Question

I have a vector class with a properly overloaded Vect*float operator and am trying to create the global/non-member float*Vect operator as follows: (Note this is a heavily edited sample)

class Vect
{
    public:
        Vect::Vect(const float p_x, const float p_y, const float p_z, const float p_w);
        Vect operator*(const float p_sclr) const;
    private:
        float x;
        float y;
        float z;
        float w;
};
Vect::Vect(const float p_x, const float p_y, const float p_z, const float p_w) {
    x = p_x;
    y = p_y;
    z = p_z;
    w = p_w;
}
Vect Vect::operator*(const float p_sclr) const {
    return Vect( (x * p_sclr), (y * p_sclr), (z * p_sclr), 1); // reset w to 1
}
//Problem Non-MemberOperator
Vect operator*(const float p_sclr, const Vect& p_vect);
Vect operator*(const float p_sclr, const Vect& p_vect) {
    return p_vect * p_sclr;
}

But when I go to test the operator with the call:

Vect A(2.0f, 3.0f, 4.0f, 5.0f);
float s = 5.0f;
Vect C, D;
C = A * s; // Fine
D = s * A; // Error as below

I receive the following compile error:

error C2678: binary ‘*’ : no operator found which takes a left-hand operand of type ‘float’ (or there is no acceptable conversion)

Can anyone provide insight to why this happens? The MS documentation is available at http://msdn.microsoft.com/en-us/library/ys0bw32s(v=VS.90).aspx and isn’t very helpful Visual Studio 2008. This is the only compile error or warning I receive.

Answer 1

The final solution was to go with a friend function because the variables were in p_vect were private:

//class definition
friend Vect operator*(const float scale, const Vect& p_vect);

//function
Vect operator*(const float p_sclr, const Vect& p_vect) {
   return Vect( (p_vect.x * p_sclr), (p_vect.y * p_sclr), (p_vect.z * p_sclr), 1.0f);
}