In the following code:
int foo(const int& f) //version 1
{
    int g = f;
    return int(foo(g)); // calls itself, turning into SO
}
int& foo(int& f) //version 2
{
    f *= -1;
    return f;
}
int main()
{
    int f = 11;
    cout << foo(f) << endl;
    cout << foo(22) << endl;
}
The first cout prints -11 as expected; f is a lvalue, so it binds to the second version of foo (although it could bind to 1st version as well, the 2nd version it's a better match).
The second call of foo is using a rvalue as parameter, so the only viable version of foo is the first one. So far, so good. Inside the first version of foo, I made a copy of the parameter so I could call the second version (with a lvalue) and return a copy of it after the call of the second version of foo. The thing is this will turn into a stack overflow; still the first version of foo will be called.
Could someone please explain to me why this happens? I would expect that g inside the first version of foo to bind to the second version of foo when passed as parameter.
It's simple really - foo at that point only means foo(const int& f). There's no second choice. Not yet. Switch up the definitions. Or separate them:
int foo(const int& f);
int& foo(int& f);
int main()
{
    int f = 11;
    cout << foo(f) << endl;
    cout << foo(22) << endl;
}
int foo(const int& f) //version 1
{
    int g = f;
    return int(foo(g)); // calls itself, turning into SO
}
int& foo(int& f) //version 2
{
    f *= -1;
    return f;
}
                        If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With