questions - smart pointers in c++ tutorial point




What is a smart pointer and when should I use one? (10)

What is a smart pointer and when should I use one?


Smart pointer is a pointer-like type with some additional functionality, e.g. automatic memory deallocation, reference counting etc.

Small intro is available on page Smart Pointers - What, Why, Which?.

One of the simple smart-pointer type is std::auto_ptr (chapter 20.4.5 of C++ standard), which allows to deallocate memory automatically when it out of scope and which is more robust than simple pointer usage when exceptions are thrown, although less flexible.

Another convenient type is boost::shared_ptr which implements reference counting and automatically deallocates memory when no references to object remains. This helps avoiding memory leaks and is easy to use to implement RAII.

Subject is covered in depth in book "C++ Templates: The Complete Guide" by David Vandevoorde, Nicolai M. Josuttis, chapter Chapter 20. Smart Pointers. Some topics covered:


http://en.wikipedia.org/wiki/Smart_pointer

In computer science, a smart pointer is an abstract data type that simulates a pointer while providing additional features, such as automatic garbage collection or bounds checking. These additional features are intended to reduce bugs caused by the misuse of pointers while retaining efficiency. Smart pointers typically keep track of the objects that point to them for the purpose of memory management. The misuse of pointers is a major source of bugs: the constant allocation, deallocation and referencing that must be performed by a program written using pointers makes it very likely that some memory leaks will occur. Smart pointers try to prevent memory leaks by making the resource deallocation automatic: when the pointer to an object (or the last in a series of pointers) is destroyed, for example because it goes out of scope, the pointed object is destroyed too.


A smart pointer is a class, a wrapper of a normal pointer. Unlike normal pointers, smart point’s life circle is based on a reference count (how many time the smart pointer object is assigned). So whenever a smart pointer is assigned to another one, the internal reference count plus plus. And whenever the object goes out of scope, the reference count minus minus.

Automatic pointer, though looks similar, is totally different from smart pointer. It is a convenient class that deallocates the resource whenever an automatic pointer object goes out of variable scope. To some extent, it makes a pointer (to dynamically allocated memory) works similar to a stack variable (statically allocated in compiling time).


A smart pointer is an object that acts like a pointer, but additionally provides control on construction, destruction, copying, moving and dereferencing.

One can implement one's own smart pointer, but many libraries also provide smart pointer implementations each with different advantages and drawbacks.

For example, Boost provides the following smart pointer implementations:

  • shared_ptr<T> is a pointer to T using a reference count to determine when the object is no longer needed.
  • scoped_ptr<T> is a pointer automatically deleted when it goes out of scope. No assignment is possible.
  • intrusive_ptr<T> is another reference counting pointer. It provides better performance than shared_ptr, but requires the type T to provide its own reference counting mechanism.
  • weak_ptr<T> is a weak pointer, working in conjunction with shared_ptr to avoid circular references.
  • shared_array<T> is like shared_ptr, but for arrays of T.
  • scoped_array<T> is like scoped_ptr, but for arrays of T.

These are just one linear descriptions of each and can be used as per need, for further detail and examples one can look at the documentation of Boost.

Additionally, the C++ standard library provides three smart pointers; std::unique_ptr for unique ownership, std::shared_ptr for shared ownership and std::weak_ptr. std::auto_ptr existed in C++03 but is now deprecated.


Definitions provided by Chris, Sergdev and Llyod are correct. I prefer a simpler definition though, just to keep my life simple: A smart pointer is simply a class that overloads the -> and * operators. Which means that your object semantically looks like a pointer but you can make it do way cooler things, including reference counting, automatic destruction etc. shared_ptr and auto_ptr are sufficient in most cases, but come along with their own set of small idiosyncrasies.


Here is the Link for similar answers : http://sickprogrammersarea.blogspot.in/2014/03/technical-interview-questions-on-c_6.html

A smart pointer is an object that acts, looks and feels like a normal pointer but offers more functionality. In C++, smart pointers are implemented as template classes that encapsulate a pointer and override standard pointer operators. They have a number of advantages over regular pointers. They are guaranteed to be initialized as either null pointers or pointers to a heap object. Indirection through a null pointer is checked. No delete is ever necessary. Objects are automatically freed when the last pointer to them has gone away. One significant problem with these smart pointers is that unlike regular pointers, they don't respect inheritance. Smart pointers are unattractive for polymorphic code. Given below is an example for the implementation of smart pointers.

Example:

template <class X>
class smart_pointer
{
          public:
               smart_pointer();                          // makes a null pointer
               smart_pointer(const X& x)            // makes pointer to copy of x

               X& operator *( );
               const X& operator*( ) const;
               X* operator->() const;

               smart_pointer(const smart_pointer <X> &);
               const smart_pointer <X> & operator =(const smart_pointer<X>&);
               ~smart_pointer();
          private:
               //...
};

This class implement a smart pointer to an object of type X. The object itself is located on the heap. Here is how to use it:

smart_pointer <employee> p= employee("Harris",1333);

Like other overloaded operators, p will behave like a regular pointer,

cout<<*p;
p->raise_salary(0.5);

I would like to add one more point to the above question, smart pointer std::shared_ptr doesn’t have subscript operator and doesn’t support ponter arithmetic, we can use get() to obtain a built in pointer.


Let T be a class in this tutorial Pointers in C++ can be divided into 3 types :

1) Raw pointers :

T a;  
T * _ptr = &a; 

They hold a memory address to a location in memory. Use with caution , as programs become complex hard to keep track.

Pointers with const data or address { Read backwards }

T a ; 
const T * ptr1 = &a ; 
T const * ptr1 = &a ;

Pointer to a data type T which is a const. Meaning you cannot change the data type using the pointer. ie *ptr1 = 19 ; will not work. But you can move the pointer. ie ptr1++ , ptr1-- ; etc will work. Read backwards : pointer to type T which is const

  T * const ptr2 ;

A const pointer to a data type T . Meaning you cannot move the pointer but you can change the value pointed to by the pointer. ie *ptr2 = 19 will work but ptr2++ ; ptr2-- etc will not work. Read backwards : const pointer to a type T

const T * const ptr3 ; 

A const pointer to a const data type T . Meaning you cannot either move the pointer nor can you change the data type pointer to be the pointer. ie . ptr3-- ; ptr3++ ; *ptr3 = 19; will not work

3) Smart Pointers : { #include <memory> }

Shared Pointer:

  T a ; 
     //shared_ptr<T> shptr(new T) ; not recommended but works 
     shared_ptr<T> shptr = make_shared<T>(); // faster + exception safe

     std::cout << shptr.use_count() ; // 1 //  gives the number of " 
things " pointing to it. 
     T * temp = shptr.get(); // gives a pointer to object

     // shared_pointer used like a regular pointer to call member functions
      shptr->memFn();
     (*shptr).memFn(); 

    //
     shptr.reset() ; // frees the object pointed to be the ptr 
     shptr = nullptr ; // frees the object 
     shptr = make_shared<T>() ; // frees the original object and points to new object

Implemented using reference counting to keep track of how many " things " point to the object pointed to by the pointer. When this count goes to 0 , the object is automatically deleted , ie objected is deleted when all the share_ptr pointing to the object goes out of scope. This gets rid of the headache of having to delete objects which you have allocated using new.

Weak Pointer : Helps deal with cyclic reference which arises when using Shared Pointer If you have two objects pointed to by two shared pointers and there is an internal shared pointer pointing to each others shared pointer then there will be a cyclic reference and the object will not be deleted when shared pointers go out of scope. To solve this , change the internal member from a shared_ptr to weak_ptr. Note : To access the element pointed to by a weak pointer use lock() , this returns a weak_ptr.

T a ; 
shared_ptr<T> shr = make_shared<T>() ; 
weak_ptr<T> wk = shr ; // initialize a weak_ptr from a shared_ptr 
wk.lock()->memFn() ; // use lock to get a shared_ptr 
//   ^^^ Can lead to exception if the shared ptr has gone out of scope
if(!wk.expired()) wk.lock()->memFn() ;
// Check if shared ptr has gone out of scope before access

See : When is std::weak_ptr useful?

Unique Pointer : Light weight smart pointer with exclusive ownership. Use when pointer points to unique objects without sharing the objects between the pointers.

unique_ptr<T> uptr(new T);
uptr->memFn(); 

//T * ptr = uptr.release(); // uptr becomes null and object is pointed to by ptr
uptr.reset() ; // deletes the object pointed to by uptr 

To change the object pointed to by the unique ptr , use move semantics

unique_ptr<T> uptr1(new T);
unique_ptr<T> uptr2(new T);
uptr2 = std::move(uptr1); 
// object pointed by uptr2 is deleted and 
// object pointed by uptr1 is pointed to by uptr2
// uptr1 becomes null 

References : They can essentially be though of as const pointers, ie a pointer which is const and cannot be moved with better syntax.

See : What are the differences between a pointer variable and a reference variable in C++?

r-value reference : reference to a temporary object   
l-value reference : reference to an object whose address can be obtained
const reference : reference to a data type which is const and cannot be modified 

Reference : https://www.youtube.com/channel/UCEOGtxYTB6vo6MQ-WQ9W_nQ Thanks to Andre for pointing out this question.


Smart Pointers are those where you don't have to worry about Memory De-Allocation, Resource Sharing and Transfer.

You can very well use these pointer in the similar way as any allocation works in Java. In java Garbage Collector does the trick, while in Smart Pointers, the trick is done by Destructors.


The existing answers are good but don't cover what to do when a smart pointer is not the (complete) answer to the problem you are trying to solve.

Among other things (explained well in other answers) using a smart pointer is a possible solution to How do we use a abstract class as a function return type? which has been marked as a duplicate of this question. However, the first question to ask if tempted to specify an abstract (or in fact, any) base class as a return type in C++ is "what do you really mean?". There is a good discussion (with further references) of idiomatic object oriented programming in C++ (and how this is different to other languages) in the documentation of the boost pointer container library. In summary, in C++ you have to think about ownership. Which smart pointers help you with, but are not the only solution, or always a complete solution (they don't give you polymorphic copy) and are not always a solution you want to expose in your interface (and a function return sounds an awful lot like an interface). It might be sufficient to return a reference, for example. But in all of these cases (smart pointer, pointer container or simply returning a reference) you have changed the return from a value to some form of reference. If you really needed copy you may need to add more boilerplate "idiom" or move beyond idiomatic (or otherwise) OOP in C++ to more generic polymorphism using libraries like Adobe Poly or Boost.TypeErasure.





c++-faq