", v.cbegin(), pp); print_seq("Second partition (all odd elements): ", pp, v.cend()); }
Possible output:
After partitioning, v: 2 4 6 8 5 3 7 1 9 Partition point is at 4; v[4] = 5 First partition (all even elements): 2 4 6 8 Second partition (all odd elements): 5 3 7 1 9
(C++20) | checks whether a range is sorted into ascending order (niebloid) |
(C++20) | returns an iterator to the first element not less than the given value (niebloid) |
(C++11) | locates the partition point of a partitioned range (function template) |
© cppreference.com
Licensed under the Creative Commons Attribution-ShareAlike Unported License v3.0.
https://en.cppreference.com/w/cpp/algorithm/ranges/partition_point