C# 3.0 Cookbook, 3rd Edition

11.1. Creating a Hash Code for a Data Type

Problem

You have created a class or structure that will be used as a key in a Hashtable or Dictionary<T,U>. You need to overload the GetHashCode method in order to return a good distribution of hash values (the Discussion section defines a good distribution of hash values). You also need to choose the best hash-code algorithm to use in the GetHashCode method of your object.

Solution

The following procedures implement hash-code algorithms and can be used to override the GetHashCode method. Included in the discussion of each method are the pros and cons of using it, as well as why you would want to use one instead of another.

In addition, it is desirable, for performance reasons, to use the return value of the GetHashCode method to determine whether the data contained within two objects is equal. Calling GetHashCode to return a hash value of two objects and comparing their hash values can be faster than calling the default implementation of Equals on the Object type, which individually tests the equality of all pertinent data within two objects. In fact, some developers even opt to compare hash-code values returned from GetHashCode within their overloaded Equals method. Using a custom implementation of the Equals method in this fashion is faster than the default implementation of the Object.Equals method.

The simple hash

This hash accepts a variable number of integer values and XORs each value to obtain a hash code. This is a well-performing and ...