Boost C++ Application Development Cookbook - Second Edition

How to do it...

We are going to see how the size of a template type can be converted to a variable of some type, and how that variable can be used for deducing the right function overload.

1. Let's define our generic and optimized versions of the process_impl function:

#include <boost/mpl/int.hpp>  namespace detail {    // Generic implementation.    template <class T, class Tag>    T process_impl(const T& val, Tag /*ignore*/) {        // ...    }     // 1 byte optimized implementation.    template <class T>    T process_impl(const T& val, boost::mpl::int_<1> /*ignore*/) {        // ...    }     // 4 bytes optimized implementation.    template <class T>    T process_impl(const T& val, boost::mpl::int_<4> /*ignore*/) {        // ...    }     // 8 bytes optimized implementation.    template <class T> T process_impl(const ...