When a compiler can infer a template parameter? - c++

Sometimes it works sometimes not:
template <class T>
void f(T t) {}
template <class T>
class MyClass {
public:
MyClass(T t) {}
};
void test () {
f<int>(5);
MyClass<int> mc(5);
f(5);
MyClass mc(5); // this doesn't work
}
Is there a way to hack around the example above? I.e. force the compiler to infer the template parameter from constructor parameter.
Will this be fixed in the future, or is there a good reason not to?
What is the general rule when compiler can infer template parameter?

Template parameters can be inferred for function templates when the parameter type can be deduced from the template parameters
So it can be inferred here:
template <typename T>
void f(T t);
template <typename T>
void f(std::vector<T> v);
but not here:
template <typename T>
T f() {
return T();
}
And not in class templates.
So the usual solution to your problem is to create a wrapper function, similar to the standard library function std::make_pair:
template <class T>
class MyClass {
public:
MyClass(T t) {}
void print(){
std::cout<<"try MyClass"<<std::endl;
}
};
template <typename T>
MyClass<T> MakeMyClass(T t) { return MyClass<T>(t); }
and then call auto a = MakeMyClass(5); to instantiate the class.

Read up on Template Argument Deduction (and ADL or Koenig lookup).

Since C++20, it is possible to infer the types of function parameters using auto:
#include <iostream>
#include <string>
auto print_stuff(auto x, auto y)
{
std::cout << x << std::endl;
std::cout << y << std::endl;
}
int main()
{
print_stuff(3,"Hello!");
print_stuff("Hello!",4);
return 0;
}

Related

Code compiles on one machine, but not on another [duplicate]

Is it legal to use an incomplete type in a template if the type is complete when the template is instantiated?
As below
#include <iostream>
struct bar;
template <typename T>
struct foo {
foo(bar* b) : b(b) {
}
void frobnicate() {
b->frobnicate();
}
T val_;
bar* b;
};
struct bar {
void frobnicate() {
std::cout << "foo\n";
}
};
int main() {
bar b;
foo<int> f(&b);
f.frobnicate();
return 0;
}
Visual Studio compiles the above without complaining. GCC issues the warning invalid use of incomplete type 'struct bar' but compiles. Clang errors out with member access into incomplete type 'bar'.
The code is ill-formed, no diagnostic required.
[temp.res.general]/6.4
The validity of a template may be checked prior to any instantiation.
The program is ill-formed, no diagnostic required, if:
...
— a hypothetical instantiation of a template immediately following its definition would be ill-formed due to a construct that does not depend on a template parameter, ...
If you absolutely can't define bar before the template, there is a workaround: you can introduce an artifical dependency on the template parameter.
template <typename T, typename, typename...>
struct dependent_type {using type = T;};
template <typename T, typename P0, typename ...P>
using dependent_type_t = typename dependent_type<T, P0, P...>::type;
Then use dependent_type_t<bar, T> instead of bar.
Clang is correct in reporting an error (as opposed to a warning or being silent about it), though MSVC's and GCC's behavior are also consistent with the standard. See #HolyBlackCat's answer for details on that.
The code you posted is ill-formed NDR. However, what you want to do is feasible.
You can defer the definition of template member functions the same way you would for a non-template class. Much like non-template classes, as long as these definitions requiring bar to be a complete type happen only once bar is complete, everything is fine.
The only hiccup is that you need to explicitly mark the method as inline to avoid ODR violations in multi-TU programs, since the definition will almost certainly be in a header.
#include <iostream>
struct bar;
template <typename T>
struct foo {
foo(bar* b) : b(b) {
}
inline void frobnicate();
T val_;
bar* b;
};
struct bar {
void frobnicate() {
std::cout << "foo\n";
}
};
template <typename T>
void foo<T>::frobnicate() {
b->frobnicate();
}
int main() {
bar b;
foo<int> f(&b);
f.frobnicate();
return 0;
}
If you want to customise a template using a forward declaration as a temlpate argument, you can do this (wihtout warnings or errors):
template <typename T, typename = T>
class print_name;
so when you do a partial specialization, you use the second, unspecialized template parameter for your calls:
struct john;
template <typename T>
class print_name<john, T>
{
public:
void operator()(const T& f) const
{
std::cout << f.name << std::endl;
}
};
In this context T is not incomplete. But when you instantiate print_name<john>, SFINAE will kick.
Here is a full example:
#include <iostream>
template <typename T, typename = T>
class print_name;
struct john;
template <typename T>
class print_name<john, T>
{
public:
void operator()(const T& f) const
{
std::cout << f.name << std::endl;
}
};
struct slim;
template <typename T>
class print_name<slim, T>
{
public:
void operator()(const T& f) const
{
std::cout << f.myName << std::endl;
}
};
#include <string>
struct john
{
std::string name;
};
struct slim
{
std::string myName;
};
int main()
{
print_name<john>{}(john{"John Cena"});
print_name<slim>{}(slim{"Slim Shady"});
return 0;
}
https://godbolt.org/z/czcGo5aaG

Automatic template deduction for function pointers

In the following code
template<class T> void f(T);
int main(){
f(3);
return 0;
}
the template argument int for deduces automatically, as usual.
But in
template<class T> void f(T);
template<class T> void (*p)(T) = f<T>;
int main(){
p(3);
return 0;
}
the compiler (clang++) insists that p(3) needs a template parameter. Why?
Besides, if I put the line template<class T> void (*p)(T) = f<T>; in a header to be included by several files, will that cause problems?
Template argument deduction works with function templates and and with CTAD from C++17. Writing a wrapper is trivial for your example.
template<class T> void f(T);
template<class T> void (*p)(T) = f<T>;
template<typename T> void Wrapper(T&& t)
{
p<T>(std::forward<T>(t));
}
int main(){
Wrapper(3);
}

Specialize member function of class template using C++20 concepts

I want to specialize a member function of a class template as follows:
#include <concepts>
template <typename T>
struct S {
void f();
};
template <typename T>
void S<T>::f() {
}
// (0) This is fine.
template <>
void S<int>::f() {
}
// (1) This triggers an error.
template <std::integral T>
void S<T>::f() {
}
The specialization (0) is fine, but specializes f() only for the int type. Instead, I would like to specialize it, e.g., for any integral type, as in (1). Is this possible using C++20 concepts? Notice that std::integral is just an example and that my specific case makes use of user-defined concepts.
Simply use a trailing requires-clause. The compiler will choose the most constrained function:
#include <concepts>
#include <iostream>
template <typename T>
struct S {
void f();
void f() requires std::integral<T>;
};
template <typename T>
void S<T>::f() {
std::cout << "general\n";
}
template <typename T>
void S<T>::f() requires std::integral<T> {
std::cout << "constrained\n";
};
int main() {
S<int> x;
S<double> y;
x.f(); // prints constrained
y.f(); // prints general
return 0;
}

Template type check parameter C++, different type execution paths

I'd like slightly different logic in a template based on the template parameter. How might I type check a template parameter?
I have the following which I am surprised does not work:
class Bar {
Bar() {}
}
template <typename T>
void Foo(const &T a) {
if (!std::is_same<T, Bar>::value) {
// Do things that will fail for type Bar e.g.
a.Fail();
}
}
I do not wish to use template specialization as in reality template specialization ends up sharing a lot of code for my specific purpose (currently working code is using template specialization)
Currently this fails during compile: "no member "Fail" in "Bar"
Each branch should be valid for every type.
In C++17, you could use if constexpr which change that:
template <typename T>
void Foo(const &T a) {
if constexpr (!std::is_same<T, Bar>::value) {
// Do things that will fail for type Bar e.g.
a.Fail();
}
}
Before, you have to rely on specialization or overload.
As for example
template <typename T>
void Fail(const T& t) { t.Fail(); }
void Fail(const Bar&) { /*Empty*/ }
template <typename T>
void Foo(const &T a) {
// ...
Fail(a);
// ...
}
Rather than specializing the entire template function Foo, specialize a helper method:
template <typename T>
void FooHelper(const T &a) {
a.fail();
}
template <>
void FooHelper<Bar>(const Bar &a) {
// something other than a.fail()
}
template <typename T>
void Foo(const T &a) {
FooHelper<T>(a);
// etc. etc.
}

Is there a way to pass templated function signature as a template template parameter

By using template template parameters one can pass to a class a templated class without specifying types on its parameters. I was wondering is there a way to pass into a template template parameter a templated signature of a function to be able to specialize which variant of the function is to be considered forward.
To be clear - I know I cannot do that:
template <class T>
void foo() { /*...*/ }
template <template <class...> class FooType>
struct Foo { /*...*/ };
int main() {
Foo<decltype(foo)> f;
}
But somehow I would like to be able to pass templated signature of function to Foo. Is it even possible?
I couldn't believe that this is not possible so I searched a bit and found a way to do exactly what I wanted. I used templated using with a syntax:
template <template<class... Args> class FooType>
struct Foo {
FooType<int> ft;
};
template <class Res, class... Args>
using FooSignature = Res(*)(Args...);
int foo() {
return 1;
}
int main() {
Foo<FooSignature> f;
f.ft = foo;
}
This however still leaves the question how can this be possible since the standard states something opposite.
In the example below one has a template template parameter that accepts the preferred signature for the function.
Because of the specialization and the lack of a body for the template class, only types for callables are accepted.
It is a generalization of what the OP actually asked:
#include<cassert>
template<typename F>
struct S;
template<typename R, typename... Args>
struct S<R(Args...)> {
using type = R(*)(Args...);
};
template<template<typename> class F>
struct T {
typename F<void(int)>::type ft;
typename F<double(double, double)>::type gt;
};
void f(int) { }
double g(double x, double y) { return x+y; }
int main() {
T<S> t;
t.ft = f;
t.gt = g;
t.ft(42);
auto v = t.gt(1., 1.);
assert(v == 2.);
}
As can be seen in this answer
Template of function pointer is illegal in C++
The C++ Standard says in $14/1,
A template defines a family of classes or functions.
Further quoting from the linked answer:
Please note that it does NOT say "A template defines a family of classes, functions or function pointers"
However, you can pass concrete function pointers, and specialise on their signature:
#include <iostream>
template <class T>
void foo(T) { }
template <typename>
struct Foo;
template<typename T>
struct Foo<void(T)>
{
void cb() { std::cout << "T\n"; }
};
template<>
struct Foo<void(int)>
{
void cb() { std::cout << "int\n"; }
};
template<>
struct Foo<void(double)>
{
void cb() { std::cout << "double\n"; }
};
int main()
{
Foo<decltype(foo<int >)>().cb(); // outputs 'int'
Foo<decltype(foo<double>)>().cb(); // outputs 'double'
Foo<decltype(foo<char >)>().cb(); // outputs 'T'
return 0;
}
template of template is still a template.
template <class T>
void foo() { /*...*/ }
template <typename T>
struct Foo { /*...*/ };
int main() {
Foo<decltype(foo<int>)> f;
}
You cannot pass a function template as an argument. What you can do is wrap a function template in a generate lambda taking a tag parameter:
template <class T> struct tag_t { using type = T; };
template <class T>
void foo() { ... }
template <class F>
void call_func_with(F f) {
f(tag_t<int>{} );
f(tag_t<double>{} );
}
call_with_func([](auto tag) { foo<decltype(tag)::type>(); } );
Here, f(tag_t<X>{} ) ends up calling foo<X>(), as desired.