Given a vector like:
template<int... elements>
struct vec;
How is it possible to create a metafunction which can do multiply element by element all provided vectors. For example
template<typename ...AllVecs>
struct multiVecs
{
using type = ....
}
where type would execute all products element by element. For example given three vecs:
multiVecs< vec<0,1,2>, vec<1,2,3>, vec<2,3,4> >
i should get a vec<0*1*2, 1*2*3, 2*3*4>
Let's start with the two vector product, and go from there.
template <typename lhs, typename rhs>
struct multiplies;
template <int... lhs, int... rhs>
struct multiplies<vec<lhs...>, vec<rhs...>> {
static_assert(sizeof...(lhs) == sizeof...(rhs), "Vector arity mismatch");
using type = vec<(lhs * rhs)...>
};
template <typename lhs, typename rhs>
using multiplies_t = typename multiplies<lhs, rhs>::type;
This works fine, so now we just fold it over the pack.
Unfortunately C++14 doesn't have fold expressions, so we have to do it longhand.
template <typename...>
struct multiVecs;
template <typename... Ts>
using multiVecs_t = typename multiVecs<Ts...>::type;
template <typename result>
struct multiVecs<result> {
using type = result;
};
template <typename first, typename second, typename... rest>
struct multiVecs<first, second, rest...> {
using type = multiVecs_t<multiplies<first, second>, rest...>;
};