Question

I’m trying to write some math code for a project and I stumbled upon a interesting problem:

template<class T> class Radian;
template<class T> class Degree;

template <class T>
class Radian
{
public:
    T mRad;
public:
    Radian(const Degree& d);
};

template <class T>
class Degree
{
public:
    T mDeg;
public:
    Degree(const Radian& d);
};

inline Radian::Radian ( const Degree& d ) 
{

}

inline Degree::Radian ( const Radian& d )
{

}

Note that the code is bigger but I removed unnecessary functions so you can understand it better.

I understand how templates work, I know they are generated at compile time but for this particular example I don’t really understand what syntax I should use. I’m aware of the fact that in the constructors Degree should have a template argument but I’m not sure how to write it. From what I see I might have to use another template but it would be nice to use the same T for both since I won’t convert from Degree_float to Radian_double.

Could anyone shed some light on this?

Thank you.

Answer 1

This is the syntax for declaration:

template <typename T>
class Radian
{
public:
    Radian(Degree<T> const& degree);
    …
};

And this is the syntax for definition:

template <typename T>
inline Radian<T>::Radian(Degree<T> const& degree) {
    …
}

That is, since your class is a template, you need to specify its type argument. The constructor is not a template so you write it as you would a normal constructor, just prefixed with its class name.