手写三种sort算法

最近要找工作，于是手写了stl风格的三种sort算法，quick sort，merge sort和heap sort，其中quick sort分别实现了递归和非递归两种，代码如下，经过了基本的正确性测试（不保证对所有类型都正确，因为没充分测试过），然后性能测试与C++的std::sort和std::stable_sort做了一些对比代码如下：

 

#include <algorithm>#include <stack>#include <cassert>#include <utility>namespace fatrat{using std::iterator_traits;enum{ _INSERT_SORT_THRESHOLD = 7,_QUICK_SORT_THRESHOLD = 16 };template<typename _RandomAccessIterator,typename _Compare>void insertion_sort(_RandomAccessIterator first,_RandomAccessIterator last, _Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;if( first == last )return;for( _RandomAccessIterator iter = first+1; iter != last; ++iter ){_ValueType val = *iter;if( comp(val,*first) == true ){std::copy_backward(first,iter,iter+1);*first = val;}else{_RandomAccessIterator next,current;next = current = iter;--next;while( comp(val,*next) == true ){*current = *next;current = next;--next;}*current = val;}}}template<typename _RandomAccessIterator>void insertion_sort(_RandomAccessIterator first,_RandomAccessIterator last){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;if( first == last )return;for( _RandomAccessIterator iter = first+1; iter != last; ++iter ){_ValueType val = *iter;if( val < *first ){std::copy_backward(first,iter,iter+1);*first = val;}else{_RandomAccessIterator next,current;current = next = iter;--next;while( val < *next ){*current = *next;current = next;--next;}*current = val;}}}/////////////////////////////////////////////////////////////////////////////////////////////////////template<typename _RandomAccessIterator,typename _RandomAccessIterator2,typename _Compare>void merge_intervals(_RandomAccessIterator first1, _RandomAccessIterator last1,_RandomAccessIterator first2,_RandomAccessIterator last2,_RandomAccessIterator2 result,_Compare comp){while( first1 < last1 && first2 < last2 ){if( comp(*first1,*first2) == true ){*result = *first1;++first1;}else{*result = *first2;++first2;}++result;}while( first1 < last1 ){*result = *first1;++first1;++result;}while( first2 < last2 ){*result = *first2;++first2;++result;}}template<typename _RandomAccessIterator,typename _RandomAccessIterator2>void merge_intervals(_RandomAccessIterator first1, _RandomAccessIterator last1, _RandomAccessIterator first2,_RandomAccessIterator last2, _RandomAccessIterator2 result){while( first1 < last1 && first2 < last2 ){if( *first1 < * first2 ){*result = *first1;++first1;}else{*result = *first2;++first2;}++result;}while( first1 < last1 ){*result = *first1;++first1;++result;}while( first2 < last2 ){*result = *first2;++first2;++result;}}template<typename _RandomAccessIterator,typename _RandomAccessIterator2,typename _Compare>void merge_sort_with_buffer(_RandomAccessIterator first,_RandomAccessIterator last,_RandomAccessIterator2 result,_Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;assert(first <= last);_DistanceTypedistance = last-first;if( distance < _INSERT_SORT_THRESHOLD )insertion_sort(first,last,comp);else{merge_sort_with_buffer(first,first+distance/2,result,comp);merge_sort_with_buffer(first+distance/2,last,result+distance/2,comp);merge_intervals(first,first+distance/2,first+distance/2,last,result,comp);std::copy(result,result+distance,first);}}template<typename _RandomAccessIterator,typename _RandomAccessIterator2>void merge_sort_with_buffer(_RandomAccessIterator first,_RandomAccessIterator last,_RandomAccessIterator2 result){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;assert(first <= last);_DistanceTypedistance = last-first;if( distance < _INSERT_SORT_THRESHOLD )insertion_sort(first,last);else{merge_sort_with_buffer(first,first+distance/2,result);merge_sort_with_buffer(first+distance/2,last,result);merge_intervals(first,first+distance/2,first+distance/2,last,result);std::copy(result,result+distance,first);}}template<typename _RandomAccessIterator,typename _Compare>void merge_sort_recursion(_RandomAccessIterator first,_RandomAccessIterator last, _Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;assert(first <= last);_DistanceTypedistance = last-first;if( distance < _INSERT_SORT_THRESHOLD )insertion_sort(first,last,comp);else{_ValueType*buffer = new _ValueType[last-first];merge_sort_with_buffer(first,first+distance/2,buffer,comp);merge_sort_with_buffer(first+distance/2,last,buffer+distance/2,comp);merge_intervals(first,first+distance/2,first+distance/2,last,buffer,comp);std::copy(buffer,buffer+distance,first);delete[]buffer;}}template<typename _RandomAccessIterator>void merge_sort_recursion(_RandomAccessIterator first,_RandomAccessIterator last){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;assert(first <= last);_DistanceTypedistance = last-first;if( distance < _INSERT_SORT_THRESHOLD )insertion_sort(first,last);else{_ValueType*buffer = new _ValueType[last-first];merge_sort_with_buffer(first,first+distance/2,buffer);merge_sort_with_buffer(first+distance/2,last,buffer+distance/2);merge_intervals(first,first+distance/2,first+distance/2,last,buffer);std::copy(buffer,buffer+distance,first);delete[]buffer;}}/////////////////////////////////////////////////////////////////////////////////////////////////////template<typename _RandomAccessIterator,typename _Compare>void shift_down(_RandomAccessIterator first,_RandomAccessIterator shift,_RandomAccessIterator last,_Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::difference_type_DistanceType;_RandomAccessIterator root,child;_DistanceType len = shift-first;root = shift;//visual studio will check iterator out of range,so we can't simply use child = first+len*2+1child = first+std::min(last-first,len*2+1);while( child < last ){if( child + 1 < last && comp(*child,*(child+1)) )++child;if( comp(*root,*child) ){std::swap(*root,*child);root = child;len = root-first;child = first+std::min(last-first,len*2+1);}elsebreak;}}template<typename _RandomAccessIterator>void shift_down(_RandomAccessIterator first,_RandomAccessIterator shift, _RandomAccessIterator last){typedef typename iterator_traits<_RandomAccessIterator>::difference_type_DistanceType;_RandomAccessIterator root,child;_DistanceType len = shift-first;root = shift;child = first+std::min(last-first,len*2+1);while( child < last ){if( child+1 < last && *child < *(child+1) )++child;if( *root < *child ){std::swap(*root,*child);root = child;len = root-first;child = first+std::min(last-first,len*2+1);}elsebreak;}}template<typename _RandomAccessIterator>void build_heap(_RandomAccessIterator first, _RandomAccessIterator last){typedef typename iterator_traits<_RandomAccessIterator>::difference_type_DistanceType;_RandomAccessIterator shift;_DistanceTypelen = last-first;assert(len >= 0 );for( shift = first+len/2; shift > first; --shift )shift_down(first,shift,last);shift_down(first,first,last);}template<typename _RandomAccessIterator,typename _Compare>void build_heap(_RandomAccessIterator first, _RandomAccessIterator last,_Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::difference_type_DistanceType;_RandomAccessIterator shift;_DistanceTypelen = last-first;assert(len >= 0 );for(shift = first+len/2; shift > first; --shift)shift_down(first,shift,last,comp);shift_down(first,first,last,comp);}template<typename _RandomAccessIterator>void heap_sort(_RandomAccessIterator first,_RandomAccessIterator last){_RandomAccessIterator iter;build_heap(first,last);for( iter = last-1; iter > first; --iter ){std::swap(*first,*iter);shift_down(first,first,iter);}}template<typename _RandomAccessIterator,typename _Compare>void heap_sort(_RandomAccessIterator first,_RandomAccessIterator last, _Compare comp){_RandomAccessIterator iter;build_heap(first,last,comp);for( iter = last-1; iter > first; --iter ){std::swap(*first,*iter);shift_down(first,first,iter,comp);}}//////////////////////////////////////////////////////////////////////////////////////////////////////template<typename _Tp>inline const _Tp& choose_pivot(const _Tp& x, const _Tp& y, const _Tp& z){if( (x > y && y > z) || (z > y && y > x ))return y;else if( (y > x && x > z)||(z > x && x > y))return x;elsereturn z;}template<typename _Tp, typename _Compare>inline const _Tp& choose_pivot(const _Tp& x, const _Tp& y, const _Tp& z, _Compare comp){if( (comp(x,y) && comp(y,z))||(comp(z,y) && comp(y,x)))return y;else if( (comp(y,x) && comp(x,z))||(comp(z,x) && comp(x,y)))return x;elsereturn z;}template<typename _RandomAccessIterator,typename _Tp>_RandomAccessIterator quick_partition(_RandomAccessIterator first,_RandomAccessIterator last, _Tp pivot){while( true ){while( *first < pivot )++first;--last;while( *last > pivot )--last;if( first >= last )return first;std::swap(*first,*last);++first;}}template<typename _RandomAccessIterator,typename _Tp,typename _Compare>_RandomAccessIterator quick_partition(_RandomAccessIterator first,_RandomAccessIterator last,_Tp pivot, _Compare comp){while(true){while(comp(*first,pivot) == true)++first;--last;while(comp(pivot,*last) == true)--last;if( first >= last )return first;std::swap(*first,*last);++first;}}template<typename _RandomAccessIterator,typename _Compare>void quick_sort_recursion(_RandomAccessIterator first,_RandomAccessIterator last,_Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;if( last-first <= (_DistanceType)_QUICK_SORT_THRESHOLD ) insertion_sort(first,last,comp);else{_RandomAccessIterator mid = first+(last-first)/2;_ValueTypepivot = choose_pivot(*first,*mid,*(last-1),comp);mid = quick_partition(first,last,pivot,comp);quick_sort_recursion(first,mid,comp);quick_sort_recursion(mid,last,comp);}}template<typename _RandomAccessIterator>void quick_sort_recursion(_RandomAccessIterator first,_RandomAccessIterator last){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;assert(first <= last);if( last-first <= (_DistanceType)_QUICK_SORT_THRESHOLD ) insertion_sort(first,last);else{_RandomAccessIterator mid = first+(last-first)/2;_ValueTypepivot = choose_pivot(*first,*mid,*(last-1));mid = quick_partition(first,last,pivot);quick_sort_recursion(first,mid);quick_sort_recursion(mid,last);}}template<typename _RandomAccessIterator,typename _Compare>void quick_sort_nonerecu(_RandomAccessIterator first,_RandomAccessIterator last,_Compare comp){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;typedef typename std::pair<_RandomAccessIterator,_RandomAccessIterator> _StackItemType;typedef typename std::stack<_StackItemType> _StackType;assert(first <= last);_StackType stack;_StackItemType item;stack.push(std::make_pair(first,last));while( stack.empty() != true ){item = stack.top();stack.pop();first = item.first;last = item.second;if( last-first <= (_DistanceType)_QUICK_SORT_THRESHOLD )insertion_sort(first,last,comp);else{_RandomAccessIterator mid = first+(last-first)/2;_ValueTypepivot = choose_pivot(*first,*mid,*(last-1),comp);mid = quick_partition(first,last,pivot,comp);stack.push(std::make_pair(first,mid));stack.push(std::make_pair(mid,last));}}}template<typename _RandomAccessIterator>void quick_sort_nonerecu(_RandomAccessIterator first,_RandomAccessIterator last){typedef typename iterator_traits<_RandomAccessIterator>::value_type _ValueType;typedef typename iterator_traits<_RandomAccessIterator>::difference_type _DistanceType;typedef typename std::pair<_RandomAccessIterator,_RandomAccessIterator> _StackItemType;typedef typename std::stack<_StackItemType> _StackType;assert(last-first>=0);_StackType stack;_StackItemType item;stack.push(std::make_pair(first,last));while( stack.empty() != true ){item = stack.top();stack.pop();first = item.first;last = item.second;if( last-first <= (_DistanceType)_QUICK_SORT_THRESHOLD )insertion_sort(first,last);else{_RandomAccessIterator mid = first+(last-first)/2;_ValueTypepivot = choose_pivot(*first,*mid,*(last-1));mid = quick_partition(first,last,pivot);stack.push(std::make_pair(first,mid));stack.push(std::make_pair(mid,last));}}}}

 

进行了基本的测试之后与stl的std::sort，std::stable_sort进行了测试对比 

 

#include "sort_collection.hpp"#include <algorithm>#include <vector>#include <cstdio>#include <unistd.h>#include <sys/time.h>#include <ctime>using namespace std;void make_data( int size,int data_range,vector<int>& vec1){ int data; vec1.clear(); //vec2.clear(); srand((unsigned int)time(0)); for( int i = 0; i < size; i++ ) { data = rand()%data_range; vec1.push_back(data); //vec2.push_back(data); }}vector<int> data_vecs[6];#define MAX_DATA_SIZE 50000000#define TEST_CASE(SORT_METHOD,VECTOR_BASE,STR_NAME) do{ / struct timeval start,end; / double response_time; / gettimeofday(&start,NULL); / SORT_METHOD(VECTOR_BASE.begin(),VECTOR_BASE.end()); / gettimeofday(&end,NULL); / response_time = ((end.tv_sec-start.tv_sec)*1000000+(end.tv_usec-start.tv_usec))/1000000.0; / printf("%s time: %0.6lfs/n",STR_NAME,response_time); / }while(0)int main(){ make_data(MAX_DATA_SIZE,MAX_DATA_SIZE,data_vecs[0]); for( int i = 1; i < 6; i++ ) data_vecs[i] = data_vecs[0]; TEST_CASE(stable_sort,data_vecs[1],"stable_sort"); TEST_CASE(fatrat::merge_sort_recursion,data_vecs[2],"merge_sort_recursion"); TEST_CASE(fatrat::heap_sort,data_vecs[3],"heap_sort"); TEST_CASE(fatrat::quick_sort_recursion,data_vecs[4],"quick_sort_recursion"); TEST_CASE(fatrat::quick_sort_nonerecu,data_vecs[5],"quick_sort_nonerecu"); TEST_CASE(sort,data_vecs[0],"sort"); return 0;}

 

测试随机生成5000w数据，分别使用std::sort,std::stable_sort,以及自己实现的merge sort,heap sort,quick sort(递归与非递归的)进行了

对比编译使用g++ -O3优化，

 

运行的平台为：Linux 2.6.26-2-686-bigmem 16 Intel Unknown 2666MHz processors, 85340.50 total bogomips, 16223M

RAM System library 2.7.0

 

stable_sort time: 7.968942s

merge_sort_recursion time: 8.704957s

heap_sort time: 19.695442s

quick_sort_recursion time: 6.323566s

quick_sort_nonerecu time: 6.283994s

sort time: 5.778701s

 

可以看到heap sort最慢（可能是由于我实现的原因quick sort最快

 

stl的std::sort采用一种混合排序的策略，优化的比较好，因此比我实现的quick快一些还可以看到其实递归与非递归其实差别不大，

quick_sort_recursion和quick_sort_nonerecu在5kw数据下，相差还不到0.1s，因此谨慎怀疑glibc里面的qsort用栈展开是否有必要。

看了一下stl的std::stable_sort，貌似它用merge sort实现的时候，实现比我的方法好一些，因此运行速度也比我的merge sort快了接近1s。