Just to add: Anonymous types are converted by the compiler to a real object. So the code will be changed to something equivalent of this (MUCH simplified, only to show that the compiler will create an actual class):
internal sealed class AnonymousClass1
{
internal string Text { get; set; }
internal int Count { get; set; }
internal string Link { get; set; }
}
And your code will then be changed to:
obj.DataSource = new AnonymousClass1[]
{
new AnonymousClass1 {Text = "Silverlight", Count = 10, Link="/Tags/Silverlight" },
new AnonymousClass1 {Text = "IIS 7", Count = 11, Link="http://iis.net" },
new AnonymousClass1 {Text = "IE 8", Count = 12, Link="/Tags/IE8" },
new AnonymousClass1 {Text = "C#", Count = 13, Link="/Tags/C#" },
new AnonymousClass1 {Text = "Azure", Count = 13, Link="?Tag=Azure" }
};
In one of my programs, I have code like this (simplified!):
var myObjects = new []{
new { Id = Guid.NewGuid(), Title = "Some Title", Description = string.Empty },
new { Id = Guid.NewGuid(), Title = "Another Title", Description = string.Empty },
new { Id = Guid.NewGuid(), Title = "Number 3", Description = "Better than No2, but not much" }
}
foreach(var myObject in myObjects) {
DoSomeThingWith(myObject.Title);
}
This works because it is just another class (I even get IntelliSense) behind the scenes. The benefit is obviously that I just saved myself from creating a class for this object. In my example, all objects need to look the same as well. (Obviously, doing this for any public members would be a bad idea as the compiler might change the name of the anonymous class if you add/remove some)