Routine in destination object

Routine in destination object

When the conversion routine is in the destination class, we use a one-argument constructor. However, things are complicated by the fact that the constructor in the destination class must be able to access the data in source class to perform the conversion. The polar data – radius & angle –are private, so we must provide special function to allow direct access it.

#include<iostream.h>

#include<conio.h>

#include <math.h>

class polar

{

      double radius,angle;          //x and y cordinate

   public:

      polar()        //constructor

      { radius=0.0;

            angle=0.0; }

    polar(double r, double a)

      {   radius=r; angle=a;   }

   void display()

   { cout<<"(" <<radius<<","<<angle<<")"; }

      double getr()

      {  return radius;   }

      double geta()

      {   return angle;   }

};

class rectangular

{

      double xco,yco;          //x and y cordinate

   public:

      rectangular()          //constructor

      {   xco=0.0; yco=0.0;   }

      rectangular(double x, double y)

      {   xco=x; yco=y;   }

      rectangular (polar p)

      {

             double r=p.getr();

             double a=p.geta();

             xco=r * cos(a);

             yco=a * sin(a);

      }

      void display()

      {    cout<<"(" <<xco<<","<<yco<<")";  }

};

void main()

{

   rectangular rec;         //rectangular using constructor1

   polar pol(10.0,0.785398);     //polar using constructor1

   rec=pol;        //convert polar to rectangular using conversion function

   clrscr();

   cout<<"\nPolar Values=";

   pol.display();

   cout<<"\nrectangular Values=";

   rec.display();

   getch();

}

Conversion between objects of Different Classes

Conversion between objects of Different Classes

The same two methods as same as conversion between basic types and user-defined types also apply to conversions between two user-defined types. That is, you can use a one-argument constructor, or you can use a conversion function. The choice depends on whether you want to put the conversion routine in the class specifier of the source object or of the destination object.

Routine in source object

When the conversion routine is in the source class, it is commonly accomplished using a conversion function. The two classes used in the next program are rectangular and polar class. The rectangular class is similar in that its objects are points in a two dimensional plane. However, it uses a rectangular coordinate system, where the location of each point is specified by x and y coordinates. Its member functions are similar to those for polar but are adapted to rectangular coordinates.

An example:

#include <math.h>

class rectangular

{

   double xco,yco;          //x and y cordinate

   public:

      rectangular()        //constructor

      {   xco=0.0; yco=0.0;   }

      rectangular(double x, double y)

      {   xco=x; yco=y;   }

      void display()

      {    cout<<"(" <<xco<<","<<yco<<")";     }

};

class polar

      {

      double radius,angle;          //x and y cordinate

     public:

      polar()       //constructor

      {   radius=0.0; angle=0.0;   }

      polar(double r, double a)

      {   radius=r; angle=a;   }

      void display()

      {    cout<<"(" <<radius<<","<<angle<<")";      }

      operator rectangular()     //casting operator function

            { double x=radius * cos(angle);

               double y=radius * sin(angle);

               return rectangular(x,y);   }  //temporary object

};

void main()

{

   rectangular rec;

   polar pol(10.0,0.785398);

   rec=pol;  //or rec=rectangular(pol);

                 // It is equivalent to pol.operator rectangular()

   cout<<"\nPolar Values=";

   pol.display();

   cout<<"\nRec Values=";

   rec.display();

   }