@Pixelchemist已经给出了一个全面的答案。但作为一名网络开发人员，我可以给出一些建议。

通常我们使用tcp来开发一个web框架，所以通常我们有一个结构:

TcpServer listen port and register the socket to epoll or something
  -> TcpServer receive new connection 
    -> HttpConenction deal the data from the connection 
      -> HttpServer call Handler to deal with HttpConnection.
        -> Handler contain codes like save into database and fetch from db

我们可以按顺序开发框架，但它对只想关心Handler的用户并不友好。是时候使用回调了。

Mutiple Handler written by user
  -> register the handler as callback property of HttpServer
    -> register the related methods in HttpServer to HttpConnection
      -> register the relate methods in HttpConnection to TcpServer

所以用户只需要注册他们的处理程序到httpserver(通常用一些路径字符串作为键)，其他的事情是框架可以做的通用的。

你会发现我们可以把回调当成一种context，我们想委托给其他人为我们做。核心是我们不知道什么时候是调用函数的最佳时间，但我们委托的人知道。

C语言中也有回调的方法:函数指针

// Define a type for the callback signature,
// it is not necessary but makes life easier

// Function pointer called CallbackType that takes a float
// and returns an int
typedef int (*CallbackType)(float);

void DoWork(CallbackType callback)
{
  float variable = 0.0f;
  
  // Do calculations
  
  // Call the callback with the variable, and retrieve the
  // result
  int result = callback(variable);

  // Do something with the result
}

int SomeCallback(float variable)
{
  int result;

  // Interpret variable

  return result;
}

int main(int argc, char ** argv)
{
  // Pass in SomeCallback to the DoWork
  DoWork(&SomeCallback);
}

现在，如果你想将类方法作为回调函数传入，对这些函数指针的声明会有更复杂的声明，例如:

// Declaration:
typedef int (ClassName::*CallbackType)(float);

// This method performs work using an object instance
void DoWorkObject(CallbackType callback)
{
  // Class instance to invoke it through
  ClassName objectInstance;

  // Invocation
  int result = (objectInstance.*callback)(1.0f);
}

//This method performs work using an object pointer
void DoWorkPointer(CallbackType callback)
{
  // Class pointer to invoke it through
  ClassName * pointerInstance;

  // Invocation
  int result = (pointerInstance->*callback)(1.0f);
}

int main(int argc, char ** argv)
{
  // Pass in SomeCallback to the DoWork
  DoWorkObject(&ClassName::Method);
  DoWorkPointer(&ClassName::Method);
}

公认的答案是全面的，但与问题有关，我只是想在这里放一个简单的例子。我有一个很久以前写的代码。我想以有序的方式遍历树(左节点然后根节点然后右节点)，每当我到达一个节点，我希望能够调用任意函数，这样它就可以做任何事情。

void inorder_traversal(Node *p, void *out, void (*callback)(Node *in, void *out))
{
    if (p == NULL)
        return;
    inorder_traversal(p->left, out, callback);
    callback(p, out); // call callback function like this.
    inorder_traversal(p->right, out, callback);
}


// Function like bellow can be used in callback of inorder_traversal.
void foo(Node *t, void *out = NULL)
{
    // You can just leave the out variable and working with specific node of tree. like bellow.
    // cout << t->item;
    // Or
    // You can assign value to out variable like below
    // Mention that the type of out is void * so that you must firstly cast it to your proper out.
    *((int *)out) += 1;
}
// This function use inorder_travesal function to count the number of nodes existing in the tree.
void number_nodes(Node *t)
{
    int sum = 0;
    inorder_traversal(t, &sum, foo);
    cout << sum;
}

 int main()
{

    Node *root = NULL;
    // What These functions perform is inserting an integer into a Tree data-structure.
    root = insert_tree(root, 6);
    root = insert_tree(root, 3);
    root = insert_tree(root, 8);
    root = insert_tree(root, 7);
    root = insert_tree(root, 9);
    root = insert_tree(root, 10);
    number_nodes(root);
}

回调函数是传递给例程的方法，在某个时刻被传递给它的例程调用。

这对于开发可重用软件非常有用。例如，许多操作系统API(如Windows API)大量使用回调。

例如，如果你想处理文件夹中的文件，你可以用你自己的例程调用一个API函数，并且你的例程在指定的文件夹中对每个文件运行一次。这使得API非常灵活。

在c++中并没有明确的回调函数的概念。回调机制通常通过函数指针、函子对象或回调对象实现。程序员必须显式地设计和实现回调功能。

根据反馈进行编辑:

尽管这个答案收到了负面的反馈，但它并没有错。我会试着更好地解释我的观点。

C和c++拥有实现回调函数所需的一切。实现回调函数最常见和最简单的方法是将函数指针作为函数参数传递。

However, callback functions and function pointers are not synonymous. A function pointer is a language mechanism, while a callback function is a semantic concept. Function pointers are not the only way to implement a callback function - you can also use functors and even garden variety virtual functions. What makes a function call a callback is not the mechanism used to identify and call the function, but the context and semantics of the call. Saying something is a callback function implies a greater than normal separation between the calling function and the specific function being called, a looser conceptual coupling between the caller and the callee, with the caller having explicit control over what gets called. It is that fuzzy notion of looser conceptual coupling and caller-driven function selection that makes something a callback function, not the use of a function pointer.

例如，IFormatProvider的. net文档说“GetFormat是一个回调方法”，即使它只是一个普通的接口方法。我认为没有人会认为所有的虚方法调用都是回调函数。使GetFormat成为回调方法的不是它如何传递或调用的机制，而是调用方选择调用哪个对象的GetFormat方法的语义。

一些语言包含带有显式回调语义的特性，通常与事件和事件处理相关。例如，c#的事件类型的语法和语义显式地围绕回调的概念设计。Visual Basic有它的Handles子句，它显式地将一个方法声明为回调函数，同时抽象了委托或函数指针的概念。在这些情况下，回调的语义概念集成到语言本身中。

另一方面，C和c++几乎没有显式地嵌入回调函数的语义概念。机制在那里，集成的语义不在那里。你可以很好地实现回调函数，但是为了获得更复杂的、包含显式回调语义的东西，你必须在c++提供的基础上构建它，比如Qt对它们的信号和插槽所做的事情。

简而言之，c++拥有实现回调所需的一切，通常使用函数指针非常简单。它没有的是语义特定于回调的关键字和特性，比如raise、emit、Handles、event +=等。如果您来自具有这些类型的元素的语言，那么c++中的本机回调支持将感觉很中性。

请参阅上面的定义，其中声明将回调函数传递给其他函数，并在某个时刻调用它。

在c++中，让回调函数调用类方法是可取的。当您这样做时，您可以访问成员数据。如果你使用C语言定义回调函数，你必须将它指向一个静态成员函数。这不是很理想。

Here is how you can use callbacks in C++. Assume 4 files. A pair of .CPP/.H files for each class. Class C1 is the class with a method we want to callback. C2 calls back to C1's method. In this example the callback function takes 1 parameter which I added for the readers sake. The example doesn't show any objects being instantiated and used. One use case for this implementation is when you have one class that reads and stores data into temporary space and another that post processes the data. With a callback function, for every row of data read the callback can then process it. This technique cuts outs the overhead of the temporary space required. It is particularly useful for SQL queries that return a large amount of data which then has to be post-processed.

/////////////////////////////////////////////////////////////////////
// C1 H file

class C1
{
    public:
    C1() {};
    ~C1() {};
    void CALLBACK F1(int i);
};

/////////////////////////////////////////////////////////////////////
// C1 CPP file

void CALLBACK C1::F1(int i)
{
// Do stuff with C1, its methods and data, and even do stuff with the passed in parameter
}

/////////////////////////////////////////////////////////////////////
// C2 H File

class C1; // Forward declaration

class C2
{
    typedef void (CALLBACK C1::* pfnCallBack)(int i);
public:
    C2() {};
    ~C2() {};

    void Fn(C1 * pThat,pfnCallBack pFn);
};

/////////////////////////////////////////////////////////////////////
// C2 CPP File

void C2::Fn(C1 * pThat,pfnCallBack pFn)
{
    // Call a non-static method in C1
    int i = 1;
    (pThat->*pFn)(i);
}