了解 Class 模板专业化示例的隐式实例化

Understanding an Implicit Instantiation of Class Template Specialization Example

所以标准的this section给出了这个例子(我的问题是内联的):

template<class T, class U>
struct Outer {
    template<class X, class Y> struct Inner;      // Where are X and Y from? Is this defining a struct?
    template<class Y> struct Inner<T, Y>;         // Is this defining a struct specialization? If so what is Y?
    template<class Y> struct Inner<T, Y> { };     // I feel like this is a redefinition of the line above, could it ever not be?
    template<class Y> struct Inner<U, Y> { };
};

诚然,我无法理解标准的链接部分,因为我无法理解这里发生的事情。我想我只是对所有飞来飞去的 template 感到困惑,但如果有人可以逐行告诉我发生了什么,那将非常有帮助。

template<class X, class Y> struct Inner;      
// Where are X and Y from? Is this defining a struct?

这是模板结构的声明InnerXY是模板参数。

template<class Y> struct Inner<T, Y>;         
// Is this defining a struct specialization? If so what is Y?

这是偏特化的声明,Y是模板参数。

template<class Y> struct Inner<T, Y> { };     
// I feel like this is a redefinition of the line above, could it ever not be?

这是偏特化的定义,这里不重定义。

然后将它们用作:

Outer<foo1, foo2>::Inner<foo3, foo4>* i1; 
// the primary template is used, with T=foo1, U=foo2, X=foo3, Y=foo4
// btw the primary template is not defined in this example

Outer<foo1, foo2>::Inner<foo1, foo3> i2; 
// the 1st partial specialization is used, with T=foo1, U=foo2, Y=foo3

Outer<foo1, foo2>::Inner<foo2, foo3> i3; 
// the 2st partial specialization is used, with T=foo1, U=foo2, Y=foo3

给你:

template<class T, class U>
struct Outer {
    // Forward-declares a struct with two template parameters
    template<class X, class Y> struct Inner;

    // Forward-declares a partial specialization of Inner
    // (Y is its only template parameter)
    template<class Y> struct Inner<T, Y>;

    // Defines the forward-declared partial specialization
    template<class Y> struct Inner<T, Y> { };

    // Declares and defines another partial specialization
    // (Y is its only template parameter)
    template<class Y> struct Inner<U, Y> { };
};