/// <summary>
/// Build a WordCount object from a text file.
/// Source: http://stackoverflow.com/questions/2161895/reading-large-text-files-with-streams-in-c-sharp
/// </summary>
/// <param name="path"></param>
/// <returns></returns>
public static WordCount LoadFromFile(string path)
{
WordCount wc = new WordCount();
using (FileStream fs = File.Open(path, FileMode.Open, FileAccess.Read, FileShare.Read))
using (BufferedStream bs = new BufferedStream(fs))
using (StreamReader sr = new StreamReader(bs))
{
string line;
while ((line = sr.ReadLine()) != null)
{
string[] words = line.Split(new[] { " ", "," }, StringSplitOptions.RemoveEmptyEntries);
words.ToList().ForEach(x => wc.Add(x));
}
}
return wc;
}
public void NegativeWordFrequency()
{
WordCount wc = new WordCount();
WordOccurences results = wc.MostFrequent(-1);
}
public void ZeroWordFrequency()
{
WordCount wc = new WordCount();
WordOccurences results = wc.MostFrequent(0);
}
/// <summary>
/// Run the test and compare the output with the test data set.
/// </summary>
/// <param name="testSet"></param>
/// <param name="nRequest">n most frequently occuring words</param>
/// <param name="nExpected">the expected number of frequently occuring words, i.e. when nRequest is greater than the number of words</param>
private void RunTest(TestData[] testSet, int nRequest, int nExpected)
{
WordCount wc = new WordCount();
AddTestData(wc, testSet);
WordOccurences results = wc.MostFrequent(nRequest);
TestResult(results, nExpected, testSet);
}
/// <summary>
/// Add a single world X times to the WordCount object.
/// </summary>
/// <param name="wc"></param>
/// <param name="word"></param>
/// <param name="repeat"></param>
private void AddWord(WordCount wc, string word, int repeat)
{
for (int i = 0; i < repeat; i++)
{
wc.Add(word);
}
}
/// <summary>
/// Add the test data to the WordCount object.
/// </summary>
/// <param name="wc"></param>
/// <param name="testSet"></param>
private void AddTestData(WordCount wc, TestData[] testSet)
{
for (int i = 0; i < testSet.Length; i++)
{
TestData testData = testSet[i];
for (int k = 0; k < testData.words.Length; k++)
{
AddWord(wc, testData.words[k], testData.frequency);
}
}
}
public void NoWords()
{
WordCount wc = new WordCount();
WordOccurences results = wc.MostFrequent(10);
Assert.IsTrue(results.Count == 0);
}