| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | #include "krasnopevtseva_v_hoare_batcher_sort/stl/include/ops_stl.hpp" | ||
| 2 | |||
| 3 | #include <algorithm> | ||
| 4 | #include <cstddef> | ||
| 5 | #include <future> | ||
| 6 | #include <stack> | ||
| 7 | #include <thread> | ||
| 8 | #include <utility> | ||
| 9 | #include <vector> | ||
| 10 | |||
| 11 | #include "krasnopevtseva_v_hoare_batcher_sort/common/include/common.hpp" | ||
| 12 | |||
| 13 | namespace krasnopevtseva_v_hoare_batcher_sort { | ||
| 14 | |||
| 15 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | KrasnopevtsevaVHoareBatcherSortSTL::KrasnopevtsevaVHoareBatcherSortSTL(const InType &in) { |
| 16 | SetTypeOfTask(GetStaticTypeOfTask()); | ||
| 17 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | GetInput() = in; |
| 18 | 48 | GetOutput() = std::vector<int>(); | |
| 19 | 48 | } | |
| 20 | |||
| 21 | 48 | bool KrasnopevtsevaVHoareBatcherSortSTL::ValidationImpl() { | |
| 22 | 48 | return !GetInput().empty(); | |
| 23 | } | ||
| 24 | |||
| 25 | 48 | bool KrasnopevtsevaVHoareBatcherSortSTL::PreProcessingImpl() { | |
| 26 | 48 | GetOutput() = std::vector<int>(); | |
| 27 | 48 | return true; | |
| 28 | } | ||
| 29 | |||
| 30 | ✗ | int KrasnopevtsevaVHoareBatcherSortSTL::GetNumThreads(int n) { | |
| 31 | ✗ | int numthreads = static_cast<int>(std::thread::hardware_concurrency()); | |
| 32 |
1/4✗ Branch 0 not taken.
✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
✗ Branch 3 not taken.
|
48 | if (numthreads == 0) { |
| 33 | ✗ | numthreads = 1; | |
| 34 | } | ||
| 35 | ✗ | return std::min(n, numthreads); | |
| 36 | } | ||
| 37 | |||
| 38 | 48 | void KrasnopevtsevaVHoareBatcherSortSTL::SetupChunks(std::vector<Chunk> &chunks, int n, int numthreads) { | |
| 39 | 48 | int thread_input_size = n / numthreads; | |
| 40 | 48 | int remainder = n % numthreads; | |
| 41 | |||
| 42 | int current_start = 0; | ||
| 43 |
2/2✓ Branch 0 taken 192 times.
✓ Branch 1 taken 48 times.
|
240 | for (int i = 0; i < numthreads; ++i) { |
| 44 | int chunk_size = thread_input_size; | ||
| 45 |
2/2✓ Branch 0 taken 48 times.
✓ Branch 1 taken 144 times.
|
192 | if (i == numthreads - 1) { |
| 46 | 48 | chunk_size += remainder; | |
| 47 | } | ||
| 48 | 192 | chunks.push_back({nullptr, chunk_size, current_start, current_start + chunk_size - 1}); | |
| 49 | current_start += chunk_size; | ||
| 50 | } | ||
| 51 | 48 | } | |
| 52 | |||
| 53 | 48 | void KrasnopevtsevaVHoareBatcherSortSTL::ParallelSortChunks(std::vector<int> &arr, std::vector<Chunk> &chunks) { | |
| 54 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | std::vector<std::future<void>> futures; |
| 55 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | futures.reserve(chunks.size()); |
| 56 | |||
| 57 |
2/2✓ Branch 0 taken 192 times.
✓ Branch 1 taken 48 times.
|
240 | for (auto &chunk : chunks) { |
| 58 |
2/4✓ Branch 1 taken 192 times.
✗ Branch 2 not taken.
✓ Branch 4 taken 192 times.
✗ Branch 5 not taken.
|
576 | futures.push_back(std::async(std::launch::async, [&arr, &chunk]() { QuickSort(arr, chunk.left, chunk.right); })); |
| 59 | } | ||
| 60 | |||
| 61 |
2/2✓ Branch 0 taken 192 times.
✓ Branch 1 taken 48 times.
|
240 | for (auto &f : futures) { |
| 62 |
1/2✓ Branch 1 taken 192 times.
✗ Branch 2 not taken.
|
192 | f.wait(); |
| 63 | } | ||
| 64 | 48 | } | |
| 65 | |||
| 66 | ✗ | void KrasnopevtsevaVHoareBatcherSortSTL::BatcherMergeBlocksStep(int *left_ptr, int &left_sz, int *right_ptr, | |
| 67 | int &right_sz) { | ||
| 68 | 192 | std::inplace_merge(left_ptr, right_ptr, right_ptr + right_sz); | |
| 69 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | left_sz += right_sz; |
| 70 | ✗ | } | |
| 71 | |||
| 72 |
1/2✓ Branch 0 taken 96 times.
✗ Branch 1 not taken.
|
96 | void KrasnopevtsevaVHoareBatcherSortSTL::BatcherMergeLevel(int step, std::vector<Chunk> &chunks, int thread_input_size, |
| 73 | int par_if_greater) { | ||
| 74 | 96 | int pack = static_cast<int>(chunks.size()); | |
| 75 | 96 | bool do_parallel = (thread_input_size / step) > par_if_greater; | |
| 76 | |||
| 77 | 144 | auto merge_pair = [&](int off) { | |
| 78 | 144 | size_t idx1 = static_cast<size_t>(2 * step) * static_cast<size_t>(off); | |
| 79 | 144 | size_t idx2 = idx1 + static_cast<size_t>(step); | |
| 80 | 144 | BatcherMergeBlocksStep(chunks[idx1].ptr, chunks[idx1].size, chunks[idx2].ptr, chunks[idx2].size); | |
| 81 | 144 | chunks[idx1].right = chunks[idx2].right; | |
| 82 | 240 | }; | |
| 83 | |||
| 84 |
1/2✓ Branch 0 taken 96 times.
✗ Branch 1 not taken.
|
96 | if (do_parallel) { |
| 85 | ✗ | std::vector<std::future<void>> futures; | |
| 86 | ✗ | futures.reserve(pack / 2); | |
| 87 | ✗ | for (int off = 0; off < pack / 2; ++off) { | |
| 88 | ✗ | futures.push_back(std::async(std::launch::async, merge_pair, off)); | |
| 89 | } | ||
| 90 | ✗ | for (auto &f : futures) { | |
| 91 | ✗ | f.wait(); | |
| 92 | } | ||
| 93 | ✗ | } else { | |
| 94 |
2/2✓ Branch 0 taken 144 times.
✓ Branch 1 taken 96 times.
|
240 | for (int off = 0; off < pack / 2; ++off) { |
| 95 | 144 | merge_pair(off); | |
| 96 | } | ||
| 97 | } | ||
| 98 | 96 | } | |
| 99 | |||
| 100 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 48 times.
|
48 | bool KrasnopevtsevaVHoareBatcherSortSTL::RunImpl() { |
| 101 | const auto &input = GetInput(); | ||
| 102 | std::size_t size = input.size(); | ||
| 103 | |||
| 104 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 48 times.
|
48 | if (size <= 1) { |
| 105 | ✗ | GetOutput() = input; | |
| 106 | ✗ | return true; | |
| 107 | } | ||
| 108 | |||
| 109 | 48 | std::vector<int> res = input; | |
| 110 | 48 | int n = static_cast<int>(size); | |
| 111 | int numthreads = GetNumThreads(n); | ||
| 112 | |||
| 113 | 48 | std::vector<Chunk> chunks; | |
| 114 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | SetupChunks(chunks, n, numthreads); |
| 115 | |||
| 116 |
2/2✓ Branch 0 taken 192 times.
✓ Branch 1 taken 48 times.
|
240 | for (auto &chunk : chunks) { |
| 117 | 192 | chunk.ptr = res.data() + chunk.left; | |
| 118 | } | ||
| 119 | |||
| 120 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | ParallelSortChunks(res, chunks); |
| 121 | |||
| 122 |
2/2✓ Branch 0 taken 96 times.
✓ Branch 1 taken 48 times.
|
144 | for (int step = 1; static_cast<int>(chunks.size()) > 1; step *= 2) { |
| 123 | int pack = static_cast<int>(chunks.size()); | ||
| 124 |
1/2✓ Branch 1 taken 96 times.
✗ Branch 2 not taken.
|
96 | BatcherMergeLevel(step, chunks, n / numthreads, 32); |
| 125 | |||
| 126 |
2/2✓ Branch 0 taken 48 times.
✓ Branch 1 taken 48 times.
|
96 | if ((pack / 2) - 1 == 0) { |
| 127 | 48 | BatcherMergeBlocksStep(chunks[0].ptr, chunks[0].size, chunks.back().ptr, chunks.back().size); | |
| 128 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | chunks[0].right = chunks.back().right; |
| 129 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | chunks.resize(1); |
| 130 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 48 times.
|
48 | } else if ((pack / 2) % 2 != 0) { |
| 131 | ✗ | size_t idx1 = static_cast<size_t>(2 * step) * static_cast<size_t>((pack / 2) - 2); | |
| 132 | ✗ | size_t idx2 = idx1 + static_cast<size_t>(step); | |
| 133 | ✗ | BatcherMergeBlocksStep(chunks[idx1].ptr, chunks[idx1].size, chunks[idx2].ptr, chunks[idx2].size); | |
| 134 | chunks.erase(chunks.begin() + static_cast<ptrdiff_t>(idx2)); | ||
| 135 | } else { | ||
| 136 | int new_size = pack / 2; | ||
| 137 |
1/2✓ Branch 1 taken 48 times.
✗ Branch 2 not taken.
|
48 | chunks.resize(new_size); |
| 138 | } | ||
| 139 | } | ||
| 140 | |||
| 141 | GetOutput() = std::move(res); | ||
| 142 | return true; | ||
| 143 | } | ||
| 144 | |||
| 145 | 48 | bool KrasnopevtsevaVHoareBatcherSortSTL::PostProcessingImpl() { | |
| 146 | 48 | return true; | |
| 147 | } | ||
| 148 | |||
| 149 | ✗ | int KrasnopevtsevaVHoareBatcherSortSTL::Partition(std::vector<int> &arr, int first, int last) { | |
| 150 | ✗ | int i = first - 1; | |
| 151 | ✗ | int value = arr[last]; | |
| 152 | |||
| 153 | ✗ | for (int j = first; j <= last - 1; ++j) { | |
| 154 | ✗ | if (arr[j] <= value) { | |
| 155 | ✗ | ++i; | |
| 156 | ✗ | std::swap(arr[i], arr[j]); | |
| 157 | } | ||
| 158 | } | ||
| 159 | ✗ | std::swap(arr[i + 1], arr[last]); | |
| 160 | ✗ | return i + 1; | |
| 161 | } | ||
| 162 | |||
| 163 | 160 | void KrasnopevtsevaVHoareBatcherSortSTL::InsertionSort(std::vector<int> &arr, int first, int last) { | |
| 164 |
2/2✓ Branch 0 taken 800 times.
✓ Branch 1 taken 160 times.
|
960 | for (int i = first + 1; i <= last; ++i) { |
| 165 | 800 | int key = arr[i]; | |
| 166 | 800 | int j = i - 1; | |
| 167 |
4/4✓ Branch 0 taken 2328 times.
✓ Branch 1 taken 328 times.
✓ Branch 2 taken 1856 times.
✓ Branch 3 taken 472 times.
|
2656 | while (j >= first && arr[j] > key) { |
| 168 | 1856 | arr[j + 1] = arr[j]; | |
| 169 | 1856 | --j; | |
| 170 | } | ||
| 171 | 800 | arr[j + 1] = key; | |
| 172 | } | ||
| 173 | 160 | } | |
| 174 | |||
| 175 | 192 | void KrasnopevtsevaVHoareBatcherSortSTL::QuickSort(std::vector<int> &arr, int first, int last) { | |
| 176 | std::stack<std::pair<int, int>> stack; | ||
| 177 | stack.emplace(first, last); | ||
| 178 | |||
| 179 |
2/2✓ Branch 0 taken 192 times.
✓ Branch 1 taken 192 times.
|
384 | while (!stack.empty()) { |
| 180 | auto [l, r] = stack.top(); | ||
| 181 | stack.pop(); | ||
| 182 | |||
| 183 |
2/2✓ Branch 0 taken 32 times.
✓ Branch 1 taken 160 times.
|
192 | if (l >= r) { |
| 184 | 192 | continue; | |
| 185 | } | ||
| 186 | |||
| 187 |
1/2✓ Branch 0 taken 160 times.
✗ Branch 1 not taken.
|
160 | if (r - l < 16) { |
| 188 | 160 | InsertionSort(arr, l, r); | |
| 189 | 160 | continue; | |
| 190 | } | ||
| 191 | |||
| 192 | ✗ | int iter = Partition(arr, l, r); | |
| 193 | |||
| 194 | ✗ | if (iter - l < r - iter) { | |
| 195 | ✗ | stack.emplace(iter + 1, r); | |
| 196 | ✗ | stack.emplace(l, iter - 1); | |
| 197 | } else { | ||
| 198 | ✗ | stack.emplace(l, iter - 1); | |
| 199 | ✗ | stack.emplace(iter + 1, r); | |
| 200 | } | ||
| 201 | } | ||
| 202 | 192 | } | |
| 203 | |||
| 204 | } // namespace krasnopevtseva_v_hoare_batcher_sort | ||
| 205 |