本文介绍了Mergesort-细分错误的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！
问题描述
代码目录结构，
./Computing$ ls -LR
.:
list file.txt mergeSort.c program.exe type.h
./list:
arrayImpl.c linkedListImpl.c list.h
编译过程:
Compilation procedure:
$./Computing
gcc -Wall -Werror -DARRAY -I. mergeSort.c ./list/*.c -o program
此处是包含文件mergeSort.c，list/*，type.h
使用List的给定表示形式，
Here is the complete code having files, mergeSort.c, list/*, type.h
With the given representation of List,
typedef struct List{
void **array;
/* Housekeeping - Array enhancement/shrink */
int lastItemPosition;
int size;
}List;
mergesort在下面的list->array上执行，其中aux维护array的浅表副本
mergesort is performed below on list->array, where aux maintains the shallow copy of array
void mergeSort(List *list, size_t listSize, isLess less){
if(list != NULL){
void **aux = malloc(list->size * sizeof(void*)); //Auxillary shadow copy
if(aux != NULL){
printf("Size of list: %d\n", listSize);
mSort(list->array, aux, 0, listSize-1, less);
}else{
fprintf(stderr, "mergeSort() - Malloc failure");
exit(EXIT_FAILURE);
}
}else{
fprintf(stderr, "mergeSort() - List is NULL");
}
}
static void mSort(void **array, void **aux, int low, int high, isLess less){
if(high <= low) return;
int mid = (low + high)/2;
mSort(array, aux, low, mid, less);
mSort(array, aux, mid+1, high, less);
merge(array, aux, low, mid, high, less);
}
static void merge(void **array, void **aux, int low, int mid, int high, isLess less){
for(int index = 0; index <= high; index++){
aux[index] = array[index]; //Shallow copy
}
printf("Low-%d, Mid-%d, High-%d\n", low, mid, high);
int leftIndex = low; int rightIndex = mid+1;
printf("leftIndex-%d, rightIndex-%d\n", leftIndex, rightIndex);
for(int index = 0; index <= high; index++){
if(leftIndex > mid) /* right array exhausted */ array[index] = aux[rightIndex++];
else if(rightIndex > high) /*left array exhausted*/ array[index] = aux[leftIndex++];
else if( less(aux[rightIndex], aux[leftIndex]) ) array[index] = aux[rightIndex++];
else array[index] = aux[leftIndex++];
}
}
使用代码所在的地方
where the use code is,
bool less(const void *key, const void *item){
printf("\nIn less function\n");
printf("left-%d, Right-%d\n\n", ((Record *)key)->age, ((Record *)item)->age);
if( ((Record *)key)->age < ((Record *)item)->age ){
printf("Return true\n");
return true;
}else{
printf("Return false\n");
return false;
}
}
int main(){
FILE *pFile = fopen("file.txt", "r");
checkHandleFailure(pFile, FILE_HANDLE_FAILURE);
char buf[MAX_RECORD_SIZE];
DBCache *cache = initCache(pFile);
readHeader(pFile, cache, buf);
readData(pFile, cache, buf);
printRecords(cache);
printf("Before calling mergesort() \n");
mergeSort(cache->records, listGetSize(cache->records), less);
}
实际输出为:
Actual output is:
$ ./program.exe
Age,LastName,FirstName
------------------------------
50,B,A
30,A,B
20,X,D
10,F,A
90,V,E
60,N,M
Records#: 6
Before calling mergesort()
Size of list: 6
Low-0, Mid-0, High-1
leftIndex-0, rightIndex-1
In less function
left-30, Right-50
Return true
Low-0, Mid-1, High-2
leftIndex-0, rightIndex-2
In less function
left-20, Right-30
Return true
Low-3, Mid-3, High-4
leftIndex-3, rightIndex-4
In less function
left-90, Right-10
Return false
Low-3, Mid-4, High-5
leftIndex-3, rightIndex-5
Segmentation fault (core dumped)
In less function
如何解决此问题?
How to resolve this problem?
Cygwin不支持使用gdb的coredump跟踪格式，提供了跟踪(上方)
推荐答案
合并中的两个for循环应该从低位开始，而不是从零开始.从0开始的第二个for循环可能导致分段错误.从0开始的第一个for循环不应引起分段错误，但会浪费额外的时间.
The two for loops in merge should start at low, not at zero. The second for loop starting at 0 is probably causing the segmentation fault. The first for loop starting at 0 shouldn't cause a segmentation fault, but it consumes extra time.
static void merge(....)
/* ... */
for(int index = low; index <= high; index++){ // low not 0
aux[index] = array[index];
}
/* ... */
for(int index = low; index <= high; index++){ // low not 0
if(leftIndex > mid) /* ... */
/* ... */
这篇关于Mergesort-细分错误的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！