public SingleCharTokenizer(TokenStream input): base(input)
{
_input = input;
_termAttribute = (TermAttribute)AddAttribute(typeof(TermAttribute));
_offsetAttribute = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
_positionIncrementAttribute = (PositionIncrementAttribute)AddAttribute(typeof(PositionIncrementAttribute));
}
override public object Clone()
{
OffsetAttribute impl = new OffsetAttribute();
impl.endOffset = endOffset;
impl.startOffset = startOffset;
return(impl);
}
public HyphenationTokenFilter(TokenStream input, Hyphenator hyphenator)
: base(input)
{
_hyphenator = hyphenator;
_termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
_typeAtt = (TypeAttribute)AddAttribute(typeof(TypeAttribute));
_ofsAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
public AnsjTokenizer(AbstractAnalysis analysis, TextReader input, HashSet<string> filter, bool pstemming)
: base(input)
{
_analysis = analysis;
_termAtt = AddAttribute<TermAttribute>();
_offsetAtt = AddAttribute<OffsetAttribute>();
_positionAttr = AddAttribute<PositionIncrementAttribute>();
_filter = filter;
_pstemming = pstemming;
}
public IntMetaDataTokenStream(string tokenText)
{
_tokenText = tokenText;
// NOTE: Calling the AddAttribute<T> method failed, so
// switched to using AddAttributeImpl.
_termAttribute = new TermAttribute();
_offsetAttribute = new OffsetAttribute();
_payloadAtt = new PayloadAttribute();
base.AddAttributeImpl(_termAttribute);
base.AddAttributeImpl(_offsetAttribute);
base.AddAttributeImpl(_payloadAtt);
}
public override bool Equals(object other)
{
if (other == this)
{
return(true);
}
if (other is OffsetAttribute)
{
OffsetAttribute o = (OffsetAttribute)other;
return(o.startOffset == startOffset && o.endOffset == endOffset);
}
return(false);
}
/**
* Creates NGramTokenFilter with given min and max n-grams.
* <param name="input"><see cref="TokenStream"/> holding the input to be tokenized</param>
* <param name="minGram">the smallest n-gram to generate</param>
* <param name="maxGram">the largest n-gram to generate</param>
*/
public NGramTokenFilter(TokenStream input, int minGram, int maxGram)
: base(input)
{
if (minGram < 1)
{
throw new System.ArgumentException("minGram must be greater than zero");
}
if (minGram > maxGram)
{
throw new System.ArgumentException("minGram must not be greater than maxGram");
}
this.minGram = minGram;
this.maxGram = maxGram;
this.termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
this.offsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
public PrefixAwareTokenFilter(TokenStream prefix, TokenStream suffix) : base(suffix)
{
Suffix = suffix;
Prefix = prefix;
_prefixExhausted = false;
// ReSharper disable DoNotCallOverridableMethodsInConstructor
_termAtt = (TermAttribute) AddAttribute(typeof (TermAttribute));
_posIncrAtt = (PositionIncrementAttribute) AddAttribute(typeof (PositionIncrementAttribute));
_payloadAtt = (PayloadAttribute) AddAttribute(typeof (PayloadAttribute));
_offsetAtt = (OffsetAttribute) AddAttribute(typeof (OffsetAttribute));
_typeAtt = (TypeAttribute) AddAttribute(typeof (TypeAttribute));
_flagsAtt = (FlagsAttribute) AddAttribute(typeof (FlagsAttribute));
// ReSharper restore DoNotCallOverridableMethodsInConstructor
_pTermAtt = (TermAttribute) prefix.AddAttribute(typeof (TermAttribute));
_pPosIncrAtt = (PositionIncrementAttribute) prefix.AddAttribute(typeof (PositionIncrementAttribute));
_pPayloadAtt = (PayloadAttribute) prefix.AddAttribute(typeof (PayloadAttribute));
_pOffsetAtt = (OffsetAttribute) prefix.AddAttribute(typeof (OffsetAttribute));
_pTypeAtt = (TypeAttribute) prefix.AddAttribute(typeof (TypeAttribute));
_pFlagsAtt = (FlagsAttribute) prefix.AddAttribute(typeof (FlagsAttribute));
}
public virtual void TestOffsetAttribute()
{
OffsetAttribute att = new OffsetAttribute();
Assert.AreEqual(0, att.StartOffset);
Assert.AreEqual(0, att.EndOffset);
att.SetOffset(12, 34);
// no string test here, because order unknown
OffsetAttribute att2 = (OffsetAttribute)AssertCloneIsEqual(att);
Assert.AreEqual(12, att2.StartOffset);
Assert.AreEqual(34, att2.EndOffset);
att2 = (OffsetAttribute)AssertCopyIsEqual(att);
Assert.AreEqual(12, att2.StartOffset);
Assert.AreEqual(34, att2.EndOffset);
att.Clear();
Assert.AreEqual(0, att.StartOffset);
Assert.AreEqual(0, att.EndOffset);
}
/**
* Creates EdgeNGramTokenFilter that can generate n-grams in the sizes of the given range
*
* @param input {@link TokenStream} holding the input to be tokenized
* @param side the {@link Side} from which to chop off an n-gram
* @param minGram the smallest n-gram to generate
* @param maxGram the largest n-gram to generate
*/
public EdgeNGramTokenFilter(TokenStream input, Side side, int minGram, int maxGram)
: base(input)
{
if (side == null)
{
throw new System.ArgumentException("sideLabel must be either front or back");
}
if (minGram < 1)
{
throw new System.ArgumentException("minGram must be greater than zero");
}
if (minGram > maxGram)
{
throw new System.ArgumentException("minGram must not be greater than maxGram");
}
this.minGram = minGram;
this.maxGram = maxGram;
this.side = side;
this.termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
this.offsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
public TokenListStream(ICollection<Token> tokens)
{
_tokens = tokens;
_termAtt = (TermAttribute) AddAttribute(typeof (TermAttribute));
_posIncrAtt = (PositionIncrementAttribute) AddAttribute(typeof (PositionIncrementAttribute));
_payloadAtt = (PayloadAttribute) AddAttribute(typeof (PayloadAttribute));
_offsetAtt = (OffsetAttribute) AddAttribute(typeof (OffsetAttribute));
_typeAtt = (TypeAttribute) AddAttribute(typeof (TypeAttribute));
_flagsAtt = (FlagsAttribute) AddAttribute(typeof (FlagsAttribute));
}
public virtual void TestOffsetAttribute()
{
OffsetAttribute att = new OffsetAttribute();
Assert.AreEqual(0, att.StartOffset);
Assert.AreEqual(0, att.EndOffset);
att.SetOffset(12, 34);
// no string test here, because order unknown
OffsetAttribute att2 = (OffsetAttribute) AssertCloneIsEqual(att);
Assert.AreEqual(12, att2.StartOffset);
Assert.AreEqual(34, att2.EndOffset);
att2 = (OffsetAttribute) AssertCopyIsEqual(att);
Assert.AreEqual(12, att2.StartOffset);
Assert.AreEqual(34, att2.EndOffset);
att.Clear();
Assert.AreEqual(0, att.StartOffset);
Assert.AreEqual(0, att.EndOffset);
}
public override void CopyTo(AttributeImpl target)
{
OffsetAttribute t = (OffsetAttribute)target;
t.SetOffset(startOffset, endOffset);
}
private void init(Side side, int minGram, int maxGram)
{
if (side == null)
{
throw new System.ArgumentException("sideLabel must be either front or back");
}
if (minGram < 1)
{
throw new System.ArgumentException("minGram must be greater than zero");
}
if (minGram > maxGram)
{
throw new System.ArgumentException("minGram must not be greater than maxGram");
}
this.minGram = minGram;
this.maxGram = maxGram;
this.side = side;
this.termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
this.offsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
protected EdgeNGramTokenFilter(TokenStream input)
: base(input)
{
this.termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
this.offsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
private void Init(System.IO.TextReader input, HebMorph.StreamLemmatizer _lemmatizer,
HebMorph.LemmaFilters.LemmaFilterBase _lemmaFilter, bool AlwaysSaveMarkedOriginal)
{
termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
offsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
posIncrAtt = (PositionIncrementAttribute)AddAttribute(typeof(PositionIncrementAttribute));
typeAtt = (TypeAttribute)AddAttribute(typeof(TypeAttribute));
//payAtt = (PayloadAttribute)AddAttribute(typeof(PayloadAttribute));
this._streamLemmatizer = _lemmatizer;
this._streamLemmatizer.SetStream(input);
this.alwaysSaveMarkedOriginal = AlwaysSaveMarkedOriginal;
this.lemmaFilter = _lemmaFilter;
}
/// <summary>
/// Constructs a ShingleFilter with the specified single size from the TokenStream
/// </summary>
/// <param name="input">input token stream</param>
/// <param name="maxShingleSize">maximum shingle size produced by the filter.</param>
public ShingleFilter(TokenStream input, int maxShingleSize) : base(input)
{
SetMaxShingleSize(maxShingleSize);
// ReSharper disable DoNotCallOverridableMethodsInConstructor
_termAtt = (TermAttribute) AddAttribute(typeof (TermAttribute));
_offsetAtt = (OffsetAttribute) AddAttribute(typeof (OffsetAttribute));
_posIncrAtt = (PositionIncrementAttribute) AddAttribute(typeof (PositionIncrementAttribute));
_typeAtt = (TypeAttribute) AddAttribute(typeof (TypeAttribute));
// ReSharper restore DoNotCallOverridableMethodsInConstructor
}
private void init(int minGram, int maxGram)
{
if (minGram < 1)
{
throw new System.ArgumentException("minGram must be greater than zero");
}
if (minGram > maxGram)
{
throw new System.ArgumentException("minGram must not be greater than maxGram");
}
this.minGram = minGram;
this.maxGram = maxGram;
this.termAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
this.offsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
/// <summary>
/// Creates a shingle filter based on a user defined matrix.
///
/// The filter /will/ delete columns from the input matrix! You will not be able to reset the filter if you used this constructor.
/// todo: don't touch the matrix! use a bool, set the input stream to null or something, and keep track of where in the matrix we are at.
///
/// </summary>
/// <param name="matrix">the input based for creating shingles. Does not need to contain any information until ShingleMatrixFilter.Next(Token) is called the first time.</param>
/// <param name="minimumShingleSize">minimum number of tokens in any shingle.</param>
/// <param name="maximumShingleSize">maximum number of tokens in any shingle.</param>
/// <param name="spacerCharacter">character to use between texts of the token parts in a shingle. null for none.</param>
/// <param name="ignoringSinglePrefixOrSuffixShingle">if true, shingles that only contains permutation of the first of the last column will not be produced as shingles. Useful when adding boundary marker tokens such as '^' and '$'.</param>
/// <param name="settingsCodec">codec used to read input token weight and matrix positioning.</param>
public ShingleMatrixFilter(Matrix.Matrix matrix, int minimumShingleSize, int maximumShingleSize, Char spacerCharacter, bool ignoringSinglePrefixOrSuffixShingle, TokenSettingsCodec settingsCodec)
{
Matrix = matrix;
MinimumShingleSize = minimumShingleSize;
MaximumShingleSize = maximumShingleSize;
SpacerCharacter = spacerCharacter;
IsIgnoringSinglePrefixOrSuffixShingle = ignoringSinglePrefixOrSuffixShingle;
_settingsCodec = settingsCodec;
// ReSharper disable DoNotCallOverridableMethodsInConstructor
_termAtt = (TermAttribute) AddAttribute(typeof (TermAttribute));
_posIncrAtt = (PositionIncrementAttribute) AddAttribute(typeof (PositionIncrementAttribute));
_payloadAtt = (PayloadAttribute) AddAttribute(typeof (PayloadAttribute));
_offsetAtt = (OffsetAttribute) AddAttribute(typeof (OffsetAttribute));
_typeAtt = (TypeAttribute) AddAttribute(typeof (TypeAttribute));
_flagsAtt = (FlagsAttribute) AddAttribute(typeof (FlagsAttribute));
// ReSharper restore DoNotCallOverridableMethodsInConstructor
// set the input to be an empty token stream, we already have the data.
_input = new EmptyTokenStream();
_inTermAtt = (TermAttribute) _input.AddAttribute(typeof (TermAttribute));
_inPosIncrAtt = (PositionIncrementAttribute) _input.AddAttribute(typeof (PositionIncrementAttribute));
_inPayloadAtt = (PayloadAttribute) _input.AddAttribute(typeof (PayloadAttribute));
_inOffsetAtt = (OffsetAttribute) _input.AddAttribute(typeof (OffsetAttribute));
_inTypeAtt = (TypeAttribute) _input.AddAttribute(typeof (TypeAttribute));
_inFlagsAtt = (FlagsAttribute) _input.AddAttribute(typeof (FlagsAttribute));
}
/// <summary>
/// Creates a shingle filter with ad hoc parameter settings.
/// </summary>
/// <param name="input">stream from which to construct the matrix</param>
/// <param name="minimumShingleSize">minimum number of tokens in any shingle.</param>
/// <param name="maximumShingleSize">maximum number of tokens in any shingle.</param>
/// <param name="spacerCharacter">character to use between texts of the token parts in a shingle. null for none.</param>
/// <param name="ignoringSinglePrefixOrSuffixShingle">if true, shingles that only contains permutation of the first of the last column will not be produced as shingles. Useful when adding boundary marker tokens such as '^' and '$'.</param>
/// <param name="settingsCodec">codec used to read input token weight and matrix positioning.</param>
public ShingleMatrixFilter(TokenStream input, int minimumShingleSize, int maximumShingleSize, Char? spacerCharacter, bool ignoringSinglePrefixOrSuffixShingle, TokenSettingsCodec settingsCodec)
{
_input = input;
MinimumShingleSize = minimumShingleSize;
MaximumShingleSize = maximumShingleSize;
SpacerCharacter = spacerCharacter;
IsIgnoringSinglePrefixOrSuffixShingle = ignoringSinglePrefixOrSuffixShingle;
_settingsCodec = settingsCodec;
// ReSharper disable DoNotCallOverridableMethodsInConstructor
_termAtt = (TermAttribute) AddAttribute(typeof (TermAttribute));
_posIncrAtt = (PositionIncrementAttribute) AddAttribute(typeof (PositionIncrementAttribute));
_payloadAtt = (PayloadAttribute) AddAttribute(typeof (PayloadAttribute));
_offsetAtt = (OffsetAttribute) AddAttribute(typeof (OffsetAttribute));
_typeAtt = (TypeAttribute) AddAttribute(typeof (TypeAttribute));
_flagsAtt = (FlagsAttribute) AddAttribute(typeof (FlagsAttribute));
// ReSharper restore DoNotCallOverridableMethodsInConstructor
_inTermAtt = (TermAttribute) input.AddAttribute(typeof (TermAttribute));
_inPosIncrAtt = (PositionIncrementAttribute) input.AddAttribute(typeof (PositionIncrementAttribute));
_inPayloadAtt = (PayloadAttribute) input.AddAttribute(typeof (PayloadAttribute));
_inOffsetAtt = (OffsetAttribute) input.AddAttribute(typeof (OffsetAttribute));
_inTypeAtt = (TypeAttribute) input.AddAttribute(typeof (TypeAttribute));
_inFlagsAtt = (FlagsAttribute) input.AddAttribute(typeof (FlagsAttribute));
}
override public System.Object Clone()
{
OffsetAttribute impl = new OffsetAttribute();
impl.endOffset = endOffset;
impl.startOffset = startOffset;
return impl;
}