Iterative Quick Sort in C++

Algorithms: Sorting November 15, 2026 C++

Quick sort partitions around a pivot and sorts the two resulting regions.

cpp
  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
#include <stdio.h>
#include <vector>
#include <algorithm>
#include <stdexcept>
#include <cstdint>
#include <limits>
// Iterative Quick Sort in C++
//
// PURPOSE
// Quick sort partitions around a pivot and sorts the two resulting regions. This iterative
// version stores pending intervals explicitly instead of using recursion.
//
// METHOD
// Choose a median-of-three pivot and partition into less-than, equal-to, and greater-than
// regions. Process the smaller region immediately and defer the larger one to keep the
// explicit stack bounded.
//
// COMPLEXITY
// O(n log n) average time, O(n squared) adversarial worst case. Equal-key input takes O(n).
// O(log n) pending intervals; 64 stack slots cover the supported bounded input lengths. Not
// stable.
//
// CONTRACT
// Sorts ascending in place and returns 0. Three-way partition handles repeated keys
// efficiently. Median-of-three reduces common fixed-endpoint pivot problems but does not
// provide introsort's worst-case guarantee.
// Public APIs infer array lengths except in C, which accepts pointer/length pairs. Vertex
// counts and matrix dimensions remain explicit. Arrays must contain signed 32-bit integers.
// Sorted order, correct tree topology, and caller-owned data-structure invariants are
// preconditions; checking them on every query would change query complexity.
// Scalar and shape checks are included. Value-range checks are included where needed to
// prevent invalid indexing or nontermination. Allocation sizes are capped at 4,000,000
// workspace entries. Arithmetic outside Python raises an error on 32-bit overflow; Python
// uses arbitrary precision. Native allocation failures propagate. In-place outputs may be
// partly updated after arithmetic failure. These are single-threaded, in-memory
// implementations.
//
// LANGUAGE DETAILS
// C++17: const vector references for inputs, mutable references for outputs, RAII workspace,
// standard lower_bound for LIS. Invalid inputs throw invalid_argument; overflow throws
// overflow_error; allocation failure propagates bad_alloc.
//
// EXAMPLE
// Expected return: 0.
// Expected a after call: [1, 1, 3, 5].
// Run: g++ -std=c++17 -O2 example.cpp -o example && ./example
// Tests for this exact source artifact run before publication. The accompanying source
// archive includes unit cases, runtime results, and source hashes.

static int alg_min(int a, int b) { return a < b ? a : b; }
static int alg_max(int a, int b) { return a > b ? a : b; }
static void alg_require(bool valid) { if (!valid) throw std::invalid_argument("invalid algorithm input"); }
static int alg_checked(int64_t v) { if (v < INT32_MIN || v > INT32_MAX) throw std::overflow_error("32-bit result overflow"); return (int)v; }
static int alg_add(int a,int b) { return alg_checked((int64_t)a+b); }
static int alg_sub(int a,int b) { return alg_checked((int64_t)a-b); }
static int alg_mul(int a,int b) { return alg_checked((int64_t)a*b); }
static size_t alg_size(int n) { alg_require(n>=0 && n<=4000000); return (size_t)n; }

int quick_sort(std::vector<int>& a) {
    alg_require(a.size() <= 1000000);
    int n = (int)a.size();
    alg_require(((n >= 0) && (n <= 1000000)));
    alg_require(((int)a.size() == n));
    std::vector<int> lefts(alg_size(64), 0);
    std::vector<int> rights(alg_size(64), 0);
    int top = 0;
    int low = 0;
    int high = alg_sub(n, 1);
    int left = 0;
    int right = 0;
    int mid = 0;
    int pivot = 0;
    int temp = 0;
    while (((low < high) || (top > 0))) {
        if ((low >= high)) {
            top = alg_sub(top, 1);
            low = lefts[top];
            high = rights[top];
        }
        mid = alg_add(low, (alg_sub(high, low) / 2));
        pivot = alg_max(alg_min(a[low], a[mid]), alg_min(alg_max(a[low], a[mid]), a[high]));
        left = low;
        mid = low;
        right = high;
        while ((mid <= right)) {
            if ((a[mid] < pivot)) {
                temp = a[left];
                a[left] = a[mid];
                a[mid] = temp;
                left = alg_add(left, 1);
                mid = alg_add(mid, 1);
            } else {
                if ((a[mid] > pivot)) {
                    temp = a[mid];
                    a[mid] = a[right];
                    a[right] = temp;
                    right = alg_sub(right, 1);
                } else {
                    mid = alg_add(mid, 1);
                }
            }
        }
        if ((alg_sub(left, low) < alg_sub(high, right))) {
            if ((alg_add(right, 1) < high)) {
                lefts[top] = alg_add(right, 1);
                rights[top] = high;
                top = alg_add(top, 1);
            }
            high = alg_sub(left, 1);
        } else {
            if ((low < alg_sub(left, 1))) {
                lefts[top] = low;
                rights[top] = alg_sub(left, 1);
                top = alg_add(top, 1);
            }
            low = alg_add(right, 1);
        }
    }
    return 0;
}

int main(void) {
    std::vector<int> a = {5, 1, 3, 1};
    int result = quick_sort(a);
    printf("%d\n", result);
    for (int z = 0; z < 4; z++) printf("%d ", a[z]);
    printf("\n");
    // Expected a: [1, 1, 3, 5]
    return 0;
}