Styjun

I need to write a template specialization which will work with pointers on arrays (these things: char(*)[]). And we will write this code for that

// our class
template<typename T>
struct CoolStruct

 static void Print(); // will print "Default"
;

//one specialization for sized arrays
template<typename T, size_t S>
struct CoolStruct<T(*)[S]>

 static void Print(); // will print "T(*)[sized]"
;

// and one for arrays without size
template<typename T>
struct CoolStruct<T(*)[]>

 static void Print(); // will print "T(*)[]"
;

And when in our code we will do something like this:

int (*arrptr)[10];
CoolClass<decltype(arrptr)>::Print(); 

The console will print "T(*)[sized]" as we expected (considering that we wrote implementation for all methods of course). But what if we will wrote our code like this:

int (**arrptr_d)[10];
CoolClass<decltype(arrptr_d)>::Print();

In this case the console will actually print "Default". We could write another specialization like this:

template<typename T, size_t S>
struct CoolStruct<T(**)[S]>

 static void Print(); // will print "T(*)[sized]"
;

but I want to find another solution (I don't want to write another dozen specializations when I would need to use int(****)[]). So considering we using C++17 standard, is there a way to do such a thing?

P.S. And yes, sorry for my terrible English; it is not my native language.

If you don't care about other pointers, you could do a partial specialization that delegates double pointers to the specialization for single pointers:

template<typename Pointee>
struct CoolStruct<Pointee**>
 static void Print()
 CoolStruct<Pointee*>::Print();
 
;