2. (Using Libraries and Code)

Let us assume that we have a library of modules for three-dimensional geometry. One module computes the distance between two points in 3D space:

// Library adapter double Taximan3D(double x1, double y1, double z1, double x2, double y2, double z2)

{ // Distance between two points in 3d taxi space return std::abs(x1 - x2) + std::abs(y1 - y2) + std::abs(z1 - z2);
}
We reuse this code to create an algorithm to define the taximan’s distance in two dimensions.
We now create a function that is suitable for two-dimensional geometry:
std::function
Taximan2D = std::bind(Taximan3D, std::placeholders::_1, std::placeholders::_2,0.0, std::placeholders::_3, std::placeholders::_4,0.0);
We explain this cryptic code. It is an example of partial function application (see Section 3.2.4) in which a function with six arguments is morphed into one with four arguments.
In particular, the third and sixth arguments (that correspond to the z coordinates of the first and second points, respectively) are bound to the value 0.0.
We now use this function to create a new function object similar to the Strategy that we created in Section 3.8:
struct Taximan { // The scenic route(aka expensive way) from A to B.
Taximan() {}
double operator ()(const Point& p, const Point& p2) const {
// For readability double x1 = p.First(); double x2 = p2.First();
double y1 = p.Second(); double y2 = p2.Second();
double result = Taximan2D(x1, y1, x2, y2);
return result;
}
};
A test program is:
Taximan myTaxi;
Point pC(1.0, 1.0, myTaxi);
Point pD(2.0, 2.0, myTaxi);
std::cout << "Taximan C and D points: " << pC.Distance(pD) << std::endl;
std::cout << "Taximan D and C points: " << pD.Distance(pC) << std::endl;
Answer the following questions:

a) Write and compile the above code.

b) Compare the efficiency of this code that uses std::bind with (possibly) more straightforward code to compute the distance between two points using the two-dimensional taximan algorithm.