P3735 R0 partial_sort_n, nth_element_n

[wg21bot]

P3735R0 partial_sort_n, nth_element_n (Jan Schultke)

[foonathan]

SG9 discussed this in Kona on Tuesday: https://wiki.edg.com/bin/view/Wg21kona2025/NotesSG9P3735R0

We want a variant of partial_sort/nth_element that takes a count n instead of an iterator mid and clamps if n is too large (proposed by P3735R0).

SF	F	N	A	SA
2	3	2	0	0

Author: SF
Attendance: 7
Consensus in favor.

We want a variant of partial_sort/nth_element that takes a count n instead of an iterator mid and has a precondition/hardened precondition/erroneous behavior/throws an exception if n is too large.

SF	F	N	A	SA
0	5	0	2	0

Author: A
Attendance: 7
Consensus in favor.
A: It couldn't just be an overload, so we need a separate function entirely just to have this extra check, seems unnecessary.
A: If we say it's EB, then we just get what we have before anyway, and I don't like any of the other options to design the interface.

We don't want to add new std:: variants of partial_sort/nth_element (as opposed to the std::ranges ones).

SF	F	N	A	SA
0	2	3	2	0

Author: N
Attendance: 7
No consensus.

Approval of naming scheme:

Name	Votes
_n	1
_at_most	7
_up_to	5
_clamped	3
_safe	0

Attendance: 7

Forward P3735R0 "partial_sort_n, nth_element_n" to LEWG for inclusion in C++29.

SF	F	N	A	SA
2	4	1	0	0

Author: SF
Attendance: 7
Strong consensus in favor.