Type erasing type erasure, `any` questions? - c++

So, suppose I want to type erase using type erasure.
I can create pseudo-methods for variants that enable a natural:
pseudo_method print = [](auto&& self, auto&& os){ os << self; };
std::variant<A,B,C> var = // create a variant of type A B or C
(var->*print)(std::cout); // print it out without knowing what it is
My question is, how do I extend this to a std::any?
It cannot be done "in the raw". But at the point where we assign to/construct a std::any we have the type information we need.
So, in theory, an augmented any:
template<class...OperationsToTypeErase>
struct super_any {
std::any data;
// or some transformation of OperationsToTypeErase?
std::tuple<OperationsToTypeErase...> operations;
// ?? what for ctor/assign/etc?
};
could somehow automatically rebind some code such that the above type of syntax would work.
Ideally it would be as terse in use as the variant case is.
template<class...Ops, class Op,
// SFINAE filter that an op matches:
std::enable_if_t< std::disjunction< std::is_same<Ops, Op>... >{}, int>* =nullptr
>
decltype(auto) operator->*( super_any<Ops...>& a, any_method<Op> ) {
return std::get<Op>(a.operations)(a.data);
}
Now can I keep this to a type, yet reasonably use the lambda syntax to keep things simple?
Ideally I want:
any_method<void(std::ostream&)> print =
[](auto&& self, auto&& os){ os << self; };
using printable_any = make_super_any<&print>;
printable_any bob = 7; // sets up the printing data attached to the any
int main() {
(bob->*print)(std::cout); // prints 7
bob = 3.14159;
(bob->*print)(std::cout); // prints 3.14159
}
or similar syntax. Is this impossible? Infeasible? Easy?

This is a solution that uses C++14 and boost::any, as I don't have a C++17 compiler.
The syntax we end up with is:
const auto print =
make_any_method<void(std::ostream&)>([](auto&& p, std::ostream& t){ t << p << "\n"; });
super_any<decltype(print)> a = 7;
(a->*print)(std::cout);
which is almost optimal. With what I believe to be simple C++17 changes, it should look like:
constexpr any_method<void(std::ostream&)> print =
[](auto&& p, std::ostream& t){ t << p << "\n"; };
super_any<&print> a = 7;
(a->*print)(std::cout);
In C++17 I'd improve this by taking a auto*... of pointers to any_method instead of the decltype noise.
Inheriting publicly from any is a bit risky, as if someone takes the any off the top and modifies it, the tuple of any_method_data will be out of date. Probably we should just mimic the entire any interface rather than inherit publicly.
#dyp wrote a proof of concept in comments to the OP. This is based off his work, cleaned up with value-semantics (stolen from boost::any) added. #cpplearner's pointer-based solution was used to shorten it (thanks!), and then I added the vtable optimization on top of that.
First we use a tag to pass around types:
template<class T>struct tag_t{constexpr tag_t(){};};
template<class T>constexpr tag_t<T> tag{};
This trait class gets the signature stored with an any_method:
This creates a function pointer type, and a factory for said function pointers, given an any_method:
template<class any_method, class Sig=any_sig_from_method<any_method>>
struct any_method_function;
template<class any_method, class R, class...Args>
struct any_method_function<any_method, R(Args...)>
{
using type = R(*)(boost::any&, any_method const*, Args...);
template<class T>
type operator()( tag_t<T> )const{
return [](boost::any& self, any_method const* method, Args...args) {
return (*method)( boost::any_cast<T&>(self), decltype(args)(args)... );
};
}
};
Now we don't want to store a function pointer per operation in our super_any. So we bundle up the function pointers into a vtable:
template<class...any_methods>
using any_method_tuple = std::tuple< typename any_method_function<any_methods>::type... >;
template<class...any_methods, class T>
any_method_tuple<any_methods...> make_vtable( tag_t<T> ) {
return std::make_tuple(
any_method_function<any_methods>{}(tag<T>)...
);
}
template<class...methods>
struct any_methods {
private:
any_method_tuple<methods...> const* vtable = 0;
template<class T>
static any_method_tuple<methods...> const* get_vtable( tag_t<T> ) {
static const auto table = make_vtable<methods...>(tag<T>);
return &table;
}
public:
any_methods() = default;
template<class T>
any_methods( tag_t<T> ): vtable(get_vtable(tag<T>)) {}
any_methods& operator=(any_methods const&)=default;
template<class T>
void change_type( tag_t<T> ={} ) { vtable = get_vtable(tag<T>); }
template<class any_method>
auto get_invoker( tag_t<any_method> ={} ) const {
return std::get<typename any_method_function<any_method>::type>( *vtable );
}
};
we could specialize this for a cases where the vtable is small (for example, 1 item), and use direct pointers stored in-class in those cases for efficiency.
Now we start the super_any. I use super_any_t to make the declaration of super_any a bit easier.
template<class...methods>
struct super_any_t;
This searches the methods that the super any supports for SFINAE:
template<class super_any, class method>
struct super_method_applies : std::false_type {};
template<class M0, class...Methods, class method>
struct super_method_applies<super_any_t<M0, Methods...>, method> :
std::integral_constant<bool, std::is_same<M0, method>{} || super_method_applies<super_any_t<Methods...>, method>{}>
{};
This is the pseudo-method pointer, like print, that we create globally and constly.
We store the object we construct this with inside the any_method. Note that if you construct it with a non-lambda things can get hairy, as the type of this any_method is used as part of the dispatch mechanism.
template<class Sig, class F>
struct any_method {
using signature=Sig;
private:
F f;
public:
template<class Any,
// SFINAE testing that one of the Anys's matches this type:
std::enable_if_t< super_method_applies< std::decay_t<Any>, any_method >{}, int>* =nullptr
>
friend auto operator->*( Any&& self, any_method const& m ) {
// we don't use the value of the any_method, because each any_method has
// a unique type (!) and we check that one of the auto*'s in the super_any
// already has a pointer to us. We then dispatch to the corresponding
// any_method_data...
return [&self, invoke = self.get_invoker(tag<any_method>), m](auto&&...args)->decltype(auto)
{
return invoke( decltype(self)(self), &m, decltype(args)(args)... );
};
}
any_method( F fin ):f(std::move(fin)) {}
template<class...Args>
decltype(auto) operator()(Args&&...args)const {
return f(std::forward<Args>(args)...);
}
};
A factory method, not needed in C++17 I believe:
template<class Sig, class F>
any_method<Sig, std::decay_t<F>>
make_any_method( F&& f ) {
return {std::forward<F>(f)};
}
This is the augmented any. It is both an any, and it carries around a bundle of type-erasure function pointers that change whenever the contained any does:
template<class... methods>
struct super_any_t:boost::any, any_methods<methods...> {
private:
template<class T>
T* get() { return boost::any_cast<T*>(this); }
public:
template<class T,
std::enable_if_t< !std::is_same<std::decay_t<T>, super_any_t>{}, int>* =nullptr
>
super_any_t( T&& t ):
boost::any( std::forward<T>(t) )
{
using dT=std::decay_t<T>;
this->change_type( tag<dT> );
}
super_any_t()=default;
super_any_t(super_any_t&&)=default;
super_any_t(super_any_t const&)=default;
super_any_t& operator=(super_any_t&&)=default;
super_any_t& operator=(super_any_t const&)=default;
template<class T,
std::enable_if_t< !std::is_same<std::decay_t<T>, super_any_t>{}, int>* =nullptr
>
super_any_t& operator=( T&& t ) {
((boost::any&)*this) = std::forward<T>(t);
using dT=std::decay_t<T>;
this->change_type( tag<dT> );
return *this;
}
};
Because we store the any_methods as const objects, this makes making a super_any a bit easier:
template<class...Ts>
using super_any = super_any_t< std::remove_const_t<std::remove_reference_t<Ts>>... >;
Test code:
const auto print = make_any_method<void(std::ostream&)>([](auto&& p, std::ostream& t){ t << p << "\n"; });
const auto wprint = make_any_method<void(std::wostream&)>([](auto&& p, std::wostream& os ){ os << p << L"\n"; });
const auto wont_work = make_any_method<void(std::ostream&)>([](auto&& p, std::ostream& t){ t << p << "\n"; });
struct X {};
int main()
{
super_any<decltype(print), decltype(wprint)> a = 7;
super_any<decltype(print), decltype(wprint)> a2 = 7;
(a->*print)(std::cout);
(a->*wprint)(std::wcout);
// (a->*wont_work)(std::cout);
double d = 4.2;
a = d;
(a->*print)(std::cout);
(a->*wprint)(std::wcout);
(a2->*print)(std::cout);
(a2->*wprint)(std::wcout);
// a = X{}; // generates an error if you try to store a non-printable
}
live example.
The error message when I try to store a non-printable struct X{}; inside the super_any seems reasonable at least on clang:
main.cpp:150:87: error: invalid operands to binary expression ('std::ostream' (aka 'basic_ostream<char>') and 'X')
const auto x0 = make_any_method<void(std::ostream&)>([](auto&& p, std::ostream& t){ t << p << "\n"; });
this happens the moment you try to assign the X{} into the super_any<decltype(x0)>.
The structure of the any_method is sufficiently compatible with the pseudo_method that acts similarly on variants that they can probably be merged.
I used a manual vtable here to keep the type erasure overhead to 1 pointer per super_any. This adds a redirection cost to every any_method call. We could store the pointers directly in the super_any very easily, and it wouldn't be hard to make that a parameter to super_any. In any case, in the 1 erased method case, we should just store it directly.
Two different any_methods of the same type (say, both containing a function pointer) spawn the same kind of super_any. This causes problems at lookup.
Distinguishing between them is a bit tricky. If we changed the super_any to take auto* any_method, we could bundle all of the identical-type any_methods up in the vtable tuple, then do a linear search for a matching pointer if there are more than 1. The linear search should be optimized away by the compiler unless you are doing something crazy like passing a reference or pointer to which particular any_method we are using.
That seems beyond the scope of this answer, however; the existence of that improvement is enough for now.
In addition, a ->* that takes a pointer (or even reference!) on the left hand side can be added, letting it detect this and pass that to the lambda as well. This can make it truly an "any method" in that it works on variants, super_anys, and pointers with that method.
With a bit of if constexpr work, the lambda can branch on doing an ADL or a method call in every case.
This should give us:
(7->*print)(std::cout);
((super_any<&print>)(7)->*print)(std::cout); // C++17 version of above syntax
((std::variant<int, double>{7})->*print)(std::cout);
int* ptr = new int(7);
(ptr->*print)(std::cout);
(std::make_unique<int>(7)->*print)(std::cout);
(std::make_shared<int>(7)->*print)(std::cout);
with the any_method just "doing the right thing" (which is feeding the value to std::cout <<).

Here's my solution. It looks shorter than Yakk's, and it does not use std::aligned_storage and placement new. It additionally supports stateful and local functors (which implies that it might never be possible to write super_any<&print>, since print could be a local variable).
any_method:
template<class F, class Sig> struct any_method;
template<class F, class Ret, class... Args> struct any_method<F,Ret(Args...)> {
F f;
template<class T>
static Ret invoker(any_method& self, boost::any& data, Args... args) {
return self.f(boost::any_cast<T&>(data), std::forward<Args>(args)...);
}
using invoker_type = Ret (any_method&, boost::any&, Args...);
};
make_any_method:
template<class Sig, class F>
any_method<std::decay_t<F>,Sig> make_any_method(F&& f) {
return { std::forward<F>(f) };
}
super_any:
template<class...OperationsToTypeErase>
struct super_any {
boost::any data;
std::tuple<typename OperationsToTypeErase::invoker_type*...> operations = {};
template<class T, class ContainedType = std::decay_t<T>>
super_any(T&& t)
: data(std::forward<T>(t))
, operations((OperationsToTypeErase::template invoker<ContainedType>)...)
{}
template<class T, class ContainedType = std::decay_t<T>>
super_any& operator=(T&& t) {
data = std::forward<T>(t);
operations = { (OperationsToTypeErase::template invoker<ContainedType>)... };
return *this;
}
};
operator->*:
template<class...Ops, class F, class Sig,
// SFINAE filter that an op matches:
std::enable_if_t< std::disjunction< std::is_same<Ops, any_method<F,Sig>>... >{}, int> = 0
>
auto operator->*( super_any<Ops...>& a, any_method<F,Sig> f) {
auto fptr = std::get<typename any_method<F,Sig>::invoker_type*>(a.operations);
return [fptr,f, &a](auto&&... args) mutable {
return fptr(f, a.data, std::forward<decltype(args)>(args)...);
};
}
Usage:
#include <iostream>
auto print = make_any_method<void(std::ostream&)>(
[](auto&& self, auto&& os){ os << self; }
);
using printable_any = super_any<decltype(print)>;
printable_any bob = 7; // sets up the printing data attached to the any
int main() {
(bob->*print)(std::cout); // prints 7
bob = 3.14159;
(bob->*print)(std::cout); // prints 3.14159
}
Live

Related

Access std::vector<std::variant> value by index

I would like to access a member of std::vector<std::variant> by index. Considering the following snippet:
struct Data {
using data_types = std::variant<std::basic_string<char>, double, int>;
public:
template <class T>
void push_back(const T& t) {
m_data.push_back(t);
}
private:
std::vector<data_types> m_data;
};
int main()
{
Data d;
d.push_back(0);
d.push_back("string");
d.push_back(3.55);
}
I would like to access the values like d[0] (should return int) or d[1] (should return std::string).
What I have tried so far but what isn't working is to add the following public method to the existing struct:
template <class T>
T& operator[](const size_t &index) {
return std::visit([](const T& value) {
return static_cast<T>(value);
}, m_data[index]);
}
Any ideas how to achieve the desired result?
The type of an expression in C++ cannot depend on runtime parameters; basically it can only depend on types of the arguments, plus non-type template arguments.
So d[0] and d[1] must have the same type, as the type of the pieces of the expression are identical, and there are no non-type template arguments.
std::get<int>(d[0]) vs std::get<double>(d[1]) can differ in type.
std::get<1>(d[0]) vs std::get<2>(d[1]) can differ in type.
std::visit is a mechanism used to get around this; here, we create every a function object call, one for each possible type, and then pick one at runtime to actually call. However, the type returned from the visit still follows the above rule: it doesn't depend on what type is stored in the variant, and every possible type in the variant must have a valid instantiation of the function.
C++ type system is not a runtime type system. It is compile-time. Stuff like variant and dynamic_cast and any give some runtime exposure to it, but it is intentionally minimal.
If you are wanting to print the contents of a variant, you can do this:
std::visit([](auto& x){
std::cout << x;
}, d[0]);
the trick here is that each of the various types of variant have a lambda function body written for them (so they all must be valid). Then, at run time, the one actually in the variant is run.
You can also test the variant and ask if it has a specific type, either via std::get or manually.
bool has_int = std::visit([](auto& x){
return std::is_same_v<int, std::decay_t<decltype(x)>>::value;
}, d[0]);
this gives you a bool saying if d[0] has an int in it or not.
The next bit is getting insane. Please don't read this unless you fully understand how to use variants and want to know more:
You can even extract out the type index of the variant and pass that around as a run time value:
template<auto I>
using konstant_t = std::integral_constant<decltype(I),I>;
template<auto I>
constexpr konstant_t<I> konstant_v{};
template<auto...Is>
using venum_t = std::variant< konstant_t<Is>... >;
template<class Is>
struct make_venum_helper;
template<class Is>
using make_venum_helper_t = typename make_venum_helper<Is>::type;
template<std::size_t...Is>
struct make_venum_helper<std::index_sequence<Is...>>{
using type=venum_t<Is...>;
};
template<std::size_t N>
using make_venum_t = typename make_venum_helper<std::make_index_sequence<N>>::type;
template<std::size_t...Is>
constexpr auto venum_v( std::index_sequence<Is...>, std::size_t I ) {
using venum = make_venum_t<sizeof...(Is)>;
constexpr venum arr[]={
venum( konstant_v<Is> )...
};
return arr[I];
}
template<std::size_t N>
constexpr auto venum_v( std::size_t I ) {
return venum_v( std::make_index_sequence<N>{}, I );
}
template<class...Ts>
constexpr auto venum_v( std::variant<Ts...> const& v ) {
return venum_v< sizeof...(Ts) >( v.index() );
}
now you can do this:
using venum = make_venum_t<3>;
venum idx = venum_v(d[0]);
and idx holds the index of the engaged type in d[0]. This is only somewhat useful, as you still need std::visit to use it usefully:
std::visit([&](auto I) {
std::cout << std::get<I>( d[0] );
}, idx );
(within the lambda, I is a std::integral_constant, which can be constexpr converted to an integer.)
but lets you do some interesting things with it.
To extract a value from variant, use std::get:
struct Data
{
...
template <class T>
T& operator[](size_t index)
{
return std::get<T>(m_data[index]);
}
};
However, because this overloaded operator is a template, you can't use simple operator syntax to call it. Use the verbose syntax:
int main()
{
Data d;
d.push_back(0);
d.push_back("string");
d.push_back(3.55);
std::cout << d.operator[]<double>(2);
}
Or rename it to use a plain name instead of the fancy operator[].
Visitor pattern:
#include <iostream>
#include <string>
#include <variant>
#include <vector>
template <class ...Ts>
struct MultiVector : std::vector<std::variant<Ts...>> {
template <class Visitor>
void visit(std::size_t i, Visitor&& v) {
std::visit(v, (*this)[i]);
}
};
int main() {
MultiVector<std::string, int, double> vec;
vec.push_back(0);
vec.push_back("string");
vec.push_back(3.55);
vec.visit(2, [](auto& e) { std::cout << e << '\n'; });
}

Transform each of parameter pack's values based on a boolean criteria

I am trying to solve this problem in C++ TMP where in i need to convert one parameter pack types into another, and then convert back the types and also values. The conversion back part is based on a boolean criteria that whether an arg in Args... was transformed or not in the first place.
Basically, i have a pack(Args...). First, i transform this (for each args[i], call a transform function). It works like this:
For each arg in Args..., just create same type in transformed_args... unless it is one of following, in that case do following conversions:
Type In Args...
Type In transformed_Args...
SomeClass
shared_ptr to SomeClass
std::vector of SomeClass
std::vector of shared_ptr to SomeClass
everything else remains the same for ex:
int remains int
std::string remains std::string
I achieve this by template specialization, of course
For the next part, i take transformed_args..., publish a class and a functor. I receive call back on this functor from(C++generated Python using Pybind, not important though). Relevant bits of that class look like this...
template<typename C, typename...transformed_args..., typename... Args>
class SomeTemplateClass
{
MethodWrapper<C,void, Args...> func;
//.....
void operator()(transformed_args... targs)
{
//....
(*func.wrapped_method_inside)(transform_back_magic(targs)...) // this is want i want to achieve.
//transform_back_magic(targs)... is a plaeholder for code that checks if type of args[i]... != type of targs[i]... and then calls a tranform_back specialization on it else just return args[i].val
}
}
targs are in transformed_args... format, but underlying C++ function they are aimed for expects Args...
template<typename... Args, typename... transformed_args, ........whatever else is needed>
transform_back_magic(....)
{
if(Args[i].type != transformed_args[i].types)
tranform_back(targs[i]...);
}
the tranform_back function template logic is specialized for different cases and all logic is in place. But how to invoke that based on this boolean criteria is hitting my TMP knowledge limits. I just got started not many weeks ago.
Here i am listing down what i have created so far.
First of all this is what i need in pseudo code
template<typename C, typename... transformed_args, typename... Args>
class SomeTemplateClass
{
MethodWrapper<C,void, Args...> func;
void operator(transformed_args... targs)
{
**//In pseudo code, this is what i need**
Args... params = CreateArgsInstanceFromTransformedArgs(targs);
(*func.wrapped_method_inside)(params...);
}
}
In my attempt to implement this, so far I have decided on creating a tuple<Args...> object by copying data from targs(with conversions where ever required)
void operator(transformed_args... targs)
{
//....
auto mytup = call1(std::tuple<args...>(), std::make_index_sequence<sizeof...(Args)>,
std::make_tuple(targs...), targs...);
// mytup can be std::tuple<Args...>(transform_back(1st_targs), transform_back(2nd_targs)....). Once available i can write some more logic to extract Args... from this tuple and pass to(*func.wrapped_method_inside)(....)
(*func.wrapped_method_inside)(ArgsExtractorFromTuple(mytup)); // this part is not implemented yet, but i think it should be possible. This is not my primary concern at the moment
}
//call1
template<typename... Args, typename... Targs, std::size_t... N>
auto call1(std::tuple<Args...> tupA, std::index_sequence<N>..., std::tuple<Targs...> tupT, Targs ..)
{
auto booltup = tuple_creator<0>(tupA, tupT, nullptr); // to create a tuple of bools
auto ret1 = std::make_tuple<Args...>(call2(booltup, targs, N)...); // targs and N are expanded together so that i get indirect access to see the corresponding type in Args...
return ret1;
}
// tuple_creator is a recursive function template with sole purpose to create a boolean tuple.
// such that std::get<0>(booltup) = true,
//if tuple_element_t<0,std::tuple<Args...>> and tuple_element_t<0,std::tuple<targs...>> are same types else false
template<size_t I, typename... Targs, typename... Args>
auto tuple_creator(std::tuple<Args...>tupA, std::tuple<Targs...>tupT, std::enable_if_t<I == sizeof...(targs)>*)
{
return std::make_tuple(std::is_same<std::tuple_element_t<I-1, std::tuple<Targs...>>, std::tuple_element_t<I-1, std::tuple<Args...>>>::value);
}
template<size_t I = 0, typename... Targs, typename... Args>
auto tuple_creator(std::tuple<Args...>tupA, std::tuple<Targs...>tupT, std::enable_if_t<I < sizeof...(targs)>*)
{
auto ret1 = tuple_creator<I+1>(tupA, tupT, nullptr);
if(!I)
return ret1;
auto ret2 = std::is_same<std::tuple_element_t<I-1, std::tuple<Targs...>>, std::tuple_element_t<I-1, std::tuple<Args...>>>::value;
return std::tuple_cat(ret1, std::make_tuple(ret2));
}
template<typename TT, typename Tuple>
auto call2(Tuple boolyup, TT t, std::size_t I)
{
auto ret = transform_back<std::get<I>(booltup)>(t); // error: I is not a compile time constant
return ret;
}
transform_back is a template that uses a bool template param and enable_if based specialization to decide whether transform an argument back or not
below are the transform_back specialization for std::vector. Similarly i have others for when T = Class etc and so on
template<bool sameTypes, typename T>
std::enable_if_t<(is_vector<T>::value, is_shared_ptr<typename T::value_type>::value &&
is_class<remove_cvref_t<typename T::value_type_element_type>>::value
&& sameTypes), T>
transform_back(T val) // it was never transfoemd in first place, return as is
{
return val;
}
template<bool sameTypes, typename T>
std::enable_if_t<(is_vector<T>::value, is_shared_ptr<typename T::value_type>::value
&& is_class<remove_cvref_t<typename T::value_type_element_type>>::value
&& !sameTypes),
typename std::vector<typename T::value_type::element_type>>
transform(T val)
{
std::vector<T::value_type::element_type> t;
for(int i = 0 ; i < val.size(); ++i)
{
typename T::value_type::element_type obj = *val[i];
t.push_back(obj);
}
return t;
}
Both these specialization are same and only differ on sameTypes boolean variable
This code currently errors out in call2 method while trying to using
std::get
auto ret = transform_back<std::get<I>(booltup)>(t); // error: I is not a compile time constant
How can you help?
1)What could be the work around to std::get issue here? Just cant figure out a way to fit in std::size_t as template arg here instead of function arg to make it work at compile time.
Other than this:
2)If you can suggest an alternative approach to implement from top level.
Args... params = CreateArgsInstanceFromTransformedArgs(targs);
That would be great. The path i took is not very convincing personally to me.
If I understand correctly, you might do something like:
template <typename> struct Tag{};
std::shared_ptr<SomeClass> transform_to(Tag<std::shared_ptr<SomeClass>>, const SomeClass& s)
{
return std::make_shared<SomeClass>(s);
}
std::vector<std::shared_ptr<SomeClass>> transform_to(Tag<std::vector<std::shared_ptr<SomeClass>>>, const std::vector<SomeClass>& v)
{
std::vector<std::shared_ptr<SomeClass>> res;
res.reserve(v.size());
for (const auto& s : v) {
res.emplace_back(std::make_shared<SomeClass>(s));
}
return res;
}
const SomeClass& transform_to(Tag<SomeClass>, const std::shared_ptr<SomeClass>& s)
{
return *s;
}
std::vector<SomeClass> transform_to(Tag<std::vector<SomeClass>>, const std::vector<std::shared_ptr<SomeClass>>& v)
{
std::vector<SomeClass> res;
res.reserve(v.size());
for (const auto& s : v) {
res.emplace_back(*s);
}
return res;
}
template <typename T>
const T& transform_to(Tag<T>, const T& t) { return t; } // No transformations
And then
std::function<void (Args...)> func;
template <typename ... transformed_args>
void operator () (transformed_args... targs) const
{
func(transform_to(Tag<Args>(), targs)...);
}
Just explaining the use case here to add some context. Consider these three methods in C++ each represented with the function pointer SomeTemplateClass::func:
void foo(vector<shared_ptr<SomeClass>>) // 1
// Args... = vector<shared_ptr<SomeClass>>, Targs... = vector<shared_ptr<SomeClass>>
void foo(vector<SomeClass>) // 2
// Args... = vector<SomeClass>, Targs... = vector<shared_ptr<SomeClass>>
void foo(vector<SomeClass>, vector<shared_ptr<SomeClass>>) // 3
// Args... = vector<SomeClass>, vector<shared_ptr<SomeClass>>, Targs... = vector<shared_ptr<SomeClass>>, vector<shared_ptr<SomeClass>>
One instance each of SomeTemplateClass is exposed to Python via Pybind. I do these transformations so that when foo is called from Python, any arg vector<T>(in C++) is received as vector<shared_ptr<T>> in SomeTemplateClass functor. This helps in to get handle to previously created objects T that i need.
But as you can see from 3 cases for foo, foo(vector<shared_ptr<T>>) does not need to be transformed to and subsequently not need to be transformed back. The case of 'tranform_to'is easily handled with template specialization, but while transforming back, vector<shared_ptr<T>> cant be blindly converted back to vector<T>. So (transform(targs...)) needs an additional logic to transform a particular arg (or targ) only when targ[i]::type != arg[i]::type
Building on Jarod's answer, i rather need something like this where in transform_to method for vector<shared_ptr> is further divided in two possible templates
template<bool wasOriginallyTransformed>
enable_if<!wasOriginallyTransformed, std::vector<std::shared_ptr<SomeClass>> transform_to(Tag<std::vector<SomeClass>>, const std::vector<std::shared_ptr<SomeClass>>& v)
{
return v;
}
template<bool wasOriginallyTransformed>
enable_if<!wasOriginallyTransformed, std::vector<<SomeClass>
transform_to(Tag<std::vector<SomeClass>>, const std::vector<std::shared_ptr<SomeClass>>& v)
{
std::vector<SomeClass> res;
res.reserve(v.size());
for (const auto& s : v) {
res.emplace_back(*s);
}
return res;
}

Function object conversion to function pointer

I am looking for a way to convert function object to function pointer.
Captureless lambda has implicit conversion that allows to:
using fptr_t = int (*)(int);
fptr_t ptr = nullptr;
ptr = [](int) { return 2; };
ptr = [](auto) { return 3; };
(*ptr)(42);
I try to do the same with old-fashioned, empty class function objects like:
struct Foo {
int operator()(int) const { return 5; }
} foo;
Or std predicates like std::less<int>.
One way I found is to wrap call of foo with lambda.
If I can assure that foo is stateless and const
I dont really need this ptr and lambda-capture:
template <typename R, typename... Args>
struct to_function_pointer<R(Args...)> {
private:
template <typename T, REQUIRES(std::is_empty<T>::value)>
static T const& stateless_const() {
return (*static_cast<T const*>(nullptr));
}
public:
using pointer = R (*)(Args...);
template <typename U>
pointer operator()(U) const {
return [](Args... args) {
return stateless_const<std::decay_t<U>>()(args...);
};
}
};
But here I do not know how to provide perfect forwarding,
cause [](Args&&...) or [](auto&&...) cannot convert to R(*)(Args...).
Such trick fails when args is noncopyable like std::unique_ptr<int>.
I know that I could use std::function, but it's kind of heavy-weight, while I am trying to get a light-weight solution.
Live example.
Any advice appreciated.
I believe you can simplify your to_function_pointer with just:
template <typename R, typename... Args>
struct to_function_pointer<R(Args...)> {
using pointer = R(*)(Args...);
template <typename U, REQUIRES(std::is_empty<U>::value && std::is_trivially_constructible<U>::value)>
pointer operator()(U ) const
{
return [](Args... args) {
return U{}(std::forward<Args>(args)...);
}
}
};
Few things to note. Args... will already be references or not, you're providing that signature. So forward<> will still do the same thing - the function just happens to not be a template here. Also, scrap the weird nullptr cast. That just looks bad - if we just require trivial constructibility we can just write U{} which seems way cleaner to me.

How can implement dynamic function call with C++11 and C++14?

Here is code that I hope explains what I want to achieve.
vector<int> ints;
vector<double> doubles;
struct Arg {
enum Type {
Int,
Double
};
Type type;
int index;
};
template <typename F>
void Call(const F& f, const vector<Arg>& args) {
// TODO:
// - First assert that count and types or arguments of <f> agree with <args>.
// - Call "f(args)"
}
// Example:
void copy(int a, double& b) {
b = a;
}
int test() {
Call(copy, {{Int, 3}, {Double, 2}}); // copy(ints[3], double[2]);
}
Can this be done in C++11 ?
If yes, can the solution be simplified in C++14 ?
I'd do this in two steps.
First, I'd wrap f in an object able to understand Arg-like parameters, and generate errors on failure. For simplicity, suppose we throw.
This is a bit simpler than your Arg to be understood at this layer, so I might translate Arg into MyArg:
struct MyArg {
MyArg(MyArg const&)=default;
MyArg(int* p):i(p){}
MyArg(double* p):d(p){}
MyArg(Arg a):MyArg(
(a.type==Arg::Int)?
MyArg(&ints.at(a.index)):
MyArg(&doubles.at(a.index))
) {}
int * i = nullptr;
double* d = nullptr;
operator int&(){ if (!i) throw std::invalid_argument(""); return *i; }
operator double&(){ if (!d) throw std::invalid_argument(""); return *d; }
};
We map void(*)(Ts...) to std::function<void(MyArg, MyArg, MyArg)> like this:
template<class T0, class T1>using second_type = T1;
template<class...Ts>
std::function<void( second_type<Ts,MyArg>... )> // auto in C++14
my_wrap( void(*f)(Ts...) ) {
return [f](second_type<Ts,MyArg>...args){
f(args...);
};
}
now all that is left is counting function parameter count vs vector size count, and unpacking the std::vector into a function call.
The last looks like:
template<class...Ts, size_t...Is>
void call( std::function<void(Ts...)> f, std::index_sequence<Is...>, std::vector<Arg> const& v ) {
f( v[Is]... );
}
template<class...Ts>
void call( std::function<void(Ts...)> f, std::vector<Arg> const& v ) {
call( std::move(f), std::index_sequence_for<Ts...>{}, v );
}
where index_sequence and index_sequence_for are C++14, but equivalents can be implemented in C++11 (there are many implementations on stack overflow).
So we end up with something like:
template<class...Ts>
void Call( void(*pf)(Ts...), std::vector<Arg> const& v ) {
if (sizeof...(Ts)>v.size())
throw std::invalid_argument("");
auto f = my_wrap(pf);
call( std::move(f), v );
}
Dealing with the throws is left as an exercise, as is handling return values.
This code has not been compiled or tested, but the design should be sound. It only supports calling function pointers -- calling generalized callable objects is tricky, because counting how many arguments they want (of type int or double) is tricky. If you passed in how many arguments they want as a compile-time constant, it is easy. You could also build a magic switch that handles counts up to some constant (10, 20, 1000, whatever), and dispatch the runtime length of the vector into a compile time constant that throws on a argument length mismatch.
This is trickier.
The hard coded pointers sort of suck.
template<class...Ts>struct types{using type=types;};
template<size_t I> using index=std::integral_constant<size_t, I>;
template<class T, class types> struct index_in;
template<class T, class...Ts>
struct index_in<T, types<T,Ts...>>:
index<0>
{};
template<class T, class T0, class...Ts>
struct index_in<T, types<T0,Ts...>>:
index<1+index_in<T, types<Ts...>>{}>
{};
is a package of types.
Here is how we can store buffers:
template<class types>
struct buffers;
template<class...Ts>
struct buffers<types<Ts...>> {
struct raw_view {
void* start = 0;
size_t length = 0;
};
template<class T>
struct view {
T* start = 0;
T* finish = 0;
view(T* s, T* f):start(s), finish(f) {}
size_t size() const { return finish-start; }
T& operator[](size_t i)const{
if (i > size()) throw std::invalid_argument("");
return start[i];
}
}
std::array< raw_view, sizeof...(Ts) > views;
template<size_t I>
using T = std::tuple_element_t< std::tuple<Ts...>, I >;
template<class T>
using I = index_of<T, types<Ts...> >;
template<size_t I>
view<T<I>> get_view() const {
raw_view raw = views[I];
if (raw.length==0) { return {0,0}; }
return { static_cast<T<I>*>(raw.start), raw.length/sizeof(T) };
}
template<class T>
view<T> get_view() const {
return get_view< I<T>{} >();
}
template<class T>
void set_view( view<T> v ) {
raw_view raw{ v.start, v.finish-v.start };
buffers[ I<T>{} ] = raw;
}
};
now we modify Call:
template<class R, class...Args, size_t...Is, class types>
R internal_call( R(*f)(Args...), std::vector<size_t> const& indexes, buffers<types> const& views, std::index_sequence<Is...> ) {
if (sizeof...(Args) != indexes.size()) throw std::invalid_argument("");
return f( views.get_view<Args>()[indexes[Is]]... );
}
template<class R, class...Args, size_t...Is, class types>
R Call( R(*f)(Args...), std::vector<size_t> const& indexes, buffers<types> const& views ) {
return internal_call( f, indexes, views, std::index_sequence_for<Args...>{} );
}
which is C++14, but most components can be translated to C++11.
This uses O(1) array lookups, no maps. You are responsible for populating buffers<types> with the buffers, sort of like this:
buffers<types<double, int>> bufs;
std::vector<double> d = {1.0, 3.14};
std::vector<int> i = {1,2,3};
bufs.set_view<int>( { i.data(), i.data()+i.size() } );
bufs.set_view<double>( { d.data(), d.data()+d.size() } );
parameter mismatch counts and index out of range generate thrown errors. It only works with raw function pointers -- making it work with anything with a fixed (non-template) signature is easy (like a std::function).
Making it work with an object with no signature is harder. Basically instead of relying on the function called for the arguments, you instead build the cross product of the types<Ts...> up to some fixed size. You build a (large) table of which of these are valid calls to the passed in call target (at compile time), then at run time walk that table and determine if the arguments passed in are valid to call the object with.
It gets messy.
This is why my above version simply asks for indexes, and deduces the types from the object being called.
I have a partial solution, using C++11 grammar.
First I make a function overloader accepting arbitrator kinds of arguments
template< typename Function >
struct overloader : Function
{
overloader( Function const& func ) : Function{ func } {}
void operator()(...) const {}
};
template< typename Function >
overloader<Function> make_overloader( Function const& func )
{
return overloader<Function>{ func };
}
then, using the overloader to deceive the compiler into believing the following code ( in switch-case block )is legal:
template <typename F>
void Call( F const& f, const vector<Arg>& args )
{
struct converter
{
Arg const& arg;
operator double&() const
{
assert( arg.type == Double );
return doubles[arg.index];
}
operator int() const
{
assert( arg.type == Int );
return ints[arg.index];
}
converter( Arg const& arg_ ): arg( arg_ ) {}
};
auto function_overloader = make_overloader( f );
unsigned long const arg_length = args.size();
switch (arg_length)
{
case 0 :
function_overloader();
break;
case 1 :
function_overloader( converter{args[0]} );
break;
case 2 :
function_overloader( converter{args[0]}, converter{args[1]} );
break;
case 3 :
function_overloader( converter{args[0]}, converter{args[1]}, converter{args[2]} );
break;
/*
case 4 :
.
.
.
case 127 :
*/
}
}
and test it this way:
void test_1()
{
Call( []( int a, double& b ){ b = a; }, vector<Arg>{ Arg{Int, 3}, Arg{Double, 2} } );
}
void test_2()
{
Call( []( double& b ){ b = 3.14; }, vector<Arg>{ Arg{Double, 0} } );
}
void my_copy( int a, double& b, double& c )
{
b = a;
c = a+a;
}
void test_3()
{
//Call( my_copy, vector<Arg>{ Arg{Int, 4}, Arg{Double, 3}, Arg{Double, 1} } ); // -- this one does not work
Call( []( int a, double& b, double& c ){ my_copy(a, b, c); }, vector<Arg>{ Arg{Int, 4}, Arg{Double, 3}, Arg{Double, 1} } );
}
the problems with this solution is:
g++5.2 accept it, clang++6.1 doesn's
when the argument(s) of function Call is/are not legal, it remains silent
the first argument of function Call cannot be a C-style function, one must wrap that into a lambda object to make it work.
the code is available here - http://melpon.org/wandbox/permlink/CHZxVfLM92h1LACf -- for you to play with.
First of all, you need some mechanism to register your argument values that are later referenced by some type and an index:
class argument_registry
{
public:
// register a range of arguments of type T
template <class T, class Iterator>
void register_range(Iterator begin, Iterator end)
{
// enclose the range in a argument_range object and put it in our map
m_registry.emplace(typeid(T), std::make_unique<argument_range<T, Iterator>>(begin, end));
}
template <class T>
const T& get_argument(size_t idx) const
{
// check if we have a registered range for this type
auto itr = m_registry.find(typeid(T));
if (itr == m_registry.end())
{
throw std::invalid_argument("no arguments registered for this type");
}
// we are certain about the type, so downcast the argument_range object and query the argument
auto range = static_cast<const argument_range_base1<T>*>(itr->second.get());
return range->get(idx);
}
private:
// base class so we can delete the range objects properly
struct argument_range_base0
{
virtual ~argument_range_base0(){};
};
// interface for querying arguments
template <class T>
struct argument_range_base1 : argument_range_base0
{
virtual const T& get(size_t idx) const = 0;
};
// implements get by querying a registered range of arguments
template <class T, class Iterator>
struct argument_range : argument_range_base1<T>
{
argument_range(Iterator begin, Iterator end)
: m_begin{ begin }, m_count{ size_t(std::distance(begin, end)) } {}
const T& get(size_t idx) const override
{
if (idx >= m_count)
throw std::invalid_argument("argument index out of bounds");
auto it = m_begin;
std::advance(it, idx);
return *it;
}
Iterator m_begin;
size_t m_count;
};
std::map<std::type_index, std::unique_ptr<argument_range_base0>> m_registry;
};
Than we define a small type to combine a type and a numerical index for referencing arguments:
typedef std::pair<std::type_index, size_t> argument_index;
// helper function for creating an argument_index
template <class T>
argument_index arg(size_t idx)
{
return{ typeid(T), idx };
}
Finally, we need some template recursion to go through all expected arguments of a function, check if the user passed an argument of matching type and query it from the registry:
// helper trait for call function; called when there are unhandled arguments left
template <bool Done>
struct call_helper
{
template <class FuncRet, class ArgTuple, size_t N, class F, class... ExpandedArgs>
static FuncRet call(F func, const argument_registry& registry, const std::vector<argument_index>& args, ExpandedArgs&&... expanded_args)
{
// check if there are any arguments left in the passed vector
if (N == args.size())
{
throw std::invalid_argument("not enough arguments");
}
// get the type of the Nth argument
typedef typename std::tuple_element<N, ArgTuple>::type arg_type;
// check if the type matches the argument_index from our vector
if (std::type_index{ typeid(arg_type) } != args[N].first)
{
throw std::invalid_argument("argument of wrong type");
}
// query the argument from the registry
auto& arg = registry.get_argument<arg_type>(args[N].second);
// add the argument to the ExpandedArgs pack and continue the recursion with the next argument N + 1
return call_helper<std::tuple_size<ArgTuple>::value == N + 1>::template call<FuncRet, ArgTuple, N + 1>(func, registry, args, std::forward<ExpandedArgs>(expanded_args)..., arg);
}
};
// helper trait for call function; called when there are no arguments left
template <>
struct call_helper<true>
{
template <class FuncRet, class ArgTuple, size_t N, class F, class... ExpandedArgs>
static FuncRet call(F func, const argument_registry&, const std::vector<argument_index>& args, ExpandedArgs&&... expanded_args)
{
if (N != args.size())
{
// unexpected arguments in the vector
throw std::invalid_argument("too many arguments");
}
// call the function with all the expanded arguments
return func(std::forward<ExpandedArgs>(expanded_args)...);
}
};
// call function can only work on "real", plain functions
// as you could never do dynamic overload resolution in C++
template <class Ret, class... Args>
Ret call(Ret(*func)(Args...), const argument_registry& registry, const std::vector<argument_index>& args)
{
// put the argument types into a tuple for easier handling
typedef std::tuple<Args...> arg_tuple;
// start the call_helper recursion
return call_helper<sizeof...(Args) == 0>::template call<Ret, arg_tuple, 0>(func, registry, args);
}
Now you can use it like this:
int foo(int i, const double& d, const char* str)
{
printf("called foo with %d, %f, %s", i, d, str);
// return something
return 0;
}
int main()
{
// prepare some arguments
std::vector<int> ints = { 1, 2, 3 };
std::vector<double> doubles = { 10., 20., 30. };
std::vector<const char*> str = { "alpha", "bravo", "charlie" };
// register them
argument_registry registry;
registry.register_range<int>(ints.begin(), ints.end());
registry.register_range<double>(doubles.begin(), doubles.end());
registry.register_range<const char*>(str.begin(), str.end());
// call function foo with arguments from the registry
return call(foo, registry, {arg<int>(2), arg<double>(0), arg<const char*>(1)});
}
Live example: http://coliru.stacked-crooked.com/a/7350319f88d86c53
This design should be open for any argument type without the need to list all the supported types somewhere.
As noted in the code comments, you cannot call any callable object like this in general, because overload resolution could never be done at runtime in C++.
Instead of clarifying the question, as I requested, you have put it up for bounty. Except if that really is the question, i.e. a homework assignment with no use case, just exercising you on general basic programming, except for that only sheer luck will then give you an answer to your real question: people have to guess about what the problem to be solved, is. That's the reason why nobody's bothered, even with the bounty, to present a solution to the when-obvious-errors-are-corrected exceedingly trivial question that you literally pose, namely how to do exactly this:
vector<int> ints;
vector<double> doubles;
struct Arg {
enum Type {
Int,
Double
};
Type type;
int index;
};
template <typename F>
void Call(const F& f, const vector<Arg>& args) {
// TODO:
// - First assert that count and types or arguments of <f> agree with <args>.
// - Call "f(args)"
}
// Example:
void copy(int a, double& b) {
b = a;
}
int test() {
Call(copy, {{Int, 3}, {Double, 2}}); // copy(ints[3], double[2]);
}
In C++11 and later one very direct way is this:
#include <assert.h>
#include <vector>
using std::vector;
namespace g {
vector<int> ints;
vector<double> doubles;
}
struct Arg {
enum Type {
Int,
Double
};
Type type;
int index;
};
template <typename F>
void Call(const F& f, const vector<Arg>& args)
{
// Was TODO:
// - First assert that count and types or arguments of <f> agree with <args>.
assert( args.size() == 2 );
assert( args[0].type == Arg::Int );
assert( int( g::ints.size() ) > args[0].index );
assert( args[1].type == Arg::Double );
assert( int( g::doubles.size() ) > args[1].index );
// - Call "f(args)"
f( g::ints[args[0].index], g::doubles[args[1].index] );
}
// Example:
void copy(int a, double& b)
{
b = a;
}
auto test()
{
Call(copy, {{Arg::Int, 3}, {Arg::Double, 2}}); // copy(ints[3], double[2]);
}
namespace h {}
auto main()
-> int
{
g::ints = {000, 100, 200, 300};
g::doubles = {1.62, 2.72, 3.14};
test();
assert( g::doubles[2] == 300 );
}
There are no particularly relevant new features in C++14.
I propose this answer following my comment on your question. Seeing that in the requirements, you stated:
Preferably we should not be required to create a struct that
enumerates all the types we want to support.
It could suggests you would like to get rid of the type enumerator in your Arg structure. Then, only the value would be left: then why not using plain C++ types directly, instead of wrapping them ?
It assumes you then know all your argument types at compile time
(This assumption could be very wrong, but I did not see any requirement in your question preventing it. I would be glad to rewrite my answer if you give more details).
The C++11 variadic template solution
Now to the solution, using C++11 variadic templates and perfect forwarding. In a file Call.h:
template <class F, class... T_Args>
void Call(F f, T_Args &&... args)
{
f(std::forward<T_Args>(args)...);
}
Solution properties
This approach seems to satisfy all your explicit requirements:
Works with C++11 standard
Checks that count and types or arguments of f agress with args.
It actually does that early, at compile time, instead of a possible runtime approach.
No need to manually enumerate the accepted types (actually works with any C++ type, be it native or user defined)
Not in your requirement, but nice to have:
Very compact, because it leverage a native features introduced in C++11.
Accepts any number of arguments
The type of the argument and the type of the corresponding f parameter do not have to match exactly, but have to be compatible (exactly like a plain C++ function call).
Example usage
You could test it in a simple main.cpp file:
#include "Call.h"
#include <iostream>
void copy(int a, double& b)
{
b = a;
}
void main()
{
int a = 5;
double b = 6.2;
std::cout << "b before: " << b << std::endl;
Call(copy, a, b);
std::cout << "b now: " << b << std::endl;
}
Which would print:
b before: 6.2
b now: 5

C++ meta-programming: overloading of arithmetic operators for types

Suppose the following class calculates an arithmetic composition of two functions (represented by functors Function1 and Function2):
template<typename Function1, typename Function2, typename BinaryOp> Compose
{
// schematically:
// int operator() = Function1().operator() BinaryOp Function2.operator();
};
With this one could write, for instance,
typedef Compose<F,G,Division> F_div_G;
to obtain a new functor implementing F/G. This could be used, for instance, like this:
F f;
G g;
F_div_G f_div_g;
if( f(3)/g(3) == f_div_g(3) ) std::cout<<"... it works ..."<<std::endl;
However, by doing several compositions in this way, the code starts to become more and more unreadable.
Thus, is it possible to somehow "overload" the operator / (and similarly other arithmetic operators) to be able to replace Compose<F,G,Division> by F/G? Do you know a macro or something similar which safely implements this? Or, provided that it is possible, would you discourage from doing this?
template<typename F>
struct Op {
// this is optional:
template<typename... Args>
auto operator()(Args&&...args)const{
return F{}(std::forward<Args>(args)...);
}
};
// `Op<Compose>` lets you chain these things:
template<typename LHS, typename RHS, typename Operation>
struct Compose:Op<Compose> {
template<typename... Args>
auto operator()(Args&&...args)const{
// with lots of care, perfect forwarding could be shoe horned in here, but it gets tricky:
return Operation{}(LHS{}(args...),RHS{}(args...));
}
};
// write one of these for each operator you want to handle:
struct Divide {
template<typename LHS, typename RHS>
auto operator()(LHS&& lhs,RHS&&rhs)const{
return std::forward<LHS>(lhs)()/std::forward<RHS>(rhs);
}
};
// write one of these as well for each operator you want to handle:
template<typename LHS,typename RHS>
Compose<LHS,RHS,Divide> operator/( Op<LHS> lhs, Op<RHS> rhs ) {
return {};
}
and bob is your uncle.
int main() {
auto f_div_g = Op<F>{}/Op<G>{};
auto f_div_g_div_h = Op<F>{}/Op<G>{}/Op<H>{};
}
and if you want a type:
decltype( Op<F>{}/Op<G>{} )
with maybe a std::decay thrown in.