private bool CreateStemmer()
{
switch (mLanguage)
{
case Language.English:
mStemmer = new EnglishStemmer();
return(true);
case Language.German:
mStemmer = new German2Stemmer();
return(true);
case Language.French:
mStemmer = new FrenchStemmer();
return(true);
case Language.Spanish:
mStemmer = new SpanishStemmer();
return(true);
case Language.Italian:
mStemmer = new ItalianStemmer();
return(true);
case Language.Portuguese:
mStemmer = new PortugueseStemmer();
return(true);
case Language.Danish:
mStemmer = new DanishStemmer();
return(true);
case Language.Dutch:
mStemmer = new DutchStemmer();
return(true);
case Language.Finnish:
mStemmer = new FinnishStemmer();
return(true);
case Language.Norwegian:
mStemmer = new NorwegianStemmer();
return(true);
case Language.Russian:
mStemmer = new RussianStemmer();
return(true);
case Language.Swedish:
mStemmer = new SwedishStemmer();
return(true);
default:
return(false);
}
}
// *** IStemmer interface implementation ***
public string GetStem(string word)
{
Utils.ThrowException(word == null ? new ArgumentNullException("word") : null);
try
{
ISnowballStemmer stemmer = CreateStemmer();
stemmer.SetCurrent(word);
stemmer.Stem();
return(stemmer.GetCurrent());
}
catch { return(word); }
}
/// <summary>
/// Processes a string of terms using the Snowball stemming algorithm.
/// </summary>
/// <param name="text">Input string</param>
/// <param name="language">Stemmer language</param>
/// <returns>Stemmed terms</returns>
public static string ProcessText(string text, SnowballStemmerEnum language)
{
StringBuilder builder = new StringBuilder();
string result = string.Empty;
string stemmedWord;
char[] delimiterChars = { ' ' };
string[] tokens = text.Split(delimiterChars);
ISnowballStemmer stemmer = SnowballStemmerUtils.GetStemmer(language);
foreach (string token in tokens)
{
stemmedWord = stemmer.Stem(token);
builder.AppendFormat("{0} ", stemmedWord);
}
result = builder.ToString().Trim();
return(result);
}
private bool CreateStemmer()
{
switch (mLanguage)
{
case Language.English:
mStemmer = new EnglishStemmer();
return true;
case Language.German:
mStemmer = new German2Stemmer();
return true;
case Language.French:
mStemmer = new FrenchStemmer();
return true;
case Language.Spanish:
mStemmer = new SpanishStemmer();
return true;
case Language.Italian:
mStemmer = new ItalianStemmer();
return true;
case Language.Portuguese:
mStemmer = new PortugueseStemmer();
return true;
case Language.Danish:
mStemmer = new DanishStemmer();
return true;
case Language.Dutch:
mStemmer = new DutchStemmer();
return true;
case Language.Finnish:
mStemmer = new FinnishStemmer();
return true;
case Language.Norwegian:
mStemmer = new NorwegianStemmer();
return true;
case Language.Russian:
mStemmer = new RussianStemmer();
return true;
case Language.Swedish:
mStemmer = new SwedishStemmer();
return true;
default:
return false;
}
}
