Bruce Eckel's Thinking in C++, 2nd Ed Contents | Prev | Next

Controlling template instantiation

Normally templates are not instantiated until they are needed. For function templates this just means the point at which you call the function, but for class templates it’s more granular than that: each individual member function of the template is not instantiated until the first point of use. This means that only the member functions you actually use will be instantiated, which is quite important since it allows greater freedom in what the template can be used with. For example: 

//: C19:DelayedInstantiation.cpp
// Member functions of class templates are not
// instantiated until they're needed.

class X {
public:
  void f() {}
};

class Y {
public:
  void g() {}
};

template <typename T> class Z {
  T t;
public:
  void a() { t.f(); }
  void b() { t.g(); }
};

int main() {
  Z<X> zx;
  zx.a(); // Doesn't create Z<X>::b()
  Z<Y> zy;
  zy.b(); // Doesn't create Z<Y>::a()
} ///:~

Here, even though the template purports to use both f( ) and g( ) member functions of T, the fact that the program compiles shows you that it only generates Z<X>::a( ) when it is explicitly called for zx (if Z<X>::b( ) were also generated at the same time, a compile-time error message would be generated). Similarly, the call to zy.b( ) doesn’t generate Z<Y>::a( ). As a result, the Z template can be used with X and Y, whereas if all the member functions were generated when the class was first created it would significantly limit the use of many templates.

The inclusion vs. separation models

The export keyword

Contents | Prev | Next