(How)
C^# You Are - 24 December 2006

1. How do you concatenate multiple strings in C#?  Answer

   The preferred way is to use either the plus (+) sign or the String.Concat method.  Under the hood the plus sign will map to String.Concat.  Only if absolutely necessary should you use String.Format.  For anything beyond simple string concatenation you should consider using StringBuilder as it is more efficient in memory usage.

2. How would you go about parsing out a list of names from a comma-separated string of values.?  Answer
   
   

   string[] values = stringValue.Split(',');

   
   
3. Given a list of names how would you go about combining them into a comma-separated string of values?  Answer
   
   

   string stringValue = String.Join(",", values);

   
   
4. How do you prevent other developers from creating instances of your class but still allow them to interact with instances of the class (i.e. a business object)?  Answer

   You should define the constructor(s) as internal to prevent others outside your assembly from creating instances of the class yet allow you to create instances yourself.  Note that a static class will not work in this case because you want to create instances inside the assembly.

5. How do you define a singleton in C#?  Answer
   
   

   public sealed class SingletonClass
   {
      private SingletonClass ( ) 
      { } 
   
      public static SingletonClass Instance
      {
         get { return m_Inst; }
      }
   
      private static SingletonClass m_Inst = new SingletonClass();
   }

   Note that the above code is thread safe because C# guarantees that static fields are initialized before first use.  Furthermore reference assignment is atomic so the creation and assignment should occur together.  There is a slight chance (maybe not given the memory model of .NET) that the instance could be created twice.  In this case you could resolve the issue by using a lock statement.

6. What is a nullable type (added in v2, not in relation to a reference type)?  Answer

   A nullable type (introduced in v2.0) is a value type that can contain a value or null.  A nullable type was introduced to make working with databases easier.  A nullable type is effectively a structure that contains a value type and a flag indicating if the value type has been set.  A nullable type is identified by the value type name followed by a question mark like so:  int?, float?  DateTime?
   
   (Updated 28 Dec 2006 w/ comments received from James Curran)

7. Why will the following code not compile?  Answer
   {
   public string GetPowerStatus ( )
   int nStatus = SystemInformation.PowerStatus.BatteryChargeStatus;
      switch(nStatus)
      {
         case BatteryChargeStatus.Charging : return "charging";
         case BatteryChargeStatus.Critical: return "critical";
         case BatteryChargeStatus.High: return "high";
         case BatteryChargeStatus.Low: return "low";
         case BatteryChargeStatus.NoSystemBattery: return "no battery";
         case BatteryChargeStatus.Unknown: return "unknown";
      };
   
      return "";
   }

   There is no implicit conversion from enumerations to integers (like in C++).  Therefore the assignment of BatteryChargeStatus to nStatus must be done through a type cast like so.

   int nStatus = (int)SystemInformation.PowerStatus.BatteryChargeStatus;

   The above code sort of solves the problem but you'll also end up having to cast the case labels as well. The best option is to change nStatus to be of type BatteryChargeStatus and allow the compiler to enforce the type checking.

   BatteryChartStatus nStatus = SystemInformation.PowerStatus.BatteryChargeStatus;

   
   (Updated 28 Dec 2006 w/ comments received from James Curran)
   
8. What is constructor chaining?  Answer

   Constructor chaining is the, efficient, process of calling an implementation of a constructor from another constructor.  This allows class developers to consolidate constructor logic into a single constructor rather than spreading the construction code across several constructors or creating a private method.  Here is an example.

   public class NamedDataObject
   {
      public NamedDataObject ( ) : this(0, null)
      { }
   
      public NamedDataObject ( int id ) : this(id, null)
      { }
   
      public NamedDataObject ( string name ) : this(0, name)
      { }
   
      public NamedDataObject ( int id, string name )
      {
         //Initialization code
      }
   } 

   
   
9. What is a finalizer?  Answer

   A finalizer is a method that is called after a reference type has been marked for garbage collection but before it is actually collected.  It works similar to a destructor in C++ and uses the same syntax.  However a finalizer will be called at some arbitrary point in time.  A finalizer may be called after the reference type fields within the class are already released.  Therefore a finalizer can only safely access value types and unmanaged resources during finalization.

   Finalizers should be avoided whenever possible as they keep objects around longer than they are needed.  Finalizers are also not guranteed to be called in all cases.  Finalizers that throw exceptions, enter an infinite loop or cause any number of other problems can prevent other finalizers from running at all.

   Only implement a finalizer if you are dealing with unmanaged resources.  You should also implement the IDisposable interface.

10. What is a critical finalizer?  Answer

    A critical finalizer is a finalizer that is guaranteed to be called.  A finalizer is actually wrapped in a CER (critical execution region).  Therefore all the logic of the finalizer is JITted in advance and ready to run.  That means the code can run even when a stack overflow or out of memory exception occurs.  Critical finalizers should only be used when dealing with the most critical of system resources.