/// <summary>
/// A static method that gets a unique word count for each of the words in a string
/// </summary>
/// <param name="str">The String that will be broken into a distinct word count</param>
/// <returns>A distinct word count in the form of a dictionary(word, count)</returns>
public static Dictionary <string, double> GetWordCount(this string str)
{
//Check to see that the user pased an actual string
//If they didn't return them an empty dictionary
if (String.IsNullOrEmpty(str))
{
return(new Dictionary <string, double>());
}
//Create the stemmer used to impliment Porters Algorithm for stemming strings
//The purpose of this is to take words like lovely and convert them to love,
//This helps attain more accurate results
var stemmer = new PorterStemmerAlgorithm.PorterStemmer();
//A dummy double used as the output for the Double.TryParse
//This eliminates numbers from the
Double num;
Regex rgx = new Regex("[^a-zA-Z0-9]");
str = rgx.Replace(str, " ");
//Split the words first removing _ characters
return((new Regex(@"\w(?<!\d)[\w'-]*")).Matches(str)
//Cast them to an enumerable of the matches.
.Cast <Match>()
//Convert the strings to lower, Stem them for consistency and select them.
.Select(m => stemmer.stemTerm(m.Value.ToLower()))
//Group Them by their text
.GroupBy(p => p)
//Select a new object where the Word is the text and the Count is the number of occurences of that word
.Select(g => new { Word = g.Key, Count = g.Count() })
//Order them by word (not necessary but I like order)
.OrderBy(p => p.Word)
//Remove all items that are found in the stop words dictionary, or are simply numbers
.Where(p => !StopWords.ContainsKey(p.Word) && !Double.TryParse(p.Word, out num))
//Convert this list to a dictionary where the word is the key and the number of its occurences is the value
.ToDictionary(p => p.Word, p => Convert.ToDouble(p.Count)));
}
