前/后函数调用实现 [英] Pre/Post function call implementation

问题描述

我想知道我是否可以以某种方式在C ++中进行函数的前/后调用.我有一个包装器类，其中包含许多函数，在每次包装器函数调用之后，我都应该调用另一个始终相同的函数.

I was wondering if I could do pre/post function call in C++ somehow. I have a wrapper class with lots of functions, and after every wrapper function call I should call another always the same function.

所以我不想将postFunction()调用放到每个这样的函数中:

So I do not want to put that postFunction() call to every single one of the functions like this:

class Foo {
    f1();
    f2();
    f3();
    .
    .
    .
    fn();
}

void Foo::f1() {
    ::f1();
    postFunction();
}

void Foo::f2() {
    ::f2();
    postFunction();
}

etc.

相反，当我调用某些Foo的成员函数时，我希望postFunction调用自动进行.是否可以?这将有助于维护.

Instead I want that postFunction call to come automatically when I call some Foo's member function. Is it possible? It would help maintenance..

推荐答案

RAII ！邓顿敦！

struct f1 {
  f1(Foo& foo) : foo(foo) {} // pre-function, if you need it
  void operator()(){} // main function
  ~f1() {} // post-function

private:
  Foo& foo;
}

然后，您只需确保每次希望调用该函数时都创建一个新的临时f1对象.重复使用它显然意味着pre/post函数不会每次都被调用.

Then you just have to make sure to create a new temporary f1 object every time you wish to call the function. Reusing it will obviously mean the pre/post functions don't get called every time.

甚至可以像这样包装它:

Could even wrap it like this:

void call_f1(Foo& foo) {
  f1(foo)(); // calls constructor (pre), operator() (the function itself) and destructor (post)
}

您可能会尝试其他方式来构造它，但总的来说，请看是否无法让构造函数/析构函数为您完成繁重的工作.

You might experiment a bit with other ways of structuring it, but in general, see if you can't get constructors/destructors to do the heavy lifting for you.

罗马M的方法也可能是一个好主意.编写一个通用包装器，该包装器将函子或函数指针作为其参数.这样，它可以在调用其参数之前和之后调用pre/post函数

Roman M's approach might be a good idea as well. Write one generic wrapper, which takes a functor or function pointer as its argument. That way, it can call pre/post functions before and after calling its argument

这篇关于前/后函数调用实现的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！