I have several similar classes inheriting from the same Base-Class/Interface (Base class 1), and they share a couple similar functions, but then also have their own distinct functions. They all also have their own member variables of different classes, and each of those inherits from the same Base-Class/Interface (Base class 2). Is it possible to define a variable in Base class 1, of type Base class 2, then in the actual implementation of classes using Base class 1, have the variable of type Base class 2 be its proper type. Kinda hard to explain, so simplified example below.
//Base-Class 1
class Shape
{
public Shape() {}
ShapeExtra m_var;
//The common functions
public GetVar(){ return m_var; }
}
class Circle : Shape
{
public Circle() { m_var = new CircleExtra(); }
public void CircleFunc()
{
m_var.CircleExtraFunc();
}
}
class Triangle : Shape
{
public Triangle() { m_var = new TriangleExtra(); }
public void TriangleFunc()
{
m_var.TriangleExtraFunc();
}
}
.
.
.
//Base_Class 2
class ShapeExtra
{
public ShapeExtra() {}
}
class CircleExtra : ExtraClass
{
public CircleExtra() {}
void CircleExtraFunc() {//Do stuff}
}
class TriangleExtra : ExtraClass
{
public TriangleExtra() {}
void TriangleExtra() {//Do stuff}
}
.
.
.
So, I need the m_var in the child classes to be kept it as its own unique version. Because right now (w/o the extra CircleExtra m_var;), the GetVar() works, but in CircleFunc, m_var is still type of ShapeExtra, and thus doesn't know that CircleExtraFunc exists. I could cast m_var each time I wanted to do that, but that is repetitive and not worth it in my real-world case. Is there a way to utilize the functions in unique classes based off of ShapeExtra, while keeping the GetVar() function in Shape?
Please ask questions if there is anything I left out.
Simply with inheritance and without using pointers it is not possible, as C++ is a statically-and-strictly-typed language.
You can inherit both the variable and the function, but you'll need to cast function return value.
You can also override the function to make it return the concrete type, but then you have to cast the variable inside the function.
You can also declare the same var with the concrete class in subclasses, but then you just hide the variable in the superclass and inherit nothing.
I'd rather go for a solution using templates. Make the type of the variable a template type and extend the template using a concrete type in subclasses. It'll work perfectly.
It's been a long time since I last programmed in C++ and I beg your pardon if there are errors in the following example. I'm sure you can easily make it work.
template <class S>
class Shape {
S m_var;
//......
public:
S var () {
return m_var;
}
//.......
}
class Circle: Shape <CircleExtra> {
// var method returns CircleExtra
//......
}
Edit:
Regarding some comment, to allow virtual invocation of the method, it is possible to use correlated return types. Something like the following example.
class Shape {
public:
virtual ShapeExtra *var () = 0;
}
template <typename SE>
class ConcreteShape: Shape {
public:
virtual SE *var() {
return &m_var;
}
// Constructor, etc.
private:
SE m_var;
}
Or some variation. Now concrete shapes can benefit from extending the template, as long as SE * is correlated with ShapeExtra * (the type parameter extends ShapeExtra). And you can vall the method transparently through Shape interface.
Using pointers, this is totally possible.
Using your example, you could do something like this:
#include <iostream>
#include <memory>
using namespace std;
//Extras
class ShapeExtra
{
public:
ShapeExtra() {}
void ShapeFunc() { std::cout << "Shape"; }
virtual ~ShapeExtra() = default; //Important!
};
class Shape
{
public:
std::unique_ptr<ShapeExtra> m_var;
//require a pointer on construction
//make sure to document, that Shape class takes ownership and handles deletion
Shape(ShapeExtra* p):m_var(p){}
//The common functions
ShapeExtra& GetVar(){ return *m_var; }
void ShapeFunc() {m_var->ShapeFunc();}
};
class CircleExtra : public ShapeExtra
{
public:
void CircleExtraFunc() {std::cout << "Circle";}
};
class Circle : public Shape
{
CircleExtra* m_var;
public:
Circle() : Shape(new CircleExtra()) {
m_var = static_cast<CircleExtra*>(Shape::m_var.get());
}
void CircleFunc()
{
m_var->CircleExtraFunc();
}
};
int main() {
Circle c;
//use the ShapeExtra Object
c.GetVar().ShapeFunc();
//call via forwarded function
c.ShapeFunc();
//call the circleExtra Function
c.CircleFunc();
return 0;
}
Test it on ideone
Note the use of pointers and a virtual destructor:
By using a virtual destructor in the ShapeExtra base class, you make it possible to destruct an object of any derived class, using a ShapeExtra*. This is important, because
by using a std::unique_ptr<ShapeExtra> instead of a plain C-pointer, we make sure that the object is properly deleted on destruction of Shape.
It is probably a good idea to document this behaviour, i.e. that Shape takes the ownership of the ShapeExtra*. Which especially means, that we do not delete CirleExtra* in the Circle destructor
I decided here to require the ShapeExtra* on construction, but its also possible to just use std::unique_ptr::reset() later and check for nullptr on dereferencing Shape::m_var
Construction order is this: On calling the constructor of Circle, we first create a new CircleExtra which we pass to Shape before finally the constructor of Circle is executed.
Destruction order is Circle first (was created last), then Shape which also destructs the ShapeExtra for us, including (via virtual function) the CircleExtra
I would recommend the following approach:
class ShapeExtra
{
public:
virtual ~ShapeExtra() { }
virtual void SomeCommonShapeFunc() { std::cout << "Shape"; }
};
class Shape
{
public:
virtual ShapeExtra &GetVar() = 0; // Accessor function.
};
Note that the class Shape does not have any data members at all. After that for each derived class you need:
class CircleExtra : public ShapeExtra
{
public:
void SomeCommonShapeFunc() { std::cout << "Circle"; }
};
class Circle : public Shape
{
CircleExtra m_var; // Data member with circle specific class.
public:
virtual ShapeExtra &GetVar() { return m_var; }
};
Implementation of virtual method in Circle will return reference to the base class ShapeExtra. This will allow using this extra in the base class.
Note that pointers and templates are not used at all. This simplifies the overall design.
Related
I'd welcome some help with C++ inheritance to get a better grasp of the concept.
Is it possible to "extend" member types when creating a derived class? I think my problem can be best demonstrated by a simple example, where I'd like to extend the class VehicleData with a new double variable:
class VehicleData {
int yearOfManufacture;
//Different routines, for example Serialize(), etc., warranting to create a class for just a bunch of variables
};
class BicycleData:VehicleData {
double frameHeight; //new property that only applies to bicycles
};
//Now I create the actual classes that use the types above
class Vehicle {
VehicleData data;
void PrintData(); //a function that works on basic vehicle data
};
class Bicycle:Vehicle {
BicycleData data; //should copy VehicleData when creating an instance of this class
};
The problem with this approach is that when I code the above and create a Bicycle instance, its BicycleData member hides the already existing VehicleData member.
Is there a way to extend the base class, i.e. simply add a new double variable (to store frame height in this example), and keep the already existing (year of manufacture) data?
As far as I can tell, there is no clean way to do exactly what you want with inheritance alone.
You could create a template out of your base class:
template <typename Data>
class BaseVehicle
{
Data data;
// etc.
};
class Vehicle : BaseVehicle<VehicleData>
{
// etc.
};
class Bicycle : BaseVehicle<BicycleData>
{
// etc.
};
Then the Vehicle and Bicycle classes would contain data field of VehicleData and BicycleData types respectively.
Since in your example Bicycle inherits from Vehicle privately (i.e. there is no support for using Bicycle polymorphically via pointer/reference to Vehicle), this would effectively be identical to what you want to achieve.
If you do want dynamic polymorphism, you should create a separate, preferably abstract, class, defining the interface for your vehicles, e.g.:
class VehicleInterface
{
public:
// Some pure virtual interface methods
virtual void moveTo(const Vector2 position) = 0;
virtual ~VehicleInterface() = default;
};
And then you can have your concrete vehicles inherit and implement this interface:
class Vehicle : BaseVehicle<VehicleData>, public VehicleInterface
{
public:
virtual void moveTo(const Vector2 position) override
{
// implementation for Vehicle
}
};
class Bicycle : BaseVehicle<BicycleData>, public VehicleInterface
{
public:
virtual void moveTo(const Vector2 position) override
{
// implementation for Bicycle
}
};
Then any function, which would like to work with vehicles polymorphically, can just accept a reference or a pointer to VehicleInterface:
void driveToWork(VehicleInterface* vehicle)
{
vehicle->moveTo(getWorkPosition());
// etc.
}
Short answer; Not in the way that you're aiming for, but you can achieve something similar.
Rather than have an instance declared as you have, if you make data a pointer. You can then have BicycleData inherit VehicleData and then just replace data with the new instance in the constructor of the Bicycle.
ie
class Vehicle {
void PrintData();
protected:
void replaceData(std::shared_ptr<VehicleData> d) {
data = d;
}
std::shared_ptr<VehicleData> getData() {
return data;
}
template<class T>
std::shared_ptr<T> getDataAs() {
return std::dynamic_pointer_cast<T>(data);
}
private:
std::shared_ptr<VehicleData> data;
};
class Bicycle:Vehicle {
Bicycle(){replaceData(std::make_shared<BicycleData>());}
std::shared_ptr<BicycleData> getData() {
return getDataAs<BicycleData>();
}
};
Here is the situation. Let's say we have a virtual base class (e.g. ShapeJuggler) which contains a method that takes a shared pointer to a virtual base class object (e.g. Shape) as argument. Let's jump into the following pseudo-code to understand:
class Shape {
}
class ShapeJuggler {
virtual void juggle(shared_ptr<Shape>) = 0;
}
// Now deriving a class from it
class Square : public Shape {
}
class SquareJuggler : public ShapeJuggler {
public:
void juggle(shared_ptr<Shape>) {
// Want to do something specific with a 'Square'
// Or transform the 'shared_ptr<Shape>' into a 'shared_ptr<Square>'
}
}
// Calling the juggle method
void main(void) {
shared_ptr<Square> square_ptr = (shared_ptr<Square>) new Square();
SquareJuggler squareJuggler;
squareJuggler.juggle(square_ptr); // how to access 'Square'-specific members?
}
make_shared or dynamic/static_cast don't seem to do the job.
Is it at all possible? Any ideas, suggestions?
Thanks
This is where std::dynamic_pointer_cast (or one of its friends) comes into play.
It's just like dynamic_cast, but for std::shared_ptrs.
In your case (assuming the Shape class is polymorphic so dynamic_cast works):
void juggle(shared_ptr<Shape> shape) {
auto const sq = std::dynamic_pointer_cast<Square>(shape);
assert(sq);
sq->squareSpecificStuff();
}
This is the multiple dispatch problem. Their are many solution to this problem, the cleanest might be using the visitor pattern, but if you just have one function that need multiple dispatch you could avoid using a visitor:
class SquareJuggler;
class TriangleJuggler;
//.... others concrete jugglers.
class Shape {
//The default behaviour for any juggler and any shape
virtual void juggle_by(Juggler& t) {
//default code for any shape an juggle
}
// list each juggler for which you may
// implement a specific behavior
virtual void juggle_by(SquareJuggler& t) {
//provides default behavior in case you will not
// create a specific behavior for a specific shape.
//for example, just call the unspecific juggler:
this->Shape::juggle_by(static_cast<Juggler&>(t));
}
virtual void juggle_by(TriangleJuggler& t) {
//provides default behavior in case you will not
//create a specific behavior for a specific shape.
//for example, just call the unspecific juggler:
this->Shape::juggle_by(static_cast<Juggler&>(t));
}
//...
};
// Now deriving a class from it
class Square : public Shape {
void juggle_by(SquareJuggler& s) override{
//code specific to SquareJuggler and Shape
}
};
class Triangle : public Shape {
void juggle_by(TriangleJuggler& t) override{
//code specific to TriangleJuggler and Shape
}
};
class ShapeJuggler {
virtual void juggle(shared_ptr<Shape> s) {
//by default (if default has sense):
s->juggle_by(*this);
}
};
class SquareJuggler: public ShapeJuggler {
public:
void juggle(shared_ptr<Shape> s) override {
s->juggle_by(*this);
}
};
class TriangleJuggler: public ShapeJuggler {
public:
void juggle(shared_ptr<Shape> s) override {
s->juggle_by(*this);
}
};
// Calling the juggle method
void main(void) {
shared_ptr<Square> square_ptr = (shared_ptr<Square>) new Square();
SquareJuggler squareJuggler;
squareJuggler.juggle(square_ptr);
//This last call, will perform two virtual calls:
// 1. SquareJuggler::juggle(shared_ptr<Shape);
// 2. Square::juggle_by(SquareJuggler&);
}
You could also defines your XXXJuggler as final, which will enable some devirtualization optimization.
in C++;
Is there a way of calling a function from a derived class through the base class
even when the function is not overridden? In other words, I'm using a base class in
order to have heterogeneous containers without boost; I want to call a member function
that is only specific to a derived class...
Example:
(I just made this code up so there's probably a syntax error but hopefully you get the gist)
class Vehicle
{
public:
virtual void do_vehicle_stuff();
// virtual void do_car_specific_stuff(); makes no sense here
}
class Car : public Vehicle
{
public:
void do_vehicle_stuff();
void do_car_specific_stuff();
}
Car a,b;
list<Vehicle> vehicle_list;
vehicle_list.push_back(a);
vehicle_list.push_back(b);
vehicle_list.front().do_car_specific_stuff();
error: 'Class Vehicle' has no member named 'do_car_specific_stuff()'
You are slicing your classes when you insert them into the list. In
C++ subtype polymorphism (the kind of polymorphism you are using) only
works through references or pointers but not values. When you insert
your carS into the list they are converted to VehicleS.
An example:
Car c;
std::vector<Vehicle> vs;
vs.push_back(c); // slicing happens
vs.front(); // not a car anymore, but just a vehicle,
// the Car-ness is lost through the copy operation
How do to it:
std::vector<std::unique_ptr<Vehicle>> vs;
vs.push_back(new Car());
vs.front(); // actually a Car
After you have resolved that fundamental flaw of your code, this might
help you:
Vehicle* vehiclep = new Car();
if(auto carp = dynamic_cast<Car*>(vehiclep)) {
carp->do_car_specific_stuff();
}
This is a rather costly operation and usually an indication of a
design smell, so you might want to rethink what you are doing.
Here's a more appropriate design:
struct Vehicle
{
virtual ~Vehicle() { }
void do_vehicle_stuff()
{
vehicle_impl();
}
private:
virtual void vehicle_impl() = 0;
};
struct Car : Vehicle
{
private:
virtual void vehicle_impl()
{
my_vehicle_stuff();
my_car_specific_stuff();
}
void my_vehicle_stuff() { /* what you had originally */ }
void my_car_specific_stuff() { /* car-only stuff */ }
};
std::list<std::unique_ptr<Vehicle>> vehicles;
vehicles.emplace_back(new Car);
vehicles.emplace_back(new Motorcycle);
vehicles.emplace_back(new Carriage);
I have this pimpl design where the implementation classes are polymorphic but the interfaces are supposed to just contain a pointer, making them polymorphic somewhat defeats the purpose of the design.
So I create my Impl and Intf base classes to provide reference counting. And then the user can create their implementations. An example:
class Impl {
mutable int _ref;
public:
Impl() : _ref(0) {}
virtual ~Impl() {}
int addRef() const { return ++_ref; }
int decRef() const { return --_ref; }
};
template <typename TImpl>
class Intf {
TImpl* impl;
public:
Intf(TImpl* t = 0) : impl(0) {}
Intf(const Intf& other) : impl(other.impl) { if (impl) impl->addRef(); }
Intf& operator=(const Intf& other) {
if (other.impl) other.impl->addRef();
if (impl && impl->decRef() <= 0) delete impl;
impl = other.impl;
}
~Intf() { if (impl && impl->decRef() <= 0) delete impl; }
protected:
TImpl* GetImpl() const { return impl; }
void SetImpl(... //etc
};
class ShapeImpl : public Impl {
public:
virtual void draw() = 0;
};
class Shape : public Intf<ShapeImpl> {
public:
Shape(ShapeImpl* i) : Intf<ShapeImpl>(i) {}
void draw() {
ShapeImpl* i = GetImpl();
if (i) i->draw();
}
};
class TriangleImpl : public ShapeImpl {
public:
void draw();
};
class PolygonImpl : public ShapeImpl {
public:
void draw();
void addSegment(Point a, Point b);
};
Here is where have the issue. There are two possible declaration for class Polygon:
class Polygon1 : public Intf<PolygonImpl> {
public:
void draw() {
PolygonImpl* i = GetImpl();
if (i) i->draw();
}
void addSegment(Point a, Point b) {
PolygonImpl* i = GetImpl();
if (i) i->addSegment(a,b);
}
};
class Polygon2 : public Shape {
void addSegment(Point a, Point b) {
ShapeImpl* i = GetImpl();
if (i) dynamic_cast<Polygon*>(i)->addSegment(a,b);
}
}
In the Polygon1, I have rewrite the code for draw because I have not inherited it. In Polygon2 I need ugly dynamic casts because GetImpl() doesn't know about PolygonImpl. What I would like to do is something like this:
template <typename TImpl>
struct Shape_Interface {
void draw() {
TImpl* i = GetImpl();
if (i) i->draw();
}
};
template <typename TImpl>
struct Polygon_Interface : public Shape_Interface<Timpl> {
void addSegment(Point a, Point b) { ... }
};
class Shape : public TIntf<ShapeImpl>, public Shape_Interface<ShapeImpl> {...};
class Polygon : public TIntf<PolygonImpl>, public Polygon_Interface<PolygonImpl> {
public:
Polygon(PolygonImpl* i) : TIntf<PolygonImpl>(i) {}
};
But of course there's a problem here. I can't access GetImpl() from the Interface classes unless I derive them from Intf. And if I do that, I need to make Intf virtual everywhere it appears.
template <typename TImpl>
class PolygonInterface : public virtual Intf<TImpl> { ... };
class Polygon : public virtual Intf<PolygonImpl>, public PolygonInterface { ... }
OR I can store a TImpl*& in each Interface and construct them with a reference to the base Intf::impl. But that just means I have a pointer pointing back into myself for every interface included.
template <typename TImpl>
class PolygonInterface {
TImpl*& impl;
public:
PolygonInterface(TImpl*& i) : impl(i) {}
...};
Both of these solutions bloat the Intf class, add an extra dereference, and basically provide no benefit over straight polymorphism.
So, the question is, is there a third way, that I've missed that would solve this issue besides just duplicating the code everywhere (with its maintenance issues)?
TOTALLY SHOULD, BUT DOESN'T WORK: I wish there were base classes unions that just overlaid the class layouts and, for polymorphic classes, required that they have the exact same vtable layout. Then both Intf and ShapeInterface would each declare a single T* element and access it identically:
class Shape : public union Intf<ShapeImpl>, public union ShapeInterface<ShapeImpl> {};
I should note that your Impl class is nothing more than the reimplementation of a shared_ptr without the thread safety and all those cast bonuses.
Pimpl is nothing but a technic to avoid needless compile-time dependencies.
You do not need to actually know how a class is implemented to inherit from it. It would defeat the purpose of encapsulation (though your compiler does...).
So... I think that you are not trying to use Pimpl here. I would rather think this is a kind of Proxy patterns, since apparently:
Polygon1 numberOne;
Polygon2 numberTwo = numberOne;
numberTwo.changeData(); // affects data from numberOne too
// since they point to the same pointer!!
If you want to hide implementation details
Use Pimpl, but the real one, it means copying in depth during copy construction and assignment rather than just passing the pointer around (whether ref-counted or not, though ref-counted is preferable of course :) ).
If you want a proxy class
Just use a plain shared_ptr.
For inheritance
It does not matter, when you inherit from a class, how its private members are implemented. So just inherit from it.
If you want to add some new private members (usual case), then:
struct DerivedImpl;
class Derived: public Base // Base implemented with a Pimpl
{
public:
private:
std::shared_ptr<DerivedImpl> _data;
};
There is not much difference with classic implementation, as you can see, just that there is a pointer in lieu of a bunch of data.
BEWARE
If you forward declare DerivedImpl (which is the goal of Pimpl), then the destructor automatically generated by the compiler is... wrong.
The problem is that in order to generate the code for the destructor, the compiler needs the definition of DerivedImpl (ie: a complete type) in order to know how to destroy it, since a call to delete is hidden in the bowels of shared_ptr. However it may only generate a warning at compilation time (but you'll have a memory leak).
Furthermore, if you want an in-depth copy (rather than a shallow one, which consists in the copy and the original both pointing to the same DerivedImpl instance), you will also have to define manually the copy-constructor AND the assignment operator.
You may decide to create a better class that shared_ptr which will have deep-copy semantics (which could be called member_ptr as in cryptopp, or just Pimpl ;) ). This introduce a subtle bug though: while the code generated for the copy-constructor and the assignement operator could be thought of as correct, they are not, since once again you need a complete type (and thus the definition of DerivedImpl), so you will have to write them manually.
This is painful... and I'm sorry for you.
EDIT: Let's have a Shape discussion.
// Shape.h
namespace detail { class ShapeImpl; }
class Shape
{
public:
virtual void draw(Board& ioBoard) const = 0;
private:
detail::ShapeImpl* m_impl;
}; // class Shape
// Rectangle.h
namespace detail { class RectangleImpl; }
class Rectangle: public Shape
{
public:
virtual void draw(Board& ioBoard) const;
size_t getWidth() const;
size_t getHeight() const;
private:
detail::RectangleImpl* m_impl;
}; // class Rectangle
// Circle.h
namespace detail { class CircleImpl; }
class Circle: public Shape
{
public:
virtual void draw(Board& ioBoard) const;
size_t getDiameter() const;
private:
detail::CircleImpl* m_impl;
}; // class Circle
You see: neither Circle nor Rectangle care if Shape uses Pimpl or not, as its name implies, Pimpl is an implementation detail, something private that is not shared with the descendants of the class.
And as I explained, both Circle and Rectangle use Pimpl too, each with their own 'implementation class' (which can be nothing more than a simple struct with no method by the way).
I think you were right in that I didn't understand your question initially.
I think you're trying to force a square shape into a round hole... it don't quite fit C++.
You can force that your container holds pointers to objects of a given base-layout, and then allow objects of arbitrary composition to be actually pointed to from there, assuming that you as a programmer only actually place objects that in fact have identical memory layouts (member-data - there's no such thing as member-function-layout for a class unless it has virtuals, which you wish to avoid).
std::vector< boost::shared_ptr<IShape> > shapes;
NOTE at the absolute MINIMUM, you must still have a virtual destructor defined in IShape, or object deletion is going to fail miserably
And you could have classes which all take a pointer to a common implementation core, so that all compositions can be initialized with the element that they share (or it could be done statically as a template via pointer - the shared data).
But the thing is, if I try to create an example, I fall flat the second I try to consider: what is the data shared by all shapes? I suppose you could have a vector of Points, which then could be as large or small as any shape required. But even so, Draw() is truly polymorphic, it isn't an implementation that can possibly be shared by multiple types - it has to be customized for various classifications of shapes. i.e. a circle and a polygon cannot possibly share the same Draw(). And without a vtable (or some other dynamic function pointer construct), you cannot vary the function called from some common implementation or client.
Your first set of code is full of confusing constructs. Maybe you can add a new, simplified example that PURELY shows - in a more realistic way - what you're trying to do (and ignore the fact that C++ doesn't have the mechanics you want - just demonstrate what your mechanic should look like).
To my mind, I just don't get the actual practical application, unless you're tyring to do something like the following:
Take a COM class, which inherits from two other COM Interfaces:
class MyShellBrowserDialog : public IShellBrowser, public ICommDlgBrowser
{
...
};
And now I have a diamond inheritence pattern: IShellBrowser inherits ultimately from IUnknown, as does ICommDlgBrowser. But it seems incredibly silly to have to write my own IUnknown:AddRef and IUnknown::Release implementation, which is a highly standard implementation, because there's no way to cause the compiler to let another inherited class supply the missing virtual functions for IShellBrowser and/or ICommDlgBrowser.
i.e., I end up having to:
class MyShellBrowserDialog : public IShellBrowser, public ICommDlgBrowser
{
public:
virtual ULONG STDMETHODCALLTYPE AddRef(void) { return ++m_refcount; }
virtual ULONG STDMETHODCALLTYPE Release(void) { return --m_refcount; }
...
}
because there's no way I know of to "inherit" or "inject" those function implementations into MyShellBrowserDialog from anywhere else which actually fill-in the needed virtual member function for either IShellBrowser or ICommDlgBrowser.
I can, if the implementations were more complex, manually link up the vtable to an inherited implementor if I wished:
class IUnknownMixin
{
ULONG m_refcount;
protected:
IUnknonwMixin() : m_refcount(0) {}
ULONG AddRef(void) { return ++m_refcount; } // NOTE: not virutal
ULONG Release(void) { return --m_refcount; } // NOTE: not virutal
};
class MyShellBrowserDialog : public IShellBrowser, public ICommDlgBrowser, private IUnknownMixin
{
public:
virtual ULONG STDMETHODCALLTYPE AddRef(void) { return IUnknownMixin::AddRef(); }
virtual ULONG STDMETHODCALLTYPE Release(void) { return IUnknownMixin::Release(); }
...
}
And if I needed the mix-in to actually refer to the most-derived class to interact with it, I could add a template parameter to IUnknownMixin, to give it access to myself.
But what common elements could my class have or benefit by that IUnknownMixin couldn't itself supply?
What common elements could any composite class have that various mixins would want to have access to, which they needed to derive from themselves? Just have the mixins take a type parameter and access that. If its instance data in the most derived, then you have something like:
template <class T>
class IUnknownMixin
{
T & const m_outter;
protected:
IUnknonwMixin(T & outter) : m_outter(outter) {}
// note: T must have a member m_refcount
ULONG AddRef(void) { return ++m_outter.m_refcount; } // NOTE: not virtual
ULONG Release(void) { return --m_outter.m_refcount; } // NOTE: not virtual
};
Ultimately your question remains somewhat confusing to me. Perhaps you could create that example that shows your preferred-natural-syntax that accomplishes something clearly, as I just don't see that in your initial post, and I can't seem to sleuth it out from toying with these ideas myself.
I have seen lots of solutions to this basic conundrum: polymorphism + variation in interfaces.
One basic approach is to provide a way to query for extended interfaces - so you have something along the lines of COM programming under Windows:
const unsigned IType_IShape = 1;
class IShape
{
public:
virtual ~IShape() {} // ensure all subclasses are destroyed polymorphically!
virtual bool isa(unsigned type) const { return type == IType_IShape; }
virtual void Draw() = 0;
virtual void Erase() = 0;
virtual void GetBounds(std::pair<Point> & bounds) const = 0;
};
const unsigned IType_ISegmentedShape = 2;
class ISegmentedShape : public IShape
{
public:
virtual bool isa(unsigned type) const { return type == IType_ISegmentedShape || IShape::isa(type); }
virtual void AddSegment(const Point & a, const Point & b) = 0;
virtual unsigned GetSegmentCount() const = 0;
};
class Line : public IShape
{
public:
Line(std::pair<Point> extent) : extent(extent) { }
virtual void Draw();
virtual void Erase();
virtual void GetBounds(std::pair<Point> & bounds);
private:
std::pair<Point> extent;
};
class Polygon : public ISegmentedShape
{
public:
virtual void Draw();
virtual void Erase();
virtual void GetBounds(std::pair<Point> & bounds);
virtual void AddSegment(const Point & a, const Point & b);
virtual unsigned GetSegmentCount() const { return vertices.size(); }
private:
std::vector<Point> vertices;
};
Another option would be to make a single richer base interface class - which has all the interfaces you need, and then to simply define a default, no-op implementation for those in the base class, which returns false or throws to indicate that it isn't supported by the subclass in question (else the subclass would have provided a functional implementation for this member function).
class Shape
{
public:
struct Unsupported
{
Unsupported(const std::string & operation) : bad_op(operation) {}
const std::string & AsString() const { return bad_op; }
std::string bad_op;
};
virtual ~Shape() {} // ensure all subclasses are destroyed polymorphically!
virtual void Draw() = 0;
virtual void Erase() = 0;
virtual void GetBounds(std::pair<Point> & bounds) const = 0;
virtual void AddSegment(const Point & a, const Point & b) { throw Unsupported("AddSegment"); }
virtual unsigned GetSegmentCount() const { throw Unsupported("GetSegmentCount"); }
};
I hope that this helps you to see some possibilities.
Smalltalk had the wonderful attribute of being able to ask the meta-type-system whether a given instance supported a particular method - and it supported having a class-handler that could execute anytime a given instance was told to perform an operation it didn't support - along with what operation that was, so you could forward it as a proxy, or you could throw a different error, or simply quietly ignore that operation as a no-op).
Objective-C supports all of those same modalities as Smalltalk! Very, very cool things can be accomplished by having access to the type-system at runtime. I assume that .NET can pull of some crazy cool stuff along those lines (though I doubt that its nearly as elegant as Smalltalk or Objective-C, from what I've seen).
Anyway, ... good luck :)
In a C++ physics simulation, I have a class called Circle, and Square. These are Shapes, and have a method called push(), which applies force to it. There is then a special case of Circle, call it SpecialCircle, in which push() should exhibit slightly different properties. But in fact, there is also SpecialSquare() which should exhibit the same force properties. So I'd like to have an abstract base class called Shape which takes care of Circles and Squares, but then I'd also like an abstract base class called Special, which applies special properties to force().
What's the best way to design this class structure?
So far, I've got:
class Shape {
virtual void push();
};
class Circle : public Shape {};
class Square : public Shape {};
class Special {
virtual void push();
};
class SpecialCircle : public Circle, Special {};
class SpecialSquare : public Square, Special {};
Of course, the above won't compile, since Special::push() and Shape::push() conflict. I get "error: request for member ‘push’ is ambiguous", as expected.
How can I re-organize my class structure so that Circle and Square can share certain properties with each other, but SpecialCircle and SpecialSquare can still inherit from Shape, and also inherit modified functionality from Special?
Thanks.
ps., is this the diamond inheritance problem?
Another solution (it may or may not fit your needs, it depends on the details of your implementation):
Have the class Behavior, and let NormalBehavior and SpecialBehavior inherit from it.
Have the class Shape, and let Square and Circle inherit from it. Let Shape be an aggregate type, with a Behavior member (i.e. you pass a Behavior object to the various Shape constructors). In other words, let a Shape have a Behavior.
Delegate the actual differences in the behavior of shapes to methods of the Behavior hierarchy.
Conversely, you can:
Have the class PhysicalObject, and let NormalObject and SpecialObject inherit from it;
Have the class Shape, and let Square and Circle inherit from it;
Let a PhysicalObject have a Shape.
Prefer aggregation over inheritance. This is an application of the Bridge pattern. The advantage of this strategy with respect to having Square, SpecialSquare, Circle, and SpecialCircle, is that tomorrow you'll have to add Rectangle, Hexagon and so on, and for each shape you add you'll have to implement two classes (duplicated code is evil); this is, in my opinion, the real issue that Bridge addresses.
It's said that every problem in software can be solved by adding an additional layer of indirection.
Herb Sutter has an excellent article on how to solve your problem: Multiple Inheritance - Part III
In short, you use intermediate classes to 'rename' the virtual functions. As Herb says:
Renaming Virtual Functions
If the two inherited functions had different signatures, there would be no problem: We would just override them independently as usual. The trick, then, is to somehow change the signature of at least one of the two inherited functions.
The way to change a base class function's signature is to create an intermediate class which derives from the base class, declares a new virtual function, and overrides the inherited version to call the new function
Here's a long example using your classes:
class Shape {
public:
virtual void push() = 0;
};
class Circle : public Shape
{
public:
void push() {
printf( "Circle::push()\n");
}
};
class Square : public Shape
{
public:
void push() {
printf( "Square::push()\n");
}
};
class Special {
public:
virtual void push() = 0;
};
class Circle2: public Circle
{
public:
virtual void pushCircle() = 0;
void push() {
pushCircle();
}
};
class Square2: public Square
{
public:
virtual void pushSquare() = 0;
void push() {
pushSquare();
}
};
class Special2 : public Special
{
public:
virtual void pushSpecial() = 0;
void push() {
pushSpecial();
}
};
class SpecialCircle : public Circle2, public Special2
{
public:
void pushSpecial() {
printf( "SpecialCircle::pushSpecial()\n");
}
void pushCircle() {
printf( "SpecialCircle::pushCircle()\n");
}
};
class SpecialSquare : public Square2, public Special2
{
public:
void pushSpecial() {
printf( "SpecialSquare::pushSpecial()\n");
}
void pushSquare() {
printf( "SpecialSquare::pushSquare()\n");
}
};
int main( int argc, char* argv[])
{
SpecialCircle sc;
SpecialSquare ss;
// sc.push(); // can't be called - ambiguous
// ss.push();
sc.pushCircle();
ss.pushSquare();
Circle* pCircle = ≻
pCircle->push();
Square* pSquare = &ss;
pSquare->push();
Special* pSpecial = ≻
pSpecial->push();
pSpecial = &ss;
pSpecial->push();
return 0;
}
Rather than thinking of code reuse through inheritance, the use of mixins will give you the code reuse you want without the problems of multiple inheritance.
If you are unfamiliar with the technique, do a search on SO or Google. Make sure you search for both "mixin" and "Curiously Recurring Template Pattern". There are heaps of great articles around to get you started.
When you have to inherit from multiple interfaces with the same method the compiler can't tell which one are you trying to call, you can fix this by overriding such method and call the one you want.
class SpecialCircle : public Circle, Special {
public:
virtual void push() { Special::push(); }
};
class SpecialSquare : public Square, Special {
public:
virtual void push() { Special::push(); }
};
But in this case I think the correct OO approach is to factor out the push behavior in its own class, like Federico Ramponi have suggested.
Have a SpecialShape from Shape and SpecialCircle and SpecialSquare from SpecialShape.
Well, if the special and normal circles can be both applied forces to, and the special circle has another method that applies special forces, why not have two interfaces and two methods?
struct Applicable {
virtual ~Applicable() { }
// if it applies force, better be explicit with naming it.
virtual void applyForce() = 0;
};
struct SpecialApplicable {
virtual ~SpecialApplicable() { }
virtual void applySpecialForce() = 0;
};
struct Shape {
virtual ~Shape() { }
Size getSize();
Point getPosition();
// ...
};
struct Circle : Shape, Applicable {
virtual void applyForce() { /* ... */ }
}
struct SpecialCircle : Circle, SpecialApplicable {
virtual void applySpecialForce() { /* .... */ }
};
If it doesn't make sense if there is both a special and a normal apply method (which the name of the class - SpecialCircle - suggests), then why not do even this:
struct Circle : Shape, Applicable {
virtual void applyForce() { /* ... */ }
}
struct SpecialCircle : Circle {
// applies force, but specially
virtual void applyForce() { /* .... */ }
};
You can also put the applyForce into the Shape class. It also depends on the environment in which those classes are used. What, in any case, you really should avoid is having the same method in two base classes that appear in two difference base-lattices. Because that inevitable will lead to such ambiguity problems. The diamond inheritance is when you use virtual inheritance. I believe there are other good answers on stackoverflow explaining that. It isn't applicable for your problem, because the ambiguity arises because the method appears in two base class sub-objects of different types. (It only solves such cases where the base classes have the same type. In those cases, it will merge the base classes and there will only be one base class sub-object contained - inherited by virtual inheritance)