Consider: [1]
int *func(){
int a[10];
//...
return a;
}
int main(){
int* a = func();
}
//在函数里开数组后往main函数回传指针
[2]
void func(int a*){
//...
}
int main(){
int a[10];
func(a);
}
//在main函数开个数组后传实参到函数里
Two programs can define an array a and use it. Which one is better? Why?在内存分配和性能上有啥区别?更推荐哪种方式?
Two codes can define an array a and use it.Which one is better? Why?
The second is better because the first is plain broken. You return a pointer to an array that will no longer exist once the function returns. In other words a dangling pointer.
Neither is what I would do in C++ I would use one of those constructs:
#include <array>
void function(int values[10])
{
}
template<std::size_t N>
void function_template(int(&values)[N])
{
}
template<std::size_t N>
void function_const_ref(const std::array<int, N>& arr) // pass by const reference if you only want to use values
{
}
template<std::size_t N>
void function_ref(std::array<int, N>& arr)
{
arr[1] = 42;
}
int main()
{
int arr[10]{}; // {} also initializes all values to 0!
function(arr);
function_template(arr); // keeps working even if you change size from 10 to something else
// or use std::array and you ca
std::array<int, 10> arr2{};
function_ref(arr2);
// this is I think the most C++ way
function_const_ref(arr2);
return 0;
}
