How do we compose functions that return multiple return values in C++? More specifically, if one function returns a tuple, can we compose this function with another that does not explicitly accept tuples? For example, in the code:
#include <tuple>
#include <iostream>
std::tuple <int,int> tuple_ints(int x,int y) {
return std::tuple <int,int> (x,y);
}
int add(int x,int y) {
return x+y;
}
int main() {
std::cout << add(tuple_ints(1,2)) << std::endl;
}
I'm trying to compose the functions add
and tuple_ints
. This rightly generates the error:
g++ -std=c++11 test01.cpp -o test01
test01.cpp: In function 'int main()':
test01.cpp:17:37: error: cannot convert 'std::tuple<int, int>' to 'int' for argument '1' to 'int add(int, int)'
std::cout << add(tuple_ints(1,2)) << std::endl;
^
Makefile:2: recipe for target 'all' failed
make: *** [all] Error 1
I don't want to modify add
to accept a tuple; I want the definition to stay largely as it is. Is there something else that we can do so that we can compose these two functions?
It turns out that there's a proposal to add this functionality to the standard library under N3802. This is similar to the code provided by @Jarod42. I'm attaching the fixed code, which uses the code out of N3802 for reference. Mostly, the difference is that the code in the proposal appears to handle the perfect forwarding correctly
#include <tuple>
#include <iostream>
#include <utility>
// This comes from N3802
template <typename F, typename Tuple, size_t... I>
decltype(auto) apply_impl(F&& f, Tuple&& t, std::index_sequence<I...>) {
return std::forward<F>(f)(std::get<I>(std::forward<Tuple>(t))...);
}
template <typename F, typename Tuple>
decltype(auto) apply(F&& f, Tuple&& t) {
using Indices =
std::make_index_sequence<std::tuple_size<std::decay_t<Tuple>>::value>;
return apply_impl(std::forward<F>(f), std::forward<Tuple>(t), Indices{});
}
// Now, for our example
std::tuple <int,int> tuple_ints(int x,int y) {
return std::tuple <int,int> (x,y);
}
int add(int x,int y) {
return x+y;
}
int main() {
std::cout << apply(add,tuple_ints(1,2)) << std::endl;
}
Also, in case there was any confusion, this solution requires C++14 for things like std::index_sequence
.
You may add a function to dispatch tuple into argument:
namespace detail
{
template <typename F, typename TUPLE, std::size_t...Is>
auto call(F f, const TUPLE& t, std::index_sequence<Is...>)
-> decltype(f(std::get<Is>(t)...))
{
return f(std::get<Is>(t)...);
}
}
template <typename F, typename TUPLE>
auto call(F f, const TUPLE& t)
-> decltype (detail::call(f, t,
std::make_index_sequence<std::tuple_size<TUPLE>::value>()))
{
return detail::call(f, t,
std::make_index_sequence<std::tuple_size<TUPLE>::value>());
}
And then call it that way
std::cout << call(add, tuple_ints(1,2)) << std::endl;
Live example
You may want to use some anonymous lambda like
[](std::tuple <int,int> t){return add(get<0>(t),get<1>(t));}(tuple_int(1,2))
hopefully the compiler would optimize that code.
Read also about currying.
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