I’ve a long version of a program that uses this code and a short version. This is the short version and oddly, this code runs perfectly fine in short version program, but because I get an access violation in the larger version, I get the feeling something is wrong. Does anyone see anything terribly wrong (that might lead to data corruption and/or access violation errors) with the following code? Or is this alright?
char strlist[5][11] = {
{ "file01.txt" },
{ "file02.txt" },
{ "file03.txt" },
{ "file04.txt" },
{ "file05.txt" }
};
typedef struct DYNAMEM_DATA {
char *string;
int strlen;
} DYNAMEM_DATA;
typedef struct DYNAMEM_STRUCT {
struct DYNAMEM_DATA **data;
int num_elements;
} DYNAMEM_STRUCT;
DYNAMEM_STRUCT *create_dynamem_struct(int num_elements, char *strlist)
{
DYNAMEM_STRUCT *ds = (DYNAMEM_STRUCT *)calloc(1, sizeof(DYNAMEM_STRUCT));
wchar_t wstring[128];
char len[3];
int i;
ds->data = (DYNAMEM_DATA **)calloc(num_elements, sizeof(DYNAMEM_DATA *));
ds->num_elements = num_elements;
for(i = 0; i < num_elements; i++) {
ds->data[i] = (DYNAMEM_DATA *)calloc(1, sizeof(DYNAMEM_DATA));
ds->data[i]->string = (char *)calloc(1, strlen(&strlist[i*11])+5);
ds->data[i]->strlen = strlen(&strlist[i*11]);
sprintf(ds->data[i]->string, "%s, %d", &strlist[i*11], ds->data[i]->strlen);
mbstowcs(wstring, ds->data[i]->string, 128);
MessageBox(NULL, wstring, TEXT("Error"), MB_OK);
}
return ds;
}
This code runs, produces 5 ‘Error’ messages (which is to be expected), and doesn’t leak any memory. Running
valgrind(3.7.0) on Mac OS X 10.7.4 (using GCC 4.7.1), it is given a clean bill of health.The problem is not, apparently, in this version of the code.