SIGN UP MEMBER LOGIN:    
ARTICLE

Downcasting in C#

Posted by Scott Lysle Articles | Learn .NET March 01, 2009
This article describes a simple approach to downcasting in C#; downcasting merely refers to the process of casting an object of a base class type to a derived class type. Upcasting is legal in C# as the process there is to convert an object of a derived class type into an object of its base class type. In spite of the general illegality of downcasting you may find that when working with generics it is sometimes handy to do it anyway.
Reader Level:
Download Files:
 

Introduction

This article describes a simple approach to downcasting in C#; downcasting merely refers to the process of casting an object of a base class type to a derived class type. Upcasting is legal in C# as the process there is to convert an object of a derived class type into an object of its base class type. In spite of the general illegality of downcasting you may find that when working with generics it is sometimes handy to do it anyway.



Figure 1: Test Application in Use

The Solution

The example solution provided is in the form of a console mode application; the entire solution is contained within the program.cs file.
 
The Code

The Program.cs file contains two classes used to provide an example of a base class and a derived class. They are the person class (the base) and the lawyer class (derived from person). The person contains some basic information about a person whilst the lawyer class contains a couple of a additional properties:

    public class Person
    {
        public string FirstName { get; set; }
        public string MiddleName { get; set; }
        public string LastName { get; set; }
        public int Age { get; set; }
    }

    public class Lawyer : Person
    {
        public int NumberOfPeopleShafted { get; set; }
        public int NumberOfLiesTold { get; set; }
    }

What now makes this all interesting is if we wanted to use a single method to save either a person or a lawyer (I'm going out on a limb here in indicating that lawyers inherit from persons). To illustrate the point, I built a static class entitled, "PeopleHandler". The sole method contained in this class can accept either a person or one of person's derived class which is in this case only lawyers. Now, if we pass the following method a object of the person type, everything is fine and all properties are visible in the code. However, if we pass in object of the lawyer type (the derived class), the object is treated as a person rather than a lawyer and the other lawyer properties are invisible to the code even though the lawyer object passed to the "SavePerson" method contains both properties. Since downcasting is not permitted, it is necessary to cast the lawyer as a lawyer rather than as a person; unfortunately, that does not work.

If you tried this: "(Lawyer)pers", the code would show an error indicating that one cannot cast T as a lawyer. Well, in this case, we know full well that "pers" is really a lawyer and not a person so it would be really helpful if we could use it as a lawyer rather than as a person.

In the following bit of code (highlighted) you can see a solution to the dilemma and it is pretty easy, just convert the lawyer to an object and then cast is over to lawyer; that works. After doing that, you will find that you can access the additional properties of the lawyer without difficulty. To avoid any problems with that, I get the type of object passed into the method and use a switch statement to limit my attempts to read lawyer properties to lawyer objects only.

The code that I used is provided in the following:

    public static class PeopleHandler
    {
        public static void SavePerson<T>(T pers) where T : Person
        {

            Type tPers = pers.GetType();
            string[] arr = tPers.ToString().Split('.');
            string personType = arr[arr.Length - 1];

            int mAge = pers.Age;
            string mFirstName = pers.FirstName;
            string mMiddleName = pers.MiddleName;
            string mLastName = pers.LastName;

            switch (personType)
            {
                case "Person":
                    Console.WriteLine("Save Type: " + personType);
                    Console.WriteLine("Name: " + mFirstName + " " +
                    mMiddleName + " " + mLastName);
                    Console.WriteLine("Age: " + mAge.ToString());
                    break;
                case "Lawyer":
                    Console.WriteLine("Save Type: " + personType);
                    Console.WriteLine("Name: " + mFirstName + " " +
                    mMiddleName + " " + mLastName);
                    Console.WriteLine("Age: " + mAge.ToString());
                    Console.WriteLine("Number of Clients Shafted: " +
                    ((Lawyer)(Object)pers).NumberOfPeopleShafted);
                    Console.WriteLine("Number of Lies Told: " +
                    ((Lawyer)(Object)pers).NumberOfLiesTold);
                    break;
                default:
                    break;
            } 
        }
    }

The Program Class

The program class itself is used to test the People Handler's "SavePerson" method by passing it an object of type person and an object of type lawyer. The results of running the executable will display information about both the person and lawyer objects passed to the method.

    
public class Program
    {
        static void Main(string[] args)
        {
            Person MacDuff = new Person();
            MacDuff.Age = 44;
            MacDuff.FirstName = "Alexander";
            MacDuff.MiddleName = "Hercules";
            MacDuff.LastName = "MacDuff";

            PeopleHandler.SavePerson(MacDuff);

            Console.WriteLine(Environment.NewLine);

            Lawyer creep = new Lawyer();
            creep.FirstName = "Gordon";
            creep.MiddleName = "Bud";
            creep.LastName = "Miller";
            creep.Age = 50;
            creep.NumberOfLiesTold = 100000;
            creep.NumberOfPeopleShafted = 5000;

            PeopleHandler.SavePerson(creep);

            Console.Read();
        }
    }

Summary

This example provided an approach to downcasting; the article provides an example showing how downcasting might be useful when saving base and derived class objects using a common save method.

Login to add your contents and source code to this article
share this article :
post comment
 

But don't propagate bad code. Place a virtual Save() method on the Base Person class and have your Lawyer class override it to add its additional properties. public class Person { public string FirstName { get; set; } public string MiddleName { get; set; } public string LastName { get; set; } public int Age { get; set; } public virtual void Save() { Console.WriteLine("Name: " + mFirstName + " " + mMiddleName + " " + mLastName); Console.WriteLine("Age: " + mAge.ToString()); } } public class Lawyer : Person { public int NumberOfPeopleShafted { get; set; } public int NumberOfLiesTold { get; set; } public override void Save() { base.Save(); Console.WriteLine("Number of Clients Shafted: " + NumberOfPeopleShafted); Console.WriteLine("Number of Lies Told: " + NumberOfLiesTold); } } // Then you don't even need the PersonHandler class but just for grins: public static void SavePerson<T>(T pers) where T : Person { Type tPers = pers.GetType(); string[] arr = tPers.ToString().Split('.'); string personType = arr[arr.Length - 1]; switch (personType) { case "Person": Console.WriteLine("Save Type: " + personType); T.Save(); break; case "Lawyer": Console.WriteLine("Save Type: " + personType); T.Save(); } } Also, for dowcasting, don't do an explicit cast... use dynamic casting: Lawyer myLawyer = (MyPersonReference as Lawyer); if the cast fails then the myLawyer reference will be null.

Posted by Jarrod Feb 18, 2011

It is if the architect does not require a single save method on the base class that handles all of the derived classes. I put this out there due to rather large and unresolved number of blog entries I encountered from people looking for a way to do exactly this. This is not preferred way of doing business and if it were it would not be prevented so naturally you are correct in what you've said, but if you don't have the luxury of being correct, this works.

Posted by Scott Lysle Mar 02, 2009

Polymorphism is the object-oriented method of doing this. You should have one method per case. You've got this use issue that the textual name you're comparing against is decoupled from the type name. If you changed the name of your type then you'd have runtime errors. Using polymorphism gets you compile-time checking. Not to mention, what method that gets called is decided upon compilation and no string comparison would be needed.

Posted by Peter Ritchie Mar 02, 2009
Nevron Gauge for SharePoint
Become a Sponsor
PREMIUM SPONSORS
  • Finally – a virtual platform that delivers next-generation Windows Server 2008 Hyper-V virtualization technology from a managed hosting partner you can truly depend on. Visit www.maximumasp.com/max for a FREE 30 day trial. Hurry offer ends soon. Climb aboard the MaxV platform and take advantage of High Availability, Intelligent Monitoring, Recurrent Backups, and Scalability – with no hassle or hidden fees. As a managed hosting partner focused solely on Microsoft technologies since 2000, MaximumASP is uniquely qualified to provide the superior support that our business is built on. Unparalleled expertise with Microsoft technologies lead to working directly with Microsoft as first to offer IIS 7 and SQL 2008 betas in a hosted environment; partnering in the Go Live Program for Hyper-V; and product co-launches built on WS 2008 with Hyper-V technology.
    The leading .NET charting control now features PDF, Flash and Silverlight export, visualization of large datasets and more. Deliver true charting functionality to your BI, Scorecard, Presentation or Scientific apps. Download evaluation now.
Team Foundation Server Hosting
Become a Sponsor