using System;
using System.IO;
using System.Collections;
namespace Lucene.Net.Analysis.De
{
///
/// A filter that stems German words. It supports a table of words that should
/// not be stemmed at all. The stemmer used can be changed at runtime after the
/// filter object is created (as long as it is a GermanStemmer).
///
public sealed class GermanStemFilter : TokenFilter
{
///
/// The actual token in the input stream.
///
private Token token = null;
private GermanStemmer stemmer = null;
private Hashtable exclusions = null;
public GermanStemFilter( TokenStream _in ) : base(_in)
{
stemmer = new GermanStemmer();
}
///
/// Builds a GermanStemFilter that uses an exclusiontable.
///
///
///
public GermanStemFilter( TokenStream _in, Hashtable exclusiontable ): this(_in)
{
exclusions = exclusiontable;
}
///
///
/// Returns the next token in the stream, or null at EOS
public override Token Next()
{
if ( ( token = input.Next() ) == null )
{
return null;
}
// Check the exclusiontable
else if ( exclusions != null && exclusions.Contains( token.TermText() ) )
{
return token;
}
else
{
String s = stemmer.Stem( token.TermText() );
// If not stemmed, dont waste the time creating a new token
if ( !s.Equals( token.TermText() ) )
{
return new Token( s, token.StartOffset(),
token.EndOffset(), token.Type() );
}
return token;
}
}
///
/// Set a alternative/custom GermanStemmer for this filter.
///
///
public void SetStemmer( GermanStemmer stemmer )
{
if ( stemmer != null )
{
this.stemmer = stemmer;
}
}
///
/// Set an alternative exclusion list for this filter.
///
///
public void SetExclusionTable( Hashtable exclusiontable )
{
exclusions = exclusiontable;
}
}
}