Skip to main content

Covariance and Contravariance-General Discussion

If you have just started the exploration of .Net Framework 4.0, two terms namely Covariance and Contravariance might have been heard. The concept that these terms encapsulate are used by most developer almost daily, however there has never been any botheration about the terminologies.

Now, what actually these terms mean and how are these going to affect us as a developer, if we dive in to the details. The simple answer is it’s always good to know your tools before actually using them.

Enough philosophy, let’s get to the business.

Starting the discussion let me reiterate that in addition to Covariance and Contravariance, there is another terminology, Invariance. I’ll by start here by diving into the details of Invariance and then proceed further.

Invariance:

Invariance can be better understood by considering the types in .Net.>net has basically two type, value-types and reference-types. Value types (int, double etc) are invariant i.e. the types can’t be interchanged either at assignment or as passing values to the parameters.

For example if the expected value is an array of integers (int []), and we try to pass an array of doubles (double []), this will definitely result in error.

(Please Note: some types of higher order may be used in place of lower order while assigning values or passing parameters to a function)

Variant

A type must be a sub-type for another type to be a variant. Variant, as the name suggests is something that can vary. Let’s consider reference Types in .Net. All reference types ultimately derive from System.Object.

As we said in the beginning any subtype of a base type can act as variant.

How does a reference type behave as variant?

When a subtype is used in place of a base type (string in place of an object), an implicit conversion takes place and is legally acceptable.

Variant has two sub-categories

1. Covariance


2. Contra-variance

Covariance:

Covariant involves being able to use a more derived type as a substitute, e.g. a string in place of an object. Let’s consider an example to prove this.

string[] a = new string[1];

object[] b = a;

Response.Write(b[0]);



String is a sub-type of an object type and hence the code works fine with no error.

Covariance is applicable not only for assignments but also for return types from methods. It permits a method to have a more derived return type than what is defined in the delegate.

Contra-variance:

Contra-variance is just the opposite(rather juxtaposition image) of variance.In this, more specific can be used in place of derived types

Please visit the link to see a working example.


Following are the links that can be helpful in understanding the topic better.

Link1:

Link2:

Link3:


Hope this was Helpful.



Till next time we connect…Happy Coding.
Labels:

Comments

Post a Comment

Popular posts from this blog

Authentication using Social Networking portals(Facebook, Gmail, and Yahoo)

There are tons of sites, which offer sign on using the social networking site credentials (Facebook, gtalk, twitter and the list continues).It can be termed a “SINGLE SIGNON” and offers a lot of benefits compared to traditional database authentication approach. However, not storing user credentials in the DB imposes an additional risk. How to track who all logged into the system. Now the question is, which approach to follow. The best approach is to use inbuilt asp.net users for storing the logging info about user activities and using single signon technique for authentication. This article will explore the approach and provide the details of implementation using some third party libraries and customizing it to the requirements. The Authentication will be done using the following networking portals Yahoo  Gmail Facebook    Special thanks to my friend Sumit Khandelwal, for implementation of Facebook part (in fact he did it all!!) Except Facebook, all other ca...
2 comments
Read more

WPF Overview-Part-II

This post is in continuation to the last post. In this Post I’ll be exploring the Dependency Properties Dependency properties are similar to CLR properties with more advanced and complex features. The main difference between the CLR properties and dependency properties is, that the value of a normal .NET property is read directly from a private member in your class, whereas the value of a DependencyProperty is resolved dynamically when calling the GetValue() method that is inherited from DependencyObject . In case this description did not make sense, no need to worry, It will become clear by the time you reach end of this article. How the Value is Resolved in Dependency properties Every time a dependency property is accessed, it internally resolves the value by following the precedence from high to low. It checks if a local value is available, if not, check if a custom style trigger is active and I the similar manner continues until it finds a value. At last the default value is alwa...
Post a Comment
Read more

Asp.Net 4.0: An Overview-Part-III

This is the last post in the series which will explore the following new features of ASP.Net 4.0  Performance Monitoring for Individual Applications in a Single Worker Process Web.config File Refactoring Permanently Redirecting a Page Expanding the Range of Allowable URLs Performance Monitoring for Individual Applications in a Single Worker Process It is a common practice to host multiple ASP.NET applications in a single worker process, In order to increase the number of Web sites that can be hosted on a single server. This practice results in difficulties for server administrators to identify an individual application that is experiencing problems. ASP.NET 4 introduces new resource-monitoring functionality introduced by the CLR. To enable this functionality, following XML configuration snippet is added to the aspnet.config configuration file.(This file is located in the directory where the .NET Framework is installed ) <?xml version="1.0" encoding="UTF-8...
Post a Comment
Read more