A Preview of What's New in C# 3.0

On the heels of the Visual Studio 2005 and C# 2.0 releases, Microsoft has given a sneak preview of what to expect in the version after the next: C# 3.0. Even though C# 3.0 is not even standardized yet, Microsoft provided a preview release at its Professional Developers Conference (PDC) in September so eager developers could try out some of the expected features. This article discusses the following major new enhancements expected in C# 3.0:

  • Implicitly typed local variables
  • Anonymous types
  • Extension methods
  • Object and collection initializers
  • Lambda expressions
  • Query expressions
  • Expression Trees

Implicitly Typed Local Variables

C# 3.0 introduces a new keyword called "var". Var allows you to declare a new variable, whose type is implicitly inferred from the expression used to initialize the variable. In other words, the following is valid syntax in C# 3.0:

var i = 1;

The preceding line initializes the variable i to value 1 and gives it the type of integer. Note that "i" is strongly typed to an integer—it is not an object or a VB6 variant, nor does it carry the overhead of an object or a variant.

To ensure the strongly typed nature of the variable that is declared with the var keyword, C# 3.0 requires that you put the assignment (initializer) on the same line as the declaration (declarator). Also, the initializer has to be an expression, not an object or collection initializer, and it cannot be null. If multiple declarators exist on the same variable, they must all evaluate to the same type at compile time.

Implicitly typed arrays, on the other hand, are possible using a slightly different syntax, as shown below:

var intArr = new[] {1,2,3,4} ;

The above line of code would end up declaring intArr as int[].

The var keyword allows you to refer to instances of anonymous types (described in the next section) and yet the instances are statically typed. So, when you create instances of a class that contain an arbitrary set of data, you don't need to predefine a class to both hold that structure and be able to hold that data in a statically typed variable.

Anonymous Types

C# 3.0 gives you the flexibility to create an instance of a class without having to write code for the class beforehand. So, you now can write code as shown below:

new {hair="black", skin="green", teethCount=64}

The preceding line of code, with the help of the "new" keyword, gives you a new type that has three properties: hair, skin, and teethCount. Behind the scenes, the C# compiler would create a class that looks as follows:

class __Anonymous1
   private string _hair = "black";
   private string _skin = "green";
   private int _teeth   = 64;

   public string hair {get { return _hair; } set { _hair = value; }}
   public string skin {get { return _skin; } set { _skin = value; }}
   public int teeth {get { return _teeth; } set { _teeth = value; }}

In fact, if another anonymous type that specified the same sequence of names and types were created, the compiler would be smart enough to create only a single anonymous type for both instances to use. Also, because the instances are, as you may have guessed, simply instances of the same class, they can be exchanged because the types are really the same.

Now you have a class, but you still need something to hold an instance of the above class. This is where the "var" keyword comes in handy; it lets you hold a statically typed instance of the above instance of the anonymous type. Here is a rather simple and easy use of an anonymous type:

var frankenstein = new {hair="black", skin="green", teethCount=64}

Extension Methods

Extension methods enable you to extend various types with additional static methods. However, they are quite limited and should be used as a last resort—only where instance methods are insufficient.

Extension methods can be declared only in static classes and are identified by the keyword "this" as a modifier on the first parameter of the method. The following is an example of a valid extension method:

public static int ToInt32(this string s)
   return Convert.ToInt32(s) ;

If the static class that contains the above method is imported using the "using" keyword, the ToInt32 method will appear in existing types (albeit in lower precedence to existing instance methods), and you will be able to compile and execute code that looks as follows:

string s = "1";
int i    = s.ToInt32();

This allows you to take advantage of the extensible nature of various built-in or defined types and add newer methods to them.

Object and Collection Initializers

C# 3.0 is expected to allow you to include an initializer that specifies the initial values of the members of a newly created object or collection. This enables you to combine declaration and initialization in one step.

For instance, if you defined a CoOrdinate class as follows:

public class CoOrdinate
   public int x ;
   public int y;

You then could declare and initialize a CoOrdinate object using an object initializer, like this:

var myCoOrd = new CoOrdinate{ x = 0, y= 0} ;

The above code may have made you raise your eyebrows and ask, "Why not just write the following:"

var myCoOrd = new CoOrdinate(0, 0) ;
Note: I never declared a constructor that accepted two parameters in my class. In fact, initializing the object using an object initializer essentially is equivalent to calling a parameterless (default) constructor of the CoOrdinate object and then assigning the relevant values.

Similarly, you should easily be able to give values to collections in a rather concise and compact manner in C# 3.0. For instance, the following C# 2.0 code:

List<string> animals = new List<string>();


Now can be shortened to simply:

List<string> animals = new List<string> {
   "monkey", "donkey", "cow", "dog", "cat" } ;

A Preview of What's New in C# 3.0

Lambda Expressions: The Espresso of Anonymous Methods

C# 1.x allowed you to write code blocks in methods, which you could invoke easily using delegates. Delegates are definitely useful, and they are used throughout the framework, but in many instances you had to declare a method or a class just to use one. Thus, to give you an easier and more concise way of writing code, C# 2.0 allowed you to replace standard calls to delegates with anonymous methods. The following code may have been written in .NET 1.1 or earlier:

class Program
   delegate void DemoDelegate();
   static void Main(string[] args)
      DemoDelegate myDelegate = new DemoDelegate(SayHi);
   void SayHi()
      Console.Writeline("Hiya!!") ;

In C# 2.0, using anonymous methods, you could rewrite the code as follows:

class Program
   delegate void DemoDelegate();
   static void Main(string[] args)
      DemoDelegate myDelegate = delegate()

Whereas anonymous methods are a step above method-based delegate invocation, lambda expressions allow you to write anonymous methods in a more concise, functional syntax.

You can write a lambda expression as a parameter list, followed by the => token, followed by an expression or statement block. The above code can now be replaced with the following code:

class Program
   delegate void DemoDelegate();
   static void Main(string[] args)
      DemoDelegate myDelegate = () => Console.WriteLine("Hiya!!") ;

Although Lambda expressions may appear to be simply a more concise way of writing anonymous methods, in reality they also are a functional superset of anonymous methods. Specifically, Lambda expressions offer the following additional functionality:

  • They permit parameter types to be inferred. Anonymous methods will require you to explicitly state each and every type.
  • They can hold either query expressions (described in the following section) or C# statements.
  • They can be treated as data using expression trees (described later). This cannot be done using Anonymous methods.

Query Expressions

Even though further enhancements may be introduced in the coming months as C# 3.0 matures, the new features described in the preceding sections make it a lot easier to work with data inside C# in general. This feature, also known as LINQ (Language Integrated Query), allows you to write SQL-like syntax in C#.

For instance, you may have a class that describes your data as follows:

public class CoOrdinate
   public int x ;
   public int y;

You now could easily declare the logical equivalent of a database table inside C# as follows:

// Use Object and collection initializers
List<CoOrdinate> coords = ... ;

And now that you have your data as a collection that implements IEnumerable<T>, you easily can query this data as follows:

var filteredCoords =
   from c in coords
   where x == 1
   select (c.x, c.y)

In the SQL-like syntax above, "from", "where", and "select" are query expressions that take advantage of C# 3.0 features such as anonymous types, extension methods, implicit typed local variables, and so forth. This way, you can leverage SQL-like syntax and work with disconnected data easily.

Each query expression is actually translated into a C#-like invocation behind the scenes. For instance, the following:

where x == 1

Translates to this:

coords.where(c => c.x == 1)

As you can see, the above looks an awful lot like a lambda expression and extension method. C# 3.0 has many other query expressions and rules that surround them.

Expression Trees

C# 3.0 includes a new type that allows expressions to be treated as data at runtime. This type, System.Expressions.Expression<T>, is simply an in-memory representation of a lambda expression. The end result is that your code can modify and inspect lambda expressions at runtime.

The following is an example of an expression tree:

Expression<DemoDelegate> filter = () => Console.WriteLine("Hiya!!") ;

With the above expression tree setup, you easily can inspect the contents of the tree by using various properties on the filter variable.

One to Grow On

C# 3.0 offers incredible new features that make your work as an application developer and architect a lot easier, and yet it remains a programming language that lends itself to stricter and cleaner architecture.

C# 3.0 is in its infancy right now and it will mature in the coming months, but given the sizable impact its changes will have on the surrounding .NET Framework, its recommended architecture, and design patterns, definitely keep your eye on it.

About the Author

Sahil Malik has worked for a number of top-notch clients in Microsoft technologies ranging from DOS to .NET. He is the author of Pro ADO.NET 2.0 and co-author of Pro ADO.NET with VB.NET 1.1. Sahil is currently also working on a multimedia series on ADO.NET 2.0 for Keystone Learning. For his community involvement, contributions, and speaking, he has also been awarded the Microsoft MVP award.


  • There are no comments yet. Be the first to comment!

Leave a Comment
  • Your email address will not be published. All fields are required.

Top White Papers and Webcasts

  • Today's competitive marketplace requires the organization to frequently release and deploy applications at the pace of user demands, with reduced cost, risk, and increased quality. This book defines the basics of application release and deployment, and provides best practices for implementation with resources for a deeper dive. Inside you will find: The business and technical drivers behind automated application release and deployment. Evaluation guides for application release and deployment solutions. …

  • With JRebel, developers get to see their code changes immediately, fine-tune their code with incremental changes, debug, explore and deploy their code with ease (both locally and remotely), and ultimately spend more time coding instead of waiting for the dreaded application redeploy to finish. Every time a developer tests a code change it takes minutes to build and deploy the application. JRebel keeps the app server running at all times, so testing is instantaneous and interactive.

Most Popular Programming Stories

More for Developers

Latest Developer Headlines

RSS Feeds