I don't understand what must be a basic C++ concept related to function parameters. It would really help me if someone could identify what is the concept I'm missing - I'd like to study it in order to deeply understand what's happening.
Consider the simple snippet involving two classes A and B. Creating an object of type A leads to creating an object of type B associated to it. Unfortunately, I'm doing something wrong passing a parameter value creating B.
Output of the following code is:
B name: name_value
B name: name_vaHy�
Instead of
B name: name_value
B name: name_value
The name of object B is modified after the constructor call...
#include <string>
#include <iostream>
using namespace std;
class B{
private:
string name;
public:
B(string const& n) : name(n){
showName();
}
void showName(){
cout << "B name: " << name << endl;
}
};
class A{
private:
B* b;
public :
A(string const& n) {
B b_ = B(n);
this->b = &b_;
}
void show(){
this->b->showName();
}
};
int main() {
string input = "name_value";
A a = A(input);
a.show();
return 0;
}
object lifetime
If you create a B object inside the constructor of A, the lifetime of that B object is the constructor of A. It is destroyed at the end of the constructor. Using a pointer to it after the end of the constructor is undefined behavior.
The B object is valid only during the line A a = A(input);
When you run the line a.show(); the B object has been destroyed.
You could write the A class like this instead:
class A{
private:
B b;
public :
A(string const& n) : b(n) {}
void show(){
this->b.showName();
}
};