Lets say i have something like this coded
class Normal_Mode;
class Fast_Mode;
class File_Control; //handles all operations with reading/writing in file
class Main_Control {
private:
some_class *root; //all other classes need access to root pointer since there is all the data(binary tree)
File_Control *c_file_control;
Fast_Mode *c_fast_mode;
...
}
Main_Control::Main_Control ( int argc, char* argv[]) {
...
if ( argc > 1 ) {
c_fast_mode = new Fast_Mode(argc, argv[]);
} else {
c_normal_mode = new Normal_Mode();
};
...
};
int main (int argc, char* argv[]) {
Main_Control c_main_control(argc,argv);
return 0;
}
Lets say user input had argc > 1 and i am happy doing stuff with users input in Fast_Mode class but when i am finished and want to write stuff to file or read something from file while in Fast_Mode.
How do people in real world access File_control class?
Do they make some global array full with pointers to these kinda of classes who need only 1 instance.
Do they pass pointers to Fast_Mode and other classes so it can have it stored in private members for access.
or they construct/destruct such classes all the time depending on when it is needed.
And what do they do with such *root pointer where all the actual data is stored and lot of other classes needs to access it
Or my design ideas are completely wrong and people in real world do it some other way?
I don't really know what you are trying to achieve with this: if possible make you goal more clear but what I would say is that people would ususally create an abstract interface called 'Mode' and then have both Normal_Mode and Fast_Mode implement it. That way you could write the following code:
class Main_Control {
private:
some_class *root; //all other classes need access to root pointer since there is all the data(binary tree)
File_Control *c_file_control;
Mode *c_mode;
...
};
And then you could set it like:
if ( argc > 1 ) {
c_mode = new Fast_Mode(argc, argv[]);
} else {
c_mode = new Normal_Mode();
};
So you just put the common functions in the Mode interface so that they behave the same way but can be placed inside the same mode variable. To learn more about class inheritance and polymorphism look here. I hope that this answers your question.
Related
I have classes DBGameAction and ServerGameAction which has common parent class GameAction. Classes DBGameAction and ServerGameAction it's a API for safety working with entity GameAction from different part of program.
My question is: is it normal at first create DBGameAction entity and then cast it to the ServerGameAction entity? Or maybe it's a wrong program design?
My program:
#include <vector>
#include <string>
#include <iostream>
class GameAction
{
protected:
/* Need use mutex or something else for having safety access to this entity */
unsigned int cost;
unsigned int id;
std::vector<std::string> players;
GameAction(){}
public:
unsigned int getCost() const
{
return cost;
}
};
class DBGameAction : public GameAction
{
public:
void setCost(unsigned int c)
{
cost = c;
}
void setId(unsigned int i)
{
id = i;
}
};
class ServerGameAction : public GameAction
{
ServerGameAction(){}
public:
void addPlayer(std::string p)
{
players.push_back(p);
}
std::string getLastPlayer() const
{
return players.back();
}
};
int main(int argc, char *argv[])
{
DBGameAction *dbga = 0;
ServerGameAction *sga = 0;
try {
dbga = new DBGameAction;
}
catch(...) /* Something happens wrong! */
{
return -1;
}
sga = reinterpret_cast<ServerGameAction*>(dbga);
sga->addPlayer("Max");
dbga->setCost(100);
std::cout << dbga->getCost() << std::endl;
std::cout << sga->getLastPlayer() << std::endl;
delete dbga;
sga = dbga = 0;
return 0;
}
It is wrong program design.
Is there a reason why you are not creating GameAction variables which you then downcast to DBGameAction and ServerGameAction?
I haven't used reinterpret_cast in many occasions but I am sure it shouldn't be used this way. You should try to find a better design for the interface of your classes. Someone who uses your classes, doesn't have a way to know that he needs to do this sort of castings to add a player.
You have to ask yourself, if adding a player is an operation that only makes sense for ServerGameActions or for DBGameActions too. If it makes sense to add players to DBGameActions, then AddPlayer should be in the interface of DBGameAction too. Then you will not need these casts. Taking it one step further, if it is an operation that makes sense for every possible GameAction you may ever have, you can put it in the interface of the base class.
I have used a similar pattern effectively in the past, but it is a little different than most interface class setups. Instead of having a consistent interface that can trigger appropriate class-specific methods for accomplishing similar tasks on different data types, this provides two completely different sets of functionality which each have their own interface, yet work on the same data layout.
The only reason I would pull out this design is for situations where the base class is data-only and shared between multiple libraries or executables. Then each lib or exe defines a child class which houses all the functionality that it's allowed to use on the base data. This way you can, for example, build your server executable with all kinds of nice extra functions for manipulating game data that the client isn't allowed to use, and the server-side functionality doesn't get built into the client executable. It's much easier for a game modder to trigger existing, dormant functionality than to write and inject their own.
The main part of your question about casting directly between the child classes is making us worry, though. If you find yourself wanting to do that, stop and rethink. You could theoretically get away with the cast as long as your classes stay non-virtual and the derived classes never add data members (the derived classes can't have any data for what you're trying to do anyway, due to object slicing), but it would be potentially dangerous and, most likely, less readable code. As #dspfnder was talking about, you would want to work with base classes for passing data around and down-cast on-demand to access functionality.
With all that said, there are many ways to isolate, restrict, or cull functionality. It may be worth reworking your design with functionality living in friend classes instead of child classes; that would require much less or no casting.
I've tried to solve my problem for 2 days now and failed miserably. Internet does not help.
What I'm trying to do is to communicate two classes which reside within another class.
This is my first "big" project so I assume my design is terrible for you guys.
Also, my program is split between a lot of files which may be confusing.
Lets hit it! For the sake readability, I've changed every member to public.
This is my MainOGLController class which is the main class that controls everything my program does:
class MainOGLController
{ // I deleted constructor/destructor from this quote
public:
DisplayController* Display;
StellarManager* Manager; // it will need to use something from Display
void RenderScene();
bool CreateNewDisplay(int, char*[]); // argc argv
}
Ok, this is how i create instance of this class in file with main():
#include "MainOGLController.h"
MainOGLController Controller;
int main(int argc, char* argv[])
{
if ( Controller.CreateNewDisplay(argc, argv) ) return 1; // if it fails then exit;
// some opengl code here
return 0;
}
Now you are probably wondering how does the CreateNewDisplay method look like:
bool MainOGLController::CreateNewDisplay(int argc, char* argv[])
{
Display = new DisplayController(argc, argv);
Manager = new StellarManager(&Display); // me trying to make reference to Display
// to be able to use it within Manager
//ogl code
else return 0;
}
OK, so I'm creating Manager there and now we should see how i created the StellarManager class:
class StellarManager
{
std::vector<Stellar*> VectorManager; // objects from this vector will need to use
// ptrDisplay to access Display
DisplayController* ptrDisplay;
StellarManager(DisplayController&);
void addEntity();
};
Now for the constructor:
StellarManager::StellarManager(DisplayController& _p) // me trying to do anything
{
*ptrDisplay = _p;
}
So at this point I should have instance of MainOGLController, and within it, a pointer to DisplayController and StellarController, where StellarController should have its own pointer to the same DisplayController.
Now somewhere withing working piece of code I'm calling the addEntity method:
void StellarManager::addEntity()
{
VectorManager.push_back(new Stellar(&ptrDisplay); // sending ptrDisplay so that the
// Stellar object can use it
}
Stellar class is defined like this:
class Stellar
{
public:
DisplayController* ptrDisplay;
Stellar(DisplayController**);
void Draw(); // finally, heres where i want to use this Display pointer
};
Stellar constructor:
Stellar::Stellar(DisplayController** _p)
{
*ptrDisplay = **_p;
}
OKAY! Thats the final piece. All i want to do now is simply call method Draw which belongs to Stellar class and use Display which is located in MainOGLController.
Manager->VectorManager[0].Draw();
Oh and the Draw looks just like this:
void Stellar::Draw(int _mode)
{
GLMatrixStack* mvm = &(ptrDisplay->modelViewMatrix);
mvm->Scale(2, 0.5, 0.5); // Scale is a method from GLMatrixStack
}
Thats all folks, if theres any better way of doing this, im all ears.
What I did does not work, I'm able to use the *ptrDisplay from Stellar class but nothing happens so I guess I'm not using its reference but a copy.
Sorry, I know this is a lot of code and it may be very confusing. I just dont know what to do now...
It looks like the problem is here:
Stellar::Stellar(DisplayController** _p)
{
*ptrDisplay = **_p;
}
You're dereferencing a pointer (ptrDisplay) that was never initialized. This results in undefined behavior. I think this captures what you wanted to do:
Stellar::Stellar(DisplayController* _p) : ptrDisplay(_p)
{
}
It's not necessary to pass a pointer-to-pointer-to-DisplayController; all your Stellar class needs is a pointer to a DisplayController. Moreover, it sounds like you don't want to dereference _p and copy it, so simply copying the pointer (via ptrDisplay(_p)) will result in ptrDisplay pointing to the same object as _p.
I very often run into the problem that I want to implement a data structure, and would like to allow users to extend it with functional functionality; that is add functionality but not bytes to the data structure. An example could be extending std::vector with a sum method:
#include <iostream>
#include <vector>
// for the header file
template<>
int std::vector<int>::sum();
//for the object file
template<>
int std::vector<int>::sum() {
int s=0;
for(auto v = this->begin(); v!=this->end(); ++v) s+=*v;
return s;
}
int main() {
std::vector<int> numbers;
numbers.push_back(5);
numbers.push_back(2);
numbers.push_back(6);
numbers.push_back(9);
std::cout << numbers.sum() << std::endl;
return 0;
}
See: http://ideone.com/YyWs5r
So this is illegal, since one may not add functions to a class like this. Obviously this is some design decision of c++(11). It can be circumvented in two ways, that is defining a
int sum(std::vector<int> &v) { ... }
This is how std::sort works, so I guess it is the way c++(11) is intended. I think this is to the best of my knowledge the best way to do it in c++. However, it does not allow me to access private properties of std::vector. Maybe I am evil by assuming access to private properties in a (sort-of) method is fair. However, often I want users of my classes to not access certain stuff, however would like to allow extenders of my to access them. For example I can imagine that std::sort can be optimized w.r.t. specific container implementation knowledge and access.
Another way is inheriting std::vector, but I find that plain unacceptable for these reasons:
if two parties have extended the class with methods, which one would like to use, then one would need to convert from one child class to another. This is ludicrous, as one converts data to data without actually changing bytes, as both child classes (can) have exactly the same memory implementation and segmentation. Please also note that data conversion in general is boilerplate code, and boilerplate code should imho be considered evil.
one is unnecessarily mixing functionality with data structures, for example, a class name sum_vector or mean_vector is completely.
As a short reminder, I am not looking for answers like "You cannot do that in c++", I already know that (Add a method to existing C++ class in other file). However, I would like to know if there is a good way to do functional class extensions. How should I manage accessing private fields? What would be reasons why it is unreasonable for me to want private field access; why can't I discriminate between extender and user access?
Note: one could say that an extender needs protected access and a user needs public access, however, like I said, that would be for the inheritance way of extending, and I dislike it strongly for the aforementioned reasons.
You never should want to access private members of Standard Containers because they are not part of their interfaces.
However, you already can extend the functionality of the Standard Containers like std::vector: namely through the judicious use of iterators and Standard Algorithms.
E.g. the sum functionality is given by a non-member function that uses the begin() and end() functionality of std::vector
#include <algorithm>
#include <iterator>
#include <vector>
template<class Container, class Ret = decltype(*begin(c))>
Ret sum(Container const& c)
{
return std::accumulate(begin(c), end(c), Ret{});
}
Consider something like this:
#include <iostream>
class Foo{
int a;
public:
Foo(int a){this->a = a;}
int getA(){return this->a;}
void * extendedMethod(void *(*func)(int, char **, Foo*), int argc, char **argv){
return func(argc, argv, this);
}
};
void * extendFooWith(int argc, char **argv, Foo* self){
/* You can call methods on self... but still no access to private fields */
std::cout << self->getA();
return self;
}
int main(int argc, char const *argv[])
{
Foo foo(5);
foo.extendedMethod(extendFooWith, 0 /*argc*/, NULL /*argv*/);
return 0;
}
That's the best way I thought of extending a class with a method. The only way to access private fields would be from inside extendedMethod() i.e. something like this is possible: return func(this->a, argv, this); but then it is not that generic any more. One way to improve it could be checking inside extendedMethod() what kind of pointer was passed and according to it access the private fields you are interested in and pass those to func(), but this will require adding code to extendedMethod() for every other method you will extend your class with.
I have a class with a complex construction process with many parameters. Multiple clients share objects of this class, and the union of these clients parameters are used to instantiate the class. Therefore I have a factory class that stores these requirements, checks consistency of the various clients' requests, and instantiates the class.
Additionally, there are a common set of use models (or sets of parameters) which multiple clients use for multiple factories.
For instance, consider an example. (Note that the actual code is C++, but my experience is in Python so I'll pseudo-code in Python. Yes, I know that this example wouldn't actually work as-is.)
class Classroom:
def __init__(self, room_size=None, n_desks=None, n_boards=None,
n_books=None, has_globe=False, ... ):
...
class ClassroomFactory:
def __init__(self):
self._requirements = dict()
def addRequirement(self, name, value):
if name.startswith("n_"):
self._requirements[name] = max(value, self._requirements.get(name, 0))
...
def createClassroom(self):
return Classroom(**self._requirements)
# instantiate the factory
factory = ClassroomFactory()
# "client 1" is a geography teaacher
factory.addRequirement("n_desks", 10)
factory.addRequirement("n_boards", 1)
factory.addRequirement("has_globe", True)
# "client 2" is a math teacher
factory.addRequirement("n_desks", 10)
factory.addRequirement("n_boards", 1)
# "client 3" is a after-school day-care
factory.addRequirement("room_size", (20,20))
factory.addRequirement("has_carpet", True)
room = factory.createClassroom()
The common use model is as a teacher, we need 10 desks and a board. I think this is best served by a non-member function/decorator, something like:
def makeTeacherRoom(factory):
factory.addRequirement("n_desks", 10)
factory.addRequirement("n_boards", 1)
return factory
This seems like a great example of the "prefer non-member/non-friend to member" paradigm.
The thing that I'm struggling with is, within the framework of a much bigger OO code, where should these types of non-member functions/decorators live, both in terms of namespace and in terms of actual file?
Should they live in the factory's file/namespace? They are closely related to the factory, but they're limitations on the general factory, and need not be used to use the factory.
Should they live in the client's file/namespace? The client understands these use models, but this would limit re-use amongst multiple clients.
Should they live with a common base class of the clients (for instance, one could imagine a "teacher" class/namespace which would also provide the non-member function makeTeacherRoom(), which would be inherited by MathTeacher and GeographyTeacher.
Should they live somewhere else completely, in a "utils" file? And if so in which namespace?
This is primarily a personal decision. Most of your options have no technical negative effects. For example:
They could, because of locality of use, but it's not necessary.
They could, because of locality of data, but again...
They could, although this one does seem like it could make things a bit messier. Making utility classes, you may have to end up inheriting them, or making parts virtual to override later, which will get ugly pretty quick.
This is my personal favorite, or a variant of this.
I typically make a relevantly-named util file (or class with static methods) and put it in the same namespace as the classes it utilates (the more helpful version of mutilate). For a Education::Teacher class, you could have a Education::TeacherUtils file or class containing the functions that operate on Teacher. This keeps a pretty obvious naming tie-in, but also puts the util functions in their own area, so they can be included from whatever needs them (in the Teacher.cpp or similar would prevent that). In the case of a class, you can make the util and base classes friends, which is occasionally helpful (but something to use rarely, as it may be a smell).
I've seen a naming variation, Education::Utils::Teacher, but that's somewhat harder to translate to files (unless you put things into a utils dir) and can also cause name resolution oddness (in some contexts, the compiler may try to use Education::Utils::Teacher instead of Education::Teacher when you didn't mean to). Because of this, I prefer to keep utils as a suffix.
You may want to handle non-member functions in a singleton class for your application. A factory maybe executed from the program, or another object.
C++ supports global functions (non member functions), but, using a single object for the application, "does the trick".
Additionally, since the "Classroom" object may be instantiated with many optional parameters, you may want to assign it, after calling the constructor ( "init" in python ).
// filename: "classrooms.cpp"
class ClassroomClass
{
protected:
int _Room_Size;
int _N_Desks;
int _N_Boards;
int _N_Books;
bool _Has_Globe;
public:
// constructor without parameters,
// but, can be declared with them
ClassroomClass()
{
_Room_Size = 0;
_N_Desks = 0;
_N_Boards = 0;
_N_Books = 0;
_Has_Globe = false;
} // ClassroomClass()
public int get_Room_Size()
{
return _Room_Size;
}
public void set_Room_Size(int Value)
{
_Room_Size = Value;
}
// other "getters" & "setters" functions
// ...
} // class ClassroomClass
class ClassroomFactoryClass
{
public:
void addRequirement(char[] AKey, char[] AValue);
} // class ClassroomFactoryClass
class MyProgramClass
{
public:
ClassroomFactoryClass Factory;
public:
void makeTeacherRoom();
void doSomething();
} // class MyProgramClass
void MyProgramClass::addRequirement(char[] AKey, char[] AValue)
{
...
} // void MyProgramClass::addRequirement(...)
void MyProgramClass::makeTeacherRoom()
{
Factory.addRequirement("n_desks", "10")
Factory.addRequirement("n_boards", "1")
} // void MyProgramClass::makeTeacherRoom(...)
void MyProgramClass::doSomething()
{
...
} // void MyProgramClass::doSomething(...)
int main(char[][] args)
{
MyProgramClass MyProgram = new MyProgramClass();
MyProgram->doSomething();
delete MyProgram();
return 0;
} // main(...)
Cheers
Personally I would make them static members of the class.
class File
{
public:
static bool load( File & file, std::string const & fileName );
private:
std::vector< char > data;
};
int main( void )
{
std::string fileName = "foo.txt";
File myFile;
File::load( myFile, fileName );
}
With static methods they have access to the private data of the class while not belonging to a specific instance of the class. It also means the methods aren't separated from the data they act on, as would be the case if you put them in a utility header somewhere.
I have a situation where I have an interface that defines how a certain class behaves in order to fill a certain role in my program, but at this point in time I'm not 100% sure how many classes I will write to fill that role. However, at the same time, I know that I want the user to be able to select, from a GUI combo/list box, which concrete class implementing the interface that they want to use to fill a certain role. I want the GUI to be able to enumerate all available classes, but I would prefer not to have to go back and change old code whenever I decide to implement a new class to fill that role (which may be months from now)
Some things I've considered:
using an enumeration
Pros:
I know how to do it
Cons
I will have to update update the enumeration when I add a new class
ugly to iterate through
using some kind of static list object in the interface, and adding a new element from within the definition file of the implementing class
Pros:
Wont have to change old code
Cons:
Not even sure if this is possible
Not sure what kind of information to store so that a factory method can choose the proper constructor ( maybe a map between a string and a function pointer that returns a pointer to an object of the interface )
I'm guessing this is a problem (or similar to a problem) that more experienced programmers have probably come across before (and often), and there is probably a common solution to this kind of problem, which is almost certainly better than anything I'm capable of coming up with. So, how do I do it?
(P.S. I searched, but all I found was this, and it's not the same: How do I enumerate all items that implement a generic interface?. It appears he already knows how to solve the problem I'm trying to figure out.)
Edit: I renamed the title to "How can I keep track of... " rather than just "How can I enumerate..." because the original question sounded like I was more interested in examining the runtime environment, where as what I'm really interested in is compile-time book-keeping.
Create a singleton where you can register your classes with a pointer to a creator function.
In the cpp files of the concrete classes you register each class.
Something like this:
class Interface;
typedef boost::function<Interface* ()> Creator;
class InterfaceRegistration
{
typedef map<string, Creator> CreatorMap;
public:
InterfaceRegistration& instance() {
static InterfaceRegistration interfaceRegistration;
return interfaceRegistration;
}
bool registerInterface( const string& name, Creator creator )
{
return (m_interfaces[name] = creator);
}
list<string> names() const
{
list<string> nameList;
transform(
m_interfaces.begin(), m_interfaces.end(),
back_inserter(nameList)
select1st<CreatorMap>::value_type>() );
}
Interface* create(cosnt string& name ) const
{
const CreatorMap::const_iterator it
= m_interfaces.find(name);
if( it!=m_interfaces.end() && (*it) )
{
return (*it)();
}
// throw exception ...
return 0;
}
private:
CreatorMap m_interfaces;
};
// in your concrete classes cpp files
namespace {
bool registerClassX = InterfaceRegistration::instance("ClassX", boost::lambda::new_ptr<ClassX>() );
}
ClassX::ClassX() : Interface()
{
//....
}
// in your concrete class Y cpp files
namespace {
bool registerClassY = InterfaceRegistration::instance("ClassY", boost::lambda::new_ptr<ClassY>() );
}
ClassY::ClassY() : Interface()
{
//....
}
I vaguely remember doing something similar to this many years ago. Your option (2) is pretty much what I did. In that case it was a std::map of std::string to std::typeinfo. In each, .cpp file I registered the class like this:
static dummy = registerClass (typeid (MyNewClass));
registerClass takes a type_info object and simply returns true. You have to initialize a variable to ensure that registerClass is called during startup time. Simply calling registerClass in the global namespace is an error. And making dummy static allow you to reuse the name across compilation units without a name collision.
I referred to this article to implement a self-registering class factory similar to the one described in TimW's answer, but it has the nice trick of using a templated factory proxy class to handle the object registration. Well worth a look :)
Self-Registering Objects in C++ -> http://www.ddj.com/184410633
Edit
Here's the test app I did (tidied up a little ;):
object_factory.h
#include <string>
#include <vector>
// Forward declare the base object class
class Object;
// Interface that the factory uses to communicate with the object proxies
class IObjectProxy {
public:
virtual Object* CreateObject() = 0;
virtual std::string GetObjectInfo() = 0;
};
// Object factory, retrieves object info from the global proxy objects
class ObjectFactory {
public:
static ObjectFactory& Instance() {
static ObjectFactory instance;
return instance;
}
// proxies add themselves to the factory here
void AddObject(IObjectProxy* object) {
objects_.push_back(object);
}
size_t NumberOfObjects() {
return objects_.size();
}
Object* CreateObject(size_t index) {
return objects_[index]->CreateObject();
}
std::string GetObjectInfo(size_t index) {
return objects_[index]->GetObjectInfo();
}
private:
std::vector<IObjectProxy*> objects_;
};
// This is the factory proxy template class
template<typename T>
class ObjectProxy : public IObjectProxy {
public:
ObjectProxy() {
ObjectFactory::Instance().AddObject(this);
}
Object* CreateObject() {
return new T;
}
virtual std::string GetObjectInfo() {
return T::TalkToMe();
};
};
objects.h
#include <iostream>
#include "object_factory.h"
// Base object class
class Object {
public:
virtual ~Object() {}
};
class ClassA : public Object {
public:
ClassA() { std::cout << "ClassA Constructor" << std::endl; }
~ClassA() { std::cout << "ClassA Destructor" << std::endl; }
static std::string TalkToMe() { return "This is ClassA"; }
};
class ClassB : public Object {
public:
ClassB() { std::cout << "ClassB Constructor" << std::endl; }
~ClassB() { std::cout << "ClassB Destructor" << std::endl; }
static std::string TalkToMe() { return "This is ClassB"; }
};
objects.cpp
#include "objects.h"
// Objects get registered here
ObjectProxy<ClassA> gClassAProxy;
ObjectProxy<ClassB> gClassBProxy;
main.cpp
#include "objects.h"
int main (int argc, char * const argv[]) {
ObjectFactory& factory = ObjectFactory::Instance();
for (int i = 0; i < factory.NumberOfObjects(); ++i) {
std::cout << factory.GetObjectInfo(i) << std::endl;
Object* object = factory.CreateObject(i);
delete object;
}
return 0;
}
output:
This is ClassA
ClassA Constructor
ClassA Destructor
This is ClassB
ClassB Constructor
ClassB Destructor
If you're on Windows, and using C++/CLI, this becomes fairly easy. The .NET framework provides this capability via reflection, and it works very cleanly in managed code.
In native C++, this gets a little bit trickier, as there's no simple way to query the library or application for runtime information. There are many frameworks that provide this (just look for IoC, DI, or plugin frameworks), but the simplest means of doing it yourself is to have some form of configuration which a factory method can use to register themselves, and return an implementation of your specific base class. You'd just need to implement loading a DLL, and registering the factory method - once you have that, it's fairly easy.
Something you can consider is an object counter. This way you don't need to change every place you allocate but just implementation definition. It's an alternative to the factory solution. Consider pros/cons.
An elegant way to do that is to use the CRTP : Curiously recurring template pattern.
The main example is such a counter :)
This way you just have to add in your concrete class implementation :
class X; // your interface
class MyConcreteX : public counter<X>
{
// whatever
};
Of course, it is not applicable if you use external implementations you do not master.
EDIT:
To handle the exact problem you need to have a counter that count only the first instance.
my 2 cents
There is no way to query the subclasses of a class in (native) C++.
How do you create the instances? Consider using a Factory Method allowing you to iterate over all subclasses you are working with. When you create an instance like this, it won't be possible to forget adding a new subclass later.