Counting Objects in C++ -----by Scott Meyers

Counting Objects in C++
Scott Meyers

It isn't hard to keep a count of all the objects allocated for a given

class in C++, unless you have to deal with distractions.

 

Sometimes easy things are easy, but they're still subtle. For example,

suppose you have a class Widget, and you'd like to have a way to find out at

 

run time how many Widget objects exist. An approach that's both easy to

implement and that gives the right answer is to create a static counter in

Widget, increment the counter each time a Widget constructor is called, and

decrement it whenever the Widget destructor is called. You also need a static

 

member function howMany to report how many Widgets currently exist. If Widget

 

did nothing but track how many of its type exist, it would look more or less

 

like this:

 

class Widget {

public:

    Widget() { ++count; }

    Widget(const Widget&) { ++count; }

    ~Widget() { --count; }

 

    static size_t howMany()

    { return count; }

 

private:

    static size_t count;

};

 

// obligatory definition of count. This

// goes in an implementation file

size_t Widget::count = 0;

 

This works fine. The only mildly tricky thing is to remember to implement the

 

copy constructor, because a compiler-generated copy constructor for Widget

wouldn't know to increment count.

 

If you had to do this only for Widget, you'd be done, but counting objects is

 

something you might want to implement for several classes. Doing the same

thing over and over gets tedious, and tedium leads to errors. To forestall

such tedium, it would be best to somehow package the above object-counting

code so it could be reused in any class that wanted it. The ideal package woul

d:

 

    * be easy to use — require minimal work on the part of class authors who

 

want to use it. Ideally, they shouldn't have to do more than one thing, that

 

is, more than basically say "I want to count the objects of this type."

    * be efficient — impose no unnecessary space or time penalties on client

 

classes employing the package.

    * be foolproof — be next to impossible to accidently yield a count that

 

is incorrect. (We're not going to worry about malicious clients, ones who

deliberately try to mess up the count. In C++, such clients can always find a

 

way to do their dirty deeds.)

 

Stop for a moment and think about how you'd implement a reusable

object-counting package that satisfies the goals above. It's probably harder

 

than you expect. If it were as easy as it seems like it should be, you

wouldn't be reading an article about it in this magazine.

new, delete, and Exceptions

 

While you're mulling over your solution to the object-counting problem, allow

 

me to switch to what seems like an unrelated topic. That topic is the

relationship between new and delete when constructors throw exceptions. When

 

you ask C++ to dynamically allocate an object, you use a new expression, as in

:

 

class ABCD { ... }; // ABCD = "A Big Complex Datatype"

ABCD *p = new ABCD; // a new expression

 

The new expression — whose meaning is built into the language and whose

behavior you cannot change — does two things. First, it calls a memory

allocation function called operator new. That function is responsible for

finding enough memory to hold an ABCD object. If the call to operator new

succeeds, the new expression then invokes an ABCD constructor on the memory

that operator new found.

 

But suppose operator new throws a std::bad_alloc exception. Exceptions of

this type indicate that an attempt to dynamically allocate memory has failed.

 

In the new expression above, there are two functions that might give rise to

 

that exception. The first is the invocation of operator new that is supposed

 

to find enough memory to hold an ABCD object. The second is the subsequent

invocation of the ABCD constructor that is supposed to turn the raw memory

into a valid ABCD object.

 

If the exception came from the call to operator new, no memory was allocated.

 

However, if the call to operator new succeeded and the invocation of the ABCD

 

constructor led to the exception, it is important that the memory allocated

by operator new be deallocated. If it's not, the program has a memory leak.

It's not possible for the client — the code requesting creation of the ABCD

 

object — to determine which function gave rise to the exception.

 

For many years this was a hole in the draft C++ language specification, but

in March 1995 the C++ Standards committee adopted the rule that if, during a

 

new expression, the invocation of operator new succeeds and the subsequent

constructor call throws an exception, the runtime system must automatically

deallocate the memory that operator new allocated. This deallocation is

performed by operator delete, the deallocation analogue of operator new. (For

 

details, see the sidebar on placement new and placement delete.)

 

It is this relationship between new expressions and operator delete affects

us in our attempt to automate the counting of object instantiations.

Counting Objects

 

In all likelihood, your solution to the object-counting problem involved the

 

development of an object-counting class. Your class probably looks remarkably

 

like, perhaps even exactly like, the Widget class I showed earlier:

 

// see below for a discussion of why

// this isn't quite right

 

class Counter { 

public:         

    Counter() { ++count; }

    Counter(const Counter&) { ++count; }

    ~Counter() { --count; }

    static size_t howMany()

        { return count; }

 

private:

    static size_t count;

};

// This still goes in an

// implementation file

size_t Counter::count = 0;

 

The idea here is that authors of classes that need to count the number of

objects in existence simply use Counter to take care of the bookkeeping.

There are two obvious ways to do this. One way is to define a Counter object

 

as a class data member, as in:

 

// embed a Counter to count objects

class Widget {

public:

    .....  // all the usual public

           // Widget stuff

    static size_t howMany()

    { return Counter::howMany(); }

private:

    .....  // all the usual private

           // Widget stuff

    Counter c;

};

 

The other way is to declare Counter as a base class, as in:

 

// inherit from Counter to count objects

class Widget: public Counter {

    .....  // all the usual public

           // Widget stuff

private:

    .....  // all the usual private

           // Widget stuff

};

 

Both approaches have advantages and disadvantages. But before we examine

them, we need to observe that neither approach will work in its current form.

 

The problem has to do with the static object count inside Counter. There's

only one such object, but we need one for each class using Counter. For

example, if we want to count both Widgets and ABCDs, we need two static

size_t objects, not one. Making Counter::count nonstatic doesn't solve the

problem, because we need one counter per class, not one counter per object.

 

We can get the behavior we want by employing one of the best-known but

oddest-named tricks in all of C++: we turn Counter into a template, and each

 

class using Counter instantiates the template with itself as the template

argument.

 

Let me say that again. Counter becomes a template:

 

template

class Counter {

public:

    Counter() { ++count; }

    Counter(const Counter&) { ++count; }

    ~Counter() { --count; }

 

    static size_t howMany()

    { return count; }

 

private:

    static size_t count;

};

 

template

size_t

Counter::count = 0; // this now can go in header

 

The first Widget implementation choice now looks like:

 

// embed a Counter to count objects

class Widget {

public:

    .....

    static size_t howMany()

    {return Counter::howMany();}

private:

    .....

    Counter c;

};

 

And the second choice now looks like:

 

// inherit from Counter to count objects

class Widget: public Counter {   

    .....

};

 

Notice how in both cases we replace Counter with Counter. As I said

earlier, each class using Counter instantiates the template with itself as

the argument.

 

The tactic of a class instantiating a template for its own use by passing

itself as the template argument was first publicized by Jim Coplien. He

showed that it's used in many languages (not just C++) and he called it "a

curiously recurring template pattern" [1]. I don't think Jim intended it, but

 

his description of the pattern has pretty much become its name. That's too

bad, because pattern names are important, and this one fails to convey

information about what it does or how it's used.

 

The naming of patterns is as much art as anything else, and I'm not very good

 

at it, but I'd probably call this pattern something like "Do It For Me."

Basically, each class generated from Counter provides a service (it counts

how many objects exist) for the class requesting the Counter instantiation.

So the class Counter counts Widgets, and the class Counter

counts ABCDs.

 

Now that Counter is a template, both the embedding design and the inheritance

 

design will work, so we're in a position to evaluate their comparative

strengths and weaknesses. One of our design criteria was that object-counting

 

functionality should be easy for clients to obtain, and the code above makes

 

clear that the inheritance-based design is easier than the embedding-based

design. That's because the former requires only the mentioning of Counter as

 

a base class, whereas the latter requires that a Counter data member be

defined and that howMany be reimplemented by clients to invoke Counter's

howMany [2]. That's not a lot of additional work (client howManys are simple

 

inline functions), but having to do one thing is easier than having to do

two. So let's first turn our attention to the design employing inheritance.

Using Public Inheritance

 

The design based on inheritance works because C++ guarantees that each time a

 

derived class object is constructed or destroyed, its base class part will

also be constructed first and destroyed last. Making Counter a base class

thus ensures that a Counter constructor or destructor will be called each

time a class inheriting from it has an object created or destroyed.

 

Any time the subject of base classes comes up, however, so does the subject

of virtual destructors. Should Counter have one? Well-established principles

 

of object-oriented design for C++ dictate that it should. If it has no

virtual destructor, deletion of a derived class object via a base class

pointer yields undefined (and typically undesirable) results:

 

class Widget: public Counter

{ ... };

Counter *pw =

    new Widget;  // get base class ptr

                 // to derived class object   

......

delete pw; // yields undefined results

           // if the base class lacks

           // a virtual destructor

 

Such behavior would violate our criterion that our object-counting design be

 

essentially foolproof, because there's nothing unreasonable about the code

above. That's a powerful argument for giving Counter a virtual destructor.

 

Another criterion, however, was maximal efficiency (imposition of no

unnecessary speed or space penalty for counting objects), and now we're in

trouble. We're in trouble because the presence of a virtual destructor (or

any virtual function) in Counter means each object of type Counter (or a

class derived from Counter) will contain a (hidden) virtual pointer, and this

 

will increase the size of such objects if they don't already support virtual

 

functions [3]. That is, if Widget itself contains no virtual functions,

objects of type Widget would increase in size if Widget started inheriting

from Counter. We don't want that.

 

The only way to avoid it is to find a way to prevent clients from deleting

derived class objects via base class pointers. It seems that a reasonable way

 

to achieve this is to declare operator delete private in Counter:

 

template

class Counter {

public:

    .....

private:

    void operator delete(void*);

    .....

};

 

Now the delete expression won't compile:

 

class Widget: public Counter { ... };

Counter *pw = new Widget;  ......

delete pw; // Error. Can't call private

// operator delete

 

Unfortunately — and this is the really interesting part — the new expression

 

shouldn't compile either!

 

Counter *pw =

    new Widget;  // this should not

                 // compile because

                 // operator delete is

                 // private

 

Remember from my earlier discussion of new, delete, and exceptions that C++'s

 

runtime system is responsible for deallocating memory allocated by operator

new if the subsequent constructor invocation fails. Recall also that operator

 

delete is the function called to perform the deallocation. But we've declared

 

operator delete private in Counter, which makes it invalid to create objects

 

on the heap via new!

 

Yes, this is counterintuitive, and don't be surprised if your compilers don't

 

yet support this rule, but the behavior I've described is correct.

Furthermore, there's no other obvious way to prevent deletion of derived

class objects via Counter* pointers, and we've already rejected the notion of

 

a virtual destructor in Counter. So I say we abandon this design and turn our

 

attention to using a Counter data member instead.

Using a Data Member

 

We've already seen that the design based on a Counter data member has one

drawback: clients must both define a Counter data member and write an inline

 

version of howMany that calls the Counter's howMany function. That's

marginally more work than we'd like to impose on clients, but it's hardly

unmanageable. There is another drawback, however. The addition of a Counter

data member to a class will often increase the size of objects of that class

 

type.

 

At first blush, this is hardly a revelation. After all, how surprising is it

 

that adding a data member to a class makes objects of that type bigger? But

blush again. Look at the definition of Counter:

 

template

class Counter {

public:

    Counter();

    Counter(const Counter&);

    ~Counter();

 

    static size_t howMany();

private:

    static size_t count;

};

 

Notice how it has no nonstatic data members. That means each object of type

Counter contains nothing. Might we hope that objects of type Counter have

size zero? We might, but it would do us no good. C++ is quite clear on this

point. All objects have a size of at least one byte, even objects with no

nonstatic data members. By definition, sizeof will yield some positive number

 

for each class instantiated from the Counter template. So each client class

containing a Counter object will contain more data than it would if it didn't

 

contain the Counter.

 

(Interestingly, this does not imply that the size of a class without a

Counter will necessarily be bigger than the size of the same class containing

 

a Counter. That's because alignment restrictions can enter into the matter.

For example, if Widget is a class containing two bytes of data but that's

required to be four-byte aligned, each object of type Widget will contain two

 

bytes of padding, and sizeof(Widget) will return 4. If, as is common,

compilers satisfy the requirement that no objects have zero size by inserting

 

a char into Counter, it's likely that sizeof(Widget) will still yield

 

4 even if Widget contains a Counter object. That object will simply

take the place of one of the bytes of padding that Widget already contained.

 

This is not a terribly common scenario, however, and we certainly can't plan

 

on it when designing a way to package object-counting capabilities.)

 

I'm writing this at the very beginning of the Christmas season. (It is in

fact Thanksgiving Day, which gives you some idea of how I celebrate major

holidays...) Already I'm in a Bah Humbug mood. All I want to do is count

objects, and I don't want to haul along any extra baggage to do it. There has

 

got to be a way.

Using Private Inheritance

 

Look again at the inheritance-based code that led to the need to consider a

virtual destructor in Counter:

 

class Widget: public Counter

{ ... };

Counter *pw = new Widget;           

......

delete

pw;  // yields undefined results

     // if Counter lacks a virtual

     // destructor

 

Earlier we tried to prevent this sequence of operations by preventing the

delete expression from compiling, but we discovered that that also prohibited

 

the new expression from compiling. But there is something else we can

prohibit. We can prohibit the implicit conversion from a Widget* pointer

(which is what new returns) to a Counter* pointer. In other words, we

 

can prevent inheritance-based pointer conversions. All we have to do is

replace the use of public inheritance with private inheritance:

 

class Widget: private Counter

{ ... };

Counter *pw =

    new Widget;  // error! no implicit

                 // conversion from

                 // Widget* to

                 // Counter*

 

Furthermore, we're likely to find that the use of Counter as a base class

does not increase the size of Widget compared to Widget's stand-alone size.

Yes, I know I just finished telling you that no class has zero size, but —

well, that's not really what I said. What I said was that no objects have

zero size. The C++ Standard makes clear that the base-class part of a more

derived object may have zero size. In fact many compilers implement what has

 

come to be known as the empty base optimization [4].

 

Thus, if a Widget contains a Counter, the size of the Widget must increase.

The Counter data member is an object in its own right, hence it must have

nonzero size. But if Widget inherits from Counter, compilers are allowed to

keep Widget the same size it was before. This suggests an interesting rule of

 

thumb for designs where space is tight and empty classes are involved: prefer

 

private inheritance to containment when both will do.

 

This last design is nearly perfect. It fulfills the efficiency criterion,

provided your compilers implement the empty base optimization, because

inheriting from Counter adds no per-object data to the inheriting class, and

 

all Counter member functions are inline. It fulfills the foolproof criterion,

 

because count manipulations are handled automatically by Counter member

functions, those functions are automatically called by C++, and the use of

private inheritance prevents implicit conversions that would allow

derived-class objects to be manipulated as if they were base-class objects.

(Okay, it's not totally foolproof: Widget's author might foolishly

instantiate Counter with a type other than Widget, i.e., Widget could be made

 

to inherit from Counter. I choose to ignore this possibility.)

 

The design is certainly easy for clients to use, but some may grumble that it

 

could be easier. The use of private inheritance means that howMany will

become private in inheriting classes, so such classes must include a using

declaration to make howMany public to their clients:

 

class Widget: private Counter {

public:

    // make howMany public

    using Counter::howMany;

 

    ..... // rest of Widget is unchanged

};

 

class ABCD: private Counter {

public:

    // make howMany public

    using Counter::howMany;

 

    ..... // rest of ABCD is unchanged

};

 

For compilers not supporting namespaces, the same thing is accomplished by

replacing the using declaration with the older (now deprecated) access

declaration:

 

class Widget: private Counter {

public:

    // make howMany public

    Counter::howMany;

 

    .....  // rest of Widget is unchanged

};

 

Hence, clients who want to count objects and who want to make that count

available (as part of their class's interface) to their clients must do two

things: declare Counter as a base class and make howMany accessible [5].

 

The use of inheritance does, however, lead to two conditions that are worth

noting. The first is ambiguity. Suppose we want to count Widgets, and we want

 

to make the count available for general use. As shown above, we have Widget

inherit from Counter and we make howMany public in Widget. Now

suppose we have a class SpecialWidget publicly inherit from Widget and we

want to offer SpecialWidget clients the same functionality Widget clients

enjoy. No problem, we just have SpecialWidget inherit from

Counter.

 

But here is the ambiguity problem. Which howMany should be made available by

 

SpecialWidget, the one it inherits from Widget or the one it inherits from

Counter? The one we want, naturally, is the one from

Counter, but there's no way to say that without actually

writing SpecialWidget::howMany. Fortunately, it's a simple inline function:

 

class SpecialWidget: public Widget,

    private Counter {

public:

    .....

    static size_t howMany()

    { return Counter::howMany(); }

    .....

};

 

The second observation about our use of inheritance to count objects is that

 

the value returned from Widget::howMany includes not just the number of

Widget objects, it includes also objects of classes derived from Widget. If

the only class derived from Widget is SpecialWidget and there are five

stand-alone Widget objects and three stand-alone SpecialWidgets,

Widget::howMany will return eight. After all, construction of each

SpecialWidget also entails construction of the base Widget part.

Summary

 

The following points are really all you need to remember:

 

    * Automating the counting of objects isn't hard, but it's not completely

 

straightforward, either. Use of the "Do It For Me" pattern (Coplien's

"curiously recurring template pattern") makes it possible to generate the

correct number of counters. The use of private inheritance makes it possible

 

to offer object-counting capabilities without increasing object sizes.

    * When clients have a choice between inheriting from an empty class or

containing an object of such a class as a data member, inheritance is

preferable, because it allows for more compact objects.

    * Because C++ endeavors to avoid memory leaks when construction fails for

 

heap objects, code that requires access to operator new generally requires

access to the corresponding operator delete too.

    * The Counter class template doesn't care whether you inherit from it or

 

you contain an object of its type. It looks the same regardless. Hence,

clients can freely choose inheritance or containment, even using different

strategies in different parts of a single application or library. o

 

Notes and References

 

[1] James O. Coplien. "The Column Without a Name: A Curiously Recurring

Template Pattern," C++ Report, February 1995.

 

[2] An alternative is to omit Widget::howMany and make clients call

Counter::howMany directly. For the purposes of this article, however,

 

we'll assume we want howMany to be part of the Widget interface.

 

[3] Scott Meyers. More Effective C++ (Addison-Wesley, 1996), pp. 113-122.

 

[4] Nathan Myers. "The Empty Member C++ Optimization," Dr. Dobb's Journal,

August 1997. Also available at http://www.cantrip.org/emptyopt.html.

 

[5] Simple variations on this design make it possible for Widget to use

Counter to count objects without making the count available to Widget

 

clients, not even by calling Counter::howMany directly. Exercise for

 

the reader with too much free time: come up with one or more such variations.

 

Further Reading

 

To learn more about the details of new and delete, read the columns by Dan

Saks on the topic (CUJ January - July 1997), or Item 8 in my More Effective

C++ (Addison-Wesley, 1996). For a broader examination of the object-counting

 

problem, including how to limit the number of instantiations of a class,

consult Item 26 of More Effective C++.

Acknowledgments

 

Mark Rodgers, Damien Watkins, Marco Dalla Gasperina, and Bobby Schmidt

provided comments on drafts of this article. Their insights and suggestions

improved it in several ways.

 

Scott Meyers authored the best-selling Effective C++, Second Edition and More

 

Effective C++ (both published by Addison Wesley). Find out more about him,

his books, his services, and his dog at http://www.aristeia.com.