/// <summary>
/// Returns a new instance of a supported string metrics calculator.
/// </summary>
/// <param name="algorithm">A supported string similarity algorithm.</param>
/// <returns>A string metric measurement class.</returns>
public static IStringMetric FromAlgorithm(SimMetricAlgorithm algorithm)
{
switch (algorithm)
{
case SimMetricAlgorithm.BlockDistance:
return(new BlockDistance());
case SimMetricAlgorithm.ChapmanLengthDeviation:
return(new ChapmanLengthDeviation());
case SimMetricAlgorithm.CosineSimilarity:
return(new CosineSimilarity());
case SimMetricAlgorithm.DiceSimilarity:
return(new DiceSimilarity());
case SimMetricAlgorithm.EuclideanDistance:
return(new EuclideanDistance());
case SimMetricAlgorithm.JaccardSimilarity:
return(new JaccardSimilarity());
case SimMetricAlgorithm.Jaro:
return(new Jaro());
case SimMetricAlgorithm.JaroWinkler:
return(new JaroWinkler());
case SimMetricAlgorithm.MatchingCoefficient:
return(new MatchingCoefficient());
case SimMetricAlgorithm.MongeElkan:
return(new MongeElkan());
case SimMetricAlgorithm.NeedlemanWunch:
return(new NeedlemanWunch());
case SimMetricAlgorithm.OverlapCoefficient:
return(new OverlapCoefficient());
case SimMetricAlgorithm.QGramsDistance:
return(new QGramsDistance());
case SimMetricAlgorithm.SmithWaterman:
return(new SmithWaterman());
case SimMetricAlgorithm.SmithWatermanGotoh:
return(new SmithWatermanGotoh());
case SimMetricAlgorithm.SmithWatermanGotohWindowedAffine:
return(new SmithWatermanGotohWindowedAffine());
case SimMetricAlgorithm.ChapmanMeanLength:
return(new ChapmanMeanLength());
default:
return(new Levenstein());
}
}
/// <summary>
/// Compares two strings for similarity.
/// </summary>
/// <param name="firstWord">The first string to compare.</param>
/// <param name="secondWord">The second string to compare.</param>
/// <param name="threshold">Any similarity score at or above this value will be considered similar.</param>
/// <param name="algorithm">Specify the algorithm to use for comparing <paramref name="firstWord"/> and <paramref name="secondWord"/>. The default
/// algorithm is <see cref="SimMetricAlgorithm.Levenstein"/>.</param>
/// <returns>`true` if the similarity score is equal or above <paramref name="threshold"/>. Otherwise, `false`.</returns>
/// <remarks><![CDATA[
/// The following code demonstrates how to filter a list of strings based on the degree of similarity:
/// ```C#
/// string word = "fooler"
/// var list = new List<string>() { "fowler", "fish", "crawler" };
/// var filtered = List<string>();
/// foreach (string item in list)
/// {
/// if (item.NearEquals(word))
/// filtered.Add(item);
/// }
/// Console.WriteLine("You typed '{0}'. Did you mean: {1}", word, filtered);
/// ```
/// ]]></remarks>
public static bool NearEquals(this string firstWord, string secondWord, double threshold, SimMetricAlgorithm algorithm)
{
IStringMetric sim = StringMetricFactory.FromAlgorithm(algorithm);
double diff = sim.GetSimilarity(firstWord, secondWord);
return((diff < threshold) ? true : false);
}