Functional programming and C + + template element programming

Let's take a look at an example:

#include　<stdio.h>
template　<int　depth>
class　Fibnacci
{
public:
　　　　static　const　int　value　=　Fibnacci<depth-1>::value　+　Fibnacci<depth-2>::value;
};

template　<>
class　Fibnacci<0>
{
public:
　　　　static　const　int　value　=　0;
};

template　<>
class　Fibnacci<1>
{
public:
　　　　static　const　int　value　=　1;
};

template　<int　depth>
void　printFibnacci()
{
　　　　printFibnacci<depth-1>();
　　　　wprintf(L"%d\n",　Fibnacci<depth>::value);
}

template　<>
void　printFibnacci<0>()
{
　　　　wprintf(L"%d\n",　Fibnacci<0>::value);
}

int　main()
{
　　　　printFibnacci<8>();
　　　　return　0;
}

Fibnacci series, I believe it is a programmer can write out, the point is, this fibnacci sequence of calculation is complete at compile time! The following print is also true, and when you make a large parameter, the runtime will not change, but you will spend a long time in the compile phase.

If you've heard of some template metaprogramming, you'll know that "C + + templates are Turing complete." How is the template element Turing complete? The answer is that the template element is the same as the functional principle.

The essence of a template is defined and replaced, and the essence of a lambda function is also defined and replaced, where the substitution actually conforms to the lambda calculus Theory in Mathematics: