Reputation: 19
// Foo.h
class Foo
{
public:
Foo();
~Foo();
void exec();
};
// Foo.cpp
Foo::~Foo()
{
// Statements A
exit(0);
}
//main.cpp
int main()
{
Foo foo;
foo.exec();
// Statements B
return 0;
}
So why both Statements A and B get executed? And is the destructor called when we reach return 0?
I have seen variants where void exec();
is int exec();
and the main
function ends with
return foo.exec();
does this calls the destructor?
Because I want to have an object that takes control of the code flow from the main function and end the program later.
Upvotes: 1
Views: 3331
Reputation: 409176
The destructor is called once the main
function returns, i.e. after the return
statement, just like local objects in any other function.
As for return someObject.someFunction();
, the expression in the return
statement must be fully evaluated before the return
statement can actually return anything, because it needs the result of that expression. So if the function contains a long-running loop (like a GUI event loop) then it might take quite some time before the return
statement actually returns.
Upvotes: 2
Reputation: 145269
The destructor, and hence statement A, is executed as part of the processing of the return 0
statement B. This is a simple guarantee in C++, that the destructor of a local object with automatic storage, is executed automatically when the execution leaves the object's scope.
Upvotes: 2