public void TestOffsetsWithTokenizer()
{
const string input = @"test1 <a href=""foo"">testlink</a> test2 test3";
Tokenizer t = new WhitespaceTokenizer(new HTMLStripCharFilter(CharReader.Get(new StringReader(input))));
string token = string.Empty;
List <Token> results = new List <Token>();
OffsetAttribute att = ((OffsetAttribute)t.GetAttribute(typeof(OffsetAttribute)));
t.IncrementToken();
Assert.AreEqual(0, att.StartOffset());
Assert.AreEqual(5, att.EndOffset() - att.StartOffset());
t.IncrementToken();
Assert.AreEqual(20, att.StartOffset());
Assert.AreEqual(8, att.EndOffset() - att.StartOffset());
t.IncrementToken();
Assert.AreEqual(33, att.StartOffset());
Assert.AreEqual(5, att.EndOffset() - att.StartOffset());
t.IncrementToken();
Assert.AreEqual(39, att.StartOffset());
Assert.AreEqual(5, att.EndOffset() - att.StartOffset());
}
public InputWindowToken(ShingleFilter outerInstance, AttributeSource attSource)
{
this.outerInstance = outerInstance;
this.attSource = attSource;
this.termAtt = attSource.GetAttribute(typeof(CharTermAttribute));
this.offsetAtt = attSource.GetAttribute(typeof(OffsetAttribute));
}
public SingleCharTokenizer(TokenStream input) : base(input)
{
_input = input;
_termAttribute = (TermAttribute)AddAttribute(typeof(TermAttribute));
_offsetAttribute = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
_positionIncrementAttribute = (PositionIncrementAttribute)AddAttribute(typeof(PositionIncrementAttribute));
}
public T ReadMetadata <T>()
{
using (var reader = CreateReader())
{
var header = new MetadataHeader(reader); //MetadataHeader self registers to reader
using (var vreader = (DependencyReader)reader.CreateVirtualReader(header.Header.HeaderSize))
{
var mainBlock = header.StructureDefinitions.First(s => s.Type == StructureType.Main).TargetIndex;
vreader.Seek(header.DataBlocks[mainBlock].Offset, SeekOrigin.Begin);
var result = vreader.ReadObject <T>();
var blockProps = typeof(T).GetProperties()
.Where(p => p.PropertyType.IsGenericType && p.PropertyType.GetGenericTypeDefinition() == typeof(BlockCollection <>));
foreach (var prop in blockProps)
{
var collection = prop.GetValue(result) as IBlockCollection;
var offset = OffsetAttribute.ValueFor(prop);
collection.LoadBlocks(mainBlock, offset, vreader);
}
return(result);
}
}
}
// we only check a few core attributes here.
// TODO: test other things
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void assertEquals(String s, org.apache.lucene.analysis.TokenStream left, org.apache.lucene.analysis.TokenStream right) throws Exception
public virtual void assertEquals(string s, TokenStream left, TokenStream right)
{
left.reset();
right.reset();
CharTermAttribute leftTerm = left.addAttribute(typeof(CharTermAttribute));
CharTermAttribute rightTerm = right.addAttribute(typeof(CharTermAttribute));
OffsetAttribute leftOffset = left.addAttribute(typeof(OffsetAttribute));
OffsetAttribute rightOffset = right.addAttribute(typeof(OffsetAttribute));
PositionIncrementAttribute leftPos = left.addAttribute(typeof(PositionIncrementAttribute));
PositionIncrementAttribute rightPos = right.addAttribute(typeof(PositionIncrementAttribute));
while (left.incrementToken())
{
assertTrue("wrong number of tokens for input: " + s, right.incrementToken());
assertEquals("wrong term text for input: " + s, leftTerm.ToString(), rightTerm.ToString());
assertEquals("wrong position for input: " + s, leftPos.PositionIncrement, rightPos.PositionIncrement);
assertEquals("wrong start offset for input: " + s, leftOffset.startOffset(), rightOffset.startOffset());
assertEquals("wrong end offset for input: " + s, leftOffset.endOffset(), rightOffset.endOffset());
}
;
assertFalse("wrong number of tokens for input: " + s, right.incrementToken());
left.end();
right.end();
assertEquals("wrong final offset for input: " + s, leftOffset.endOffset(), rightOffset.endOffset());
left.close();
right.close();
}
public LuceneTokenizerAdapter(Tokenizer tokenizer)
{
this.tokenizer = tokenizer;
this.termAttr = (CharTermAttribute)this.tokenizer.GetAttribute <ICharTermAttribute>();
this.offsetAttr = (OffsetAttribute)this.tokenizer.GetAttribute <IOffsetAttribute>();
}
// LUCENE-3642: normalize BMP->SMP and check that offsets are correct
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testCrossPlaneNormalization2() throws java.io.IOException
public virtual void testCrossPlaneNormalization2()
{
Analyzer analyzer = new AnalyzerAnonymousInnerClassHelper2(this);
int num = 1000 * RANDOM_MULTIPLIER;
for (int i = 0; i < num; i++)
{
string s = TestUtil.randomUnicodeString(random());
TokenStream ts = analyzer.tokenStream("foo", s);
try
{
ts.reset();
OffsetAttribute offsetAtt = ts.addAttribute(typeof(OffsetAttribute));
while (ts.incrementToken())
{
string highlightedText = StringHelperClass.SubstringSpecial(s, offsetAtt.startOffset(), offsetAtt.endOffset());
for (int j = 0, cp = 0; j < highlightedText.Length; j += char.charCount(cp))
{
cp = char.ConvertToUtf32(highlightedText, j);
assertTrue("non-letter:" + cp.ToString("x"), char.IsLetter(cp));
}
}
ts.end();
}
finally
{
IOUtils.closeWhileHandlingException(ts);
}
}
// just for fun
checkRandomData(random(), analyzer, num);
}
public BoostingTokenFilter(TokenStream input, int[] startOffsets, float[] boosts)
: base(input)
{
_startOffsets = startOffsets;
_boosts = boosts;
_offsetAttr = (OffsetAttribute)base.addAttribute(typeof(OffsetAttribute));
_payloadAttr = (PayloadAttribute)base.addAttribute(typeof(PayloadAttribute));
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: @Override public void reset() throws java.io.IOException
public override void reset()
{
base.reset();
hasMoreTokensInClone = false;
clonedToken = null;
clonedTermAtt = null;
clonedOffsetAtt = null;
}
public void TestOffsetValueExpression()
{
Assert.AreEqual(0, OffsetAttribute.ValueFor((DataClass01 obj) => obj.Property1));
Assert.AreEqual(10, OffsetAttribute.ValueFor((DataClass01 obj) => obj.Property2, 15));
Assert.AreEqual(15, OffsetAttribute.ValueFor((DataClass01 obj) => obj.Property2, 25));
Assert.AreEqual(20, OffsetAttribute.ValueFor((DataClass01 obj) => obj.Property3, 15));
}
private static IntPtr ParseField(FieldInfo field, PatternFinder pf)
{
OffsetAttribute offset = (OffsetAttribute)Attribute.GetCustomAttributes(field, typeof(OffsetAttribute))
.FirstOrDefault();
OffsetValueCN valcn = (OffsetValueCN)Attribute.GetCustomAttributes(field, typeof(OffsetValueCN))
.FirstOrDefault();
OffsetValueNA valna = (OffsetValueNA)Attribute.GetCustomAttributes(field, typeof(OffsetValueNA))
.FirstOrDefault();
IntPtr result = IntPtr.Zero;
if (Constants.Lang == Language.Chn)
{
if (valcn != null)
{
return((IntPtr)valcn.Value);
}
if (offset == null)
{
return(IntPtr.Zero);
}
bool b1 = true;
IntPtr[] results = pf.FindMany(offset.PatternCN, ref b1);
if (results != null)
{
result = results[0];
}
}
else
{
if (valna != null)
{
return((IntPtr)valna.Value);
}
if (offset == null)
{
return(IntPtr.Zero);
}
bool b1 = true;
IntPtr[] results = pf.FindMany(offset.Pattern, ref b1);
if (results != null)
{
result = results[0];
}
}
Logger.Info("[OffsetManager][{0:,27}] {1}", field.Name, result.ToString("X"));
return(result);
}
public void SetTokenStream(TokenStream ts)
{
this.ts = ts;
if (this.ts.HasAttribute <IOffsetAttribute>())
{
this.offsetAttr = (OffsetAttribute)this.ts.GetAttribute <IOffsetAttribute>();
}
if (this.ts.HasAttribute <ICharTermAttribute>())
{
this.termAttr = (CharTermAttribute)this.ts.GetAttribute <ICharTermAttribute>();
}
}
private void Init(int gramSize)
{
if (gramSize < 1)
{
throw new ArgumentException(
"minGram must be greater than zero");
}
_mGramSize = gramSize;
_mTermAtt = (TermAttribute)AddAttribute(typeof(TermAttribute));
_mOffsetAtt = (OffsetAttribute)AddAttribute(typeof(OffsetAttribute));
}
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 void TestOffsetValue()
{
var prop1 = typeof(DataClass01).GetProperty(nameof(DataClass01.Property1));
var prop2 = typeof(DataClass01).GetProperty(nameof(DataClass01.Property2));
var prop3 = typeof(DataClass01).GetProperty(nameof(DataClass01.Property3));
Assert.AreEqual(0, OffsetAttribute.ValueFor(prop1));
Assert.AreEqual(10, OffsetAttribute.ValueFor(prop2, 15));
Assert.AreEqual(15, OffsetAttribute.ValueFor(prop2, 25));
Assert.AreEqual(20, OffsetAttribute.ValueFor(prop3, 15));
}
/**
* 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));
}
/**
* 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");
}
_minGram = minGram;
_maxGram = maxGram;
_termAtt = (TermAttribute)AddAttribute <ITermAttribute>();
_offsetAtt = (OffsetAttribute)AddAttribute <IOffsetAttribute>();
}
/**
* 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 (minGram < 1)
{
throw new IllegalArgumentException("minGram must be greater than zero");
}
if (minGram > maxGram)
{
throw new IllegalArgumentException("minGram must not be greater than maxGram");
}
this.minGram = minGram;
this.maxGram = maxGram;
this.side = side;
termAtt = (TermAttribute)addAttribute(typeof(TermAttribute));
offsetAtt = (OffsetAttribute)addAttribute(typeof(OffsetAttribute));
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testSupplementaryCharacters() throws java.io.IOException
public virtual void testSupplementaryCharacters()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String s = org.apache.lucene.util.TestUtil.randomUnicodeString(random(), 10);
string s = TestUtil.randomUnicodeString(random(), 10);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int codePointCount = s.codePointCount(0, s.length());
int codePointCount = s.codePointCount(0, s.Length);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int minGram = org.apache.lucene.util.TestUtil.nextInt(random(), 1, 3);
int minGram = TestUtil.Next(random(), 1, 3);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int maxGram = org.apache.lucene.util.TestUtil.nextInt(random(), minGram, 10);
int maxGram = TestUtil.Next(random(), minGram, 10);
TokenStream tk = new KeywordTokenizer(new StringReader(s));
tk = new NGramTokenFilter(TEST_VERSION_CURRENT, tk, minGram, maxGram);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.tokenattributes.CharTermAttribute termAtt = tk.addAttribute(org.apache.lucene.analysis.tokenattributes.CharTermAttribute.class);
CharTermAttribute termAtt = tk.addAttribute(typeof(CharTermAttribute));
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.tokenattributes.OffsetAttribute offsetAtt = tk.addAttribute(org.apache.lucene.analysis.tokenattributes.OffsetAttribute.class);
OffsetAttribute offsetAtt = tk.addAttribute(typeof(OffsetAttribute));
tk.reset();
for (int start = 0; start < codePointCount; ++start)
{
for (int end = start + minGram; end <= Math.Min(codePointCount, start + maxGram); ++end)
{
assertTrue(tk.incrementToken());
assertEquals(0, offsetAtt.startOffset());
assertEquals(s.Length, offsetAtt.endOffset());
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int startIndex = Character.offsetByCodePoints(s, 0, start);
int startIndex = char.offsetByCodePoints(s, 0, start);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int endIndex = Character.offsetByCodePoints(s, 0, end);
int endIndex = char.offsetByCodePoints(s, 0, end);
assertEquals(s.Substring(startIndex, endIndex - startIndex), termAtt.ToString());
}
}
assertFalse(tk.incrementToken());
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testOtherLetterOffset() throws java.io.IOException
public virtual void testOtherLetterOffset()
{
string s = "a天b";
ChineseTokenizer tokenizer = new ChineseTokenizer(new StringReader(s));
int correctStartOffset = 0;
int correctEndOffset = 1;
OffsetAttribute offsetAtt = tokenizer.getAttribute(typeof(OffsetAttribute));
tokenizer.reset();
while (tokenizer.incrementToken())
{
assertEquals(correctStartOffset, offsetAtt.startOffset());
assertEquals(correctEndOffset, offsetAtt.endOffset());
correctStartOffset++;
correctEndOffset++;
}
tokenizer.end();
tokenizer.close();
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testFilterTokens() throws Exception
public virtual void testFilterTokens()
{
SnowballFilter filter = new SnowballFilter(new TestTokenStream(this), "English");
CharTermAttribute termAtt = filter.getAttribute(typeof(CharTermAttribute));
OffsetAttribute offsetAtt = filter.getAttribute(typeof(OffsetAttribute));
TypeAttribute typeAtt = filter.getAttribute(typeof(TypeAttribute));
PayloadAttribute payloadAtt = filter.getAttribute(typeof(PayloadAttribute));
PositionIncrementAttribute posIncAtt = filter.getAttribute(typeof(PositionIncrementAttribute));
FlagsAttribute flagsAtt = filter.getAttribute(typeof(FlagsAttribute));
filter.incrementToken();
assertEquals("accent", termAtt.ToString());
assertEquals(2, offsetAtt.startOffset());
assertEquals(7, offsetAtt.endOffset());
assertEquals("wrd", typeAtt.type());
assertEquals(3, posIncAtt.PositionIncrement);
assertEquals(77, flagsAtt.Flags);
assertEquals(new BytesRef(new sbyte[] { 0, 1, 2, 3 }), payloadAtt.Payload);
}
internal void LoadBlocks(int currentBlock, long collectionOffset, DependencyReader reader)
{
if (blockCount == 0)
{
return;
}
var structdef = metadata.StructureDefinitions.First(s => s.FieldBlock == currentBlock && s.FieldOffset == collectionOffset);
if (structdef.TargetIndex < 0)
{
return;
}
var block = metadata.DataBlocks[structdef.TargetIndex];
var blockSize = FixedSizeAttribute.ValueFor(typeof(T));
reader.Seek(block.Offset, SeekOrigin.Begin);
for (int i = 0; i < blockCount; i++)
{
Add(reader.ReadObject <T>());
}
var blockProps = typeof(T).GetProperties()
.Where(p => p.PropertyType.IsGenericType && p.PropertyType.GetGenericTypeDefinition() == typeof(BlockCollection <>));
int index = 0;
foreach (var item in this)
{
var adjustedBase = blockSize * index++;
foreach (var prop in blockProps)
{
var collection = prop.GetValue(item) as IBlockCollection;
var offset = OffsetAttribute.ValueFor(prop);
collection.LoadBlocks(structdef.TargetIndex, adjustedBase + offset, reader);
}
}
}
/*
* Need to worry about multiple scenarios:
* - need to go for the longest match
* a b => foo #shouldn't match if "a b" is followed by "c d"
* a b c d => bar
* - need to backtrack - retry matches for tokens already read
* a b c d => foo
* b c => bar
* If the input stream is "a b c x", one will consume "a b c d"
* trying to match the first rule... all but "a" should be
* pushed back so a match may be made on "b c".
* - don't try and match generated tokens (thus need separate queue)
* matching is not recursive.
* - handle optional generation of original tokens in all these cases,
* merging token streams to preserve token positions.
* - preserve original positionIncrement of first matched token
*/
public override bool IncrementToken()
{
while (true)
{
// if there are any generated tokens, return them... don't try any
// matches against them, as we specifically don't want recursion.
if (replacement != null && replacement.MoveNext())
{
copy(this, replacement.Current);
return(true);
}
// common case fast-path of first token not matching anything
AttributeSource firstTok = nextTok();
if (firstTok == null)
{
return(false);
}
var termAtt = firstTok.AddAttribute <ICharTermAttribute>();
SlowSynonymMap result = map.submap != null?map.submap.Get(termAtt.Buffer(), 0, termAtt.Length) : null;
if (result == null)
{
copy(this, firstTok);
return(true);
}
// fast-path failed, clone ourselves if needed
if (firstTok == this)
{
firstTok = CloneAttributes();
}
// OK, we matched a token, so find the longest match.
matched = new LinkedList <>();
result = match(result);
if (result == null)
{
// no match, simply return the first token read.
copy(this, firstTok);
return(true);
}
// reuse, or create new one each time?
List <AttributeSource> generated = new List <AttributeSource>(result.synonyms.Length + matched.Count + 1);
//
// there was a match... let's generate the new tokens, merging
// in the matched tokens (position increments need adjusting)
//
AttributeSource lastTok = matched.Count == 0 ? firstTok : matched.Last.Value;
bool includeOrig = result.IncludeOrig;
AttributeSource origTok = includeOrig ? firstTok : null;
PositionIncrementAttribute firstPosIncAtt = firstTok.addAttribute(typeof(PositionIncrementAttribute));
int origPos = firstPosIncAtt.PositionIncrement; // position of origTok in the original stream
int repPos = 0; // curr position in replacement token stream
int pos = 0; // current position in merged token stream
for (int i = 0; i < result.synonyms.Length; i++)
{
Token repTok = result.synonyms[i];
AttributeSource newTok = firstTok.cloneAttributes();
CharTermAttribute newTermAtt = newTok.addAttribute(typeof(CharTermAttribute));
OffsetAttribute newOffsetAtt = newTok.addAttribute(typeof(OffsetAttribute));
PositionIncrementAttribute newPosIncAtt = newTok.addAttribute(typeof(PositionIncrementAttribute));
OffsetAttribute lastOffsetAtt = lastTok.addAttribute(typeof(OffsetAttribute));
newOffsetAtt.setOffset(newOffsetAtt.startOffset(), lastOffsetAtt.endOffset());
newTermAtt.copyBuffer(repTok.buffer(), 0, repTok.length());
repPos += repTok.PositionIncrement;
if (i == 0) // make position of first token equal to original
{
repPos = origPos;
}
// if necessary, insert original tokens and adjust position increment
while (origTok != null && origPos <= repPos)
{
PositionIncrementAttribute origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPosInc.PositionIncrement = origPos - pos;
generated.Add(origTok);
pos += origPosInc.PositionIncrement;
origTok = matched.Count == 0 ? null : matched.RemoveFirst();
if (origTok != null)
{
origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPos += origPosInc.PositionIncrement;
}
}
newPosIncAtt.PositionIncrement = repPos - pos;
generated.Add(newTok);
pos += newPosIncAtt.PositionIncrement;
}
// finish up any leftover original tokens
while (origTok != null)
{
PositionIncrementAttribute origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPosInc.PositionIncrement = origPos - pos;
generated.Add(origTok);
pos += origPosInc.PositionIncrement;
origTok = matched.Count == 0 ? null : matched.RemoveFirst();
if (origTok != null)
{
origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPos += origPosInc.PositionIncrement;
}
}
// what if we replaced a longer sequence with a shorter one?
// a/0 b/5 => foo/0
// should I re-create the gap on the next buffered token?
replacement = generated.GetEnumerator();
// Now return to the top of the loop to read and return the first
// generated token.. The reason this is done is that we may have generated
// nothing at all, and may need to continue with more matching logic.
}
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: @Override public boolean incrementToken() throws java.io.IOException
public override bool incrementToken()
{
if (hasMoreTokensInClone)
{
int start = breaker.current();
int end = breaker.next();
if (end != BreakIterator.DONE)
{
clonedToken.copyTo(this);
termAtt.copyBuffer(clonedTermAtt.buffer(), start, end - start);
if (hasIllegalOffsets)
{
offsetAtt.setOffset(clonedOffsetAtt.startOffset(), clonedOffsetAtt.endOffset());
}
else
{
offsetAtt.setOffset(clonedOffsetAtt.startOffset() + start, clonedOffsetAtt.startOffset() + end);
}
if (handlePosIncr)
{
posAtt.PositionIncrement = 1;
}
return(true);
}
hasMoreTokensInClone = false;
}
if (!input.incrementToken())
{
return(false);
}
if (termAtt.length() == 0 || char.UnicodeBlock.of(termAtt.charAt(0)) != char.UnicodeBlock.THAI)
{
return(true);
}
hasMoreTokensInClone = true;
// if length by start + end offsets doesn't match the term text then assume
// this is a synonym and don't adjust the offsets.
hasIllegalOffsets = offsetAtt.endOffset() - offsetAtt.startOffset() != termAtt.length();
// we lazy init the cloned token, as in ctor not all attributes may be added
if (clonedToken == null)
{
clonedToken = cloneAttributes();
clonedTermAtt = clonedToken.getAttribute(typeof(CharTermAttribute));
clonedOffsetAtt = clonedToken.getAttribute(typeof(OffsetAttribute));
}
else
{
this.copyTo(clonedToken);
}
// reinit CharacterIterator
charIterator.setText(clonedTermAtt.buffer(), 0, clonedTermAtt.length());
breaker.Text = charIterator;
int end = breaker.next();
if (end != BreakIterator.DONE)
{
termAtt.Length = end;
if (hasIllegalOffsets)
{
offsetAtt.setOffset(clonedOffsetAtt.startOffset(), clonedOffsetAtt.endOffset());
}
else
{
offsetAtt.setOffset(clonedOffsetAtt.startOffset(), clonedOffsetAtt.startOffset() + end);
}
// position increment keeps as it is for first token
return(true);
}
return(false);
}
public override void CopyTo(IAttribute target)
{
OffsetAttribute t = (OffsetAttribute)target;
t.SetOffset(start, end);
}
/// <summary>
/// Retrieve suggestions.
/// </summary>
public virtual IList <LookupResult> Lookup(string key, HashSet <BytesRef> contexts, int num)
{
if (contexts != null)
{
throw new System.ArgumentException("this suggester doesn't support contexts");
}
TokenStream ts = queryAnalyzer.TokenStream("", key.ToString());
try
{
TermToBytesRefAttribute termBytesAtt = ts.AddAttribute <TermToBytesRefAttribute>();
OffsetAttribute offsetAtt = ts.AddAttribute <OffsetAttribute>();
PositionLengthAttribute posLenAtt = ts.AddAttribute <PositionLengthAttribute>();
PositionIncrementAttribute posIncAtt = ts.AddAttribute <PositionIncrementAttribute>();
ts.Reset();
var lastTokens = new BytesRef[grams];
//System.out.println("lookup: key='" + key + "'");
// Run full analysis, but save only the
// last 1gram, last 2gram, etc.:
BytesRef tokenBytes = termBytesAtt.BytesRef;
int maxEndOffset = -1;
bool sawRealToken = false;
while (ts.IncrementToken())
{
termBytesAtt.FillBytesRef();
sawRealToken |= tokenBytes.Length > 0;
// TODO: this is somewhat iffy; today, ShingleFilter
// sets posLen to the gram count; maybe we should make
// a separate dedicated att for this?
int gramCount = posLenAtt.PositionLength;
Debug.Assert(gramCount <= grams);
// Safety: make sure the recalculated count "agrees":
if (CountGrams(tokenBytes) != gramCount)
{
throw new System.ArgumentException("tokens must not contain separator byte; got token=" + tokenBytes + " but gramCount=" + gramCount + " does not match recalculated count=" + countGrams(tokenBytes));
}
maxEndOffset = Math.Max(maxEndOffset, offsetAtt.EndOffset());
lastTokens[gramCount - 1] = BytesRef.DeepCopyOf(tokenBytes);
}
ts.End();
if (!sawRealToken)
{
throw new System.ArgumentException("no tokens produced by analyzer, or the only tokens were empty strings");
}
// Carefully fill last tokens with _ tokens;
// ShingleFilter appraently won't emit "only hole"
// tokens:
int endPosInc = posIncAtt.PositionIncrement;
// Note this will also be true if input is the empty
// string (in which case we saw no tokens and
// maxEndOffset is still -1), which in fact works out OK
// because we fill the unigram with an empty BytesRef
// below:
bool lastTokenEnded = offsetAtt.EndOffset() > maxEndOffset || endPosInc > 0;
//System.out.println("maxEndOffset=" + maxEndOffset + " vs " + offsetAtt.endOffset());
if (lastTokenEnded)
{
//System.out.println(" lastTokenEnded");
// If user hit space after the last token, then
// "upgrade" all tokens. This way "foo " will suggest
// all bigrams starting w/ foo, and not any unigrams
// starting with "foo":
for (int i = grams - 1; i > 0; i--)
{
BytesRef token = lastTokens[i - 1];
if (token == null)
{
continue;
}
token.Grow(token.Length + 1);
token.Bytes[token.Length] = separator;
token.Length++;
lastTokens[i] = token;
}
lastTokens[0] = new BytesRef();
}
var arc = new FST.Arc <long?>();
var bytesReader = fst.BytesReader;
// Try highest order models first, and if they return
// results, return that; else, fallback:
double backoff = 1.0;
IList <LookupResult> results = new List <LookupResult>(num);
// We only add a given suffix once, from the highest
// order model that saw it; for subsequent lower order
// models we skip it:
var seen = new HashSet <BytesRef>();
for (int gram = grams - 1; gram >= 0; gram--)
{
BytesRef token = lastTokens[gram];
// Don't make unigram predictions from empty string:
if (token == null || (token.Length == 0 && key.Length > 0))
{
// Input didn't have enough tokens:
//System.out.println(" gram=" + gram + ": skip: not enough input");
continue;
}
if (endPosInc > 0 && gram <= endPosInc)
{
// Skip hole-only predictions; in theory we
// shouldn't have to do this, but we'd need to fix
// ShingleFilter to produce only-hole tokens:
//System.out.println(" break: only holes now");
break;
}
//System.out.println("try " + (gram+1) + " gram token=" + token.utf8ToString());
// TODO: we could add fuzziness here
// match the prefix portion exactly
//Pair<Long,BytesRef> prefixOutput = null;
long?prefixOutput = null;
prefixOutput = LookupPrefix(fst, bytesReader, token, arc);
//System.out.println(" prefixOutput=" + prefixOutput);
if (prefixOutput == null)
{
// This model never saw this prefix, e.g. the
// trigram model never saw context "purple mushroom"
backoff *= ALPHA;
continue;
}
// TODO: we could do this division at build time, and
// bake it into the FST?
// Denominator for computing scores from current
// model's predictions:
long contextCount = totTokens;
BytesRef lastTokenFragment = null;
for (int i = token.Length - 1; i >= 0; i--)
{
if (token.Bytes[token.Offset + i] == separator)
{
BytesRef context = new BytesRef(token.Bytes, token.Offset, i);
long? output = Util.Get(fst, Lucene.Net.Util.Fst.Util.ToIntsRef(context, new IntsRef()));
Debug.Assert(output != null);
contextCount = DecodeWeight(output);
lastTokenFragment = new BytesRef(token.Bytes, token.Offset + i + 1, token.Length - i - 1);
break;
}
}
BytesRef finalLastToken;
if (lastTokenFragment == null)
{
finalLastToken = BytesRef.DeepCopyOf(token);
}
else
{
finalLastToken = BytesRef.DeepCopyOf(lastTokenFragment);
}
Debug.Assert(finalLastToken.Offset == 0);
CharsRef spare = new CharsRef();
// complete top-N
Util.Fst.Util.TopResults <long?> completions = null;
try
{
// Because we store multiple models in one FST
// (1gram, 2gram, 3gram), we must restrict the
// search so that it only considers the current
// model. For highest order model, this is not
// necessary since all completions in the FST
// must be from this model, but for lower order
// models we have to filter out the higher order
// ones:
// Must do num+seen.size() for queue depth because we may
// reject up to seen.size() paths in acceptResult():
Util.Fst.Util.TopNSearcher <long?> searcher = new TopNSearcherAnonymousInnerClassHelper(this, fst, num, num + seen.Count, weightComparator, seen, finalLastToken);
// since this search is initialized with a single start node
// it is okay to start with an empty input path here
searcher.AddStartPaths(arc, prefixOutput, true, new IntsRef());
completions = searcher.Search();
Debug.Assert(completions.IsComplete);
}
catch (IOException bogus)
{
throw new Exception(bogus);
}
int prefixLength = token.Length;
BytesRef suffix = new BytesRef(8);
//System.out.println(" " + completions.length + " completions");
foreach (Util.Fst.Util.Result <long?> completion in completions)
{
token.Length = prefixLength;
// append suffix
Util.Fst.Util.ToBytesRef(completion.Input, suffix);
token.Append(suffix);
//System.out.println(" completion " + token.utf8ToString());
// Skip this path if a higher-order model already
// saw/predicted its last token:
BytesRef lastToken = token;
for (int i = token.Length - 1; i >= 0; i--)
{
if (token.Bytes[token.Offset + i] == separator)
{
Debug.Assert(token.Length - i - 1 > 0);
lastToken = new BytesRef(token.Bytes, token.Offset + i + 1, token.Length - i - 1);
break;
}
}
if (seen.Contains(lastToken))
{
//System.out.println(" skip dup " + lastToken.utf8ToString());
goto nextCompletionContinue;
}
seen.Add(BytesRef.DeepCopyOf(lastToken));
spare.Grow(token.Length);
UnicodeUtil.UTF8toUTF16(token, spare);
LookupResult result = new LookupResult(spare.ToString(), (long)(long.MaxValue * backoff * ((double)DecodeWeight(completion.Output)) / contextCount));
results.Add(result);
Debug.Assert(results.Count == seen.Count);
//System.out.println(" add result=" + result);
nextCompletionContinue :;
}
nextCompletionBreak :
backoff *= ALPHA;
}
results.Sort(new ComparatorAnonymousInnerClassHelper(this));
if (results.Count > num)
{
results.SubList(num, results.Count).Clear();
}
return(results);
}
finally
{
IOUtils.CloseWhileHandlingException(ts);
}
}
public static void AssertTokenStreamContents(TokenStream ts, System.String[] output, int[] startOffsets, int[] endOffsets, System.String[] types, int[] posIncrements, int?finalOffset)
{
Assert.IsNotNull(output);
CheckClearAttributesAttribute checkClearAtt = (CheckClearAttributesAttribute)ts.AddAttribute(typeof(CheckClearAttributesAttribute));
Assert.IsTrue(ts.HasAttribute(typeof(TermAttribute)), "has no TermAttribute");
TermAttribute termAtt = (TermAttribute)ts.GetAttribute(typeof(TermAttribute));
OffsetAttribute offsetAtt = null;
if (startOffsets != null || endOffsets != null || finalOffset != null)
{
Assert.IsTrue(ts.HasAttribute(typeof(OffsetAttribute)), "has no OffsetAttribute");
offsetAtt = (OffsetAttribute)ts.GetAttribute(typeof(OffsetAttribute));
}
TypeAttribute typeAtt = null;
if (types != null)
{
Assert.IsTrue(ts.HasAttribute(typeof(TypeAttribute)), "has no TypeAttribute");
typeAtt = (TypeAttribute)ts.GetAttribute(typeof(TypeAttribute));
}
PositionIncrementAttribute posIncrAtt = null;
if (posIncrements != null)
{
Assert.IsTrue(ts.HasAttribute(typeof(PositionIncrementAttribute)), "has no PositionIncrementAttribute");
posIncrAtt = (PositionIncrementAttribute)ts.GetAttribute(typeof(PositionIncrementAttribute));
}
ts.Reset();
for (int i = 0; i < output.Length; i++)
{
// extra safety to enforce, that the state is not preserved and also assign bogus values
ts.ClearAttributes();
termAtt.SetTermBuffer("bogusTerm");
if (offsetAtt != null)
{
offsetAtt.SetOffset(14584724, 24683243);
}
if (typeAtt != null)
{
typeAtt.SetType("bogusType");
}
if (posIncrAtt != null)
{
posIncrAtt.SetPositionIncrement(45987657);
}
checkClearAtt.GetAndResetClearCalled(); // reset it, because we called clearAttribute() before
Assert.IsTrue(ts.IncrementToken(), "token " + i + " does not exist");
Assert.IsTrue(checkClearAtt.GetAndResetClearCalled(), "clearAttributes() was not called correctly in TokenStream chain");
Assert.AreEqual(output[i], termAtt.Term(), "term " + i);
if (startOffsets != null)
{
Assert.AreEqual(startOffsets[i], offsetAtt.StartOffset(), "startOffset " + i);
}
if (endOffsets != null)
{
Assert.AreEqual(endOffsets[i], offsetAtt.EndOffset(), "endOffset " + i);
}
if (types != null)
{
Assert.AreEqual(types[i], typeAtt.Type(), "type " + i);
}
if (posIncrements != null)
{
Assert.AreEqual(posIncrements[i], posIncrAtt.GetPositionIncrement(), "posIncrement " + i);
}
}
Assert.IsFalse(ts.IncrementToken(), "end of stream");
ts.End();
if (finalOffset.HasValue)
{
Assert.AreEqual(finalOffset, offsetAtt.EndOffset(), "finalOffset ");
}
ts.Close();
}
private void btnAnalyze_Click(object sender, System.EventArgs e)
{
ShowError("");
try
{
if (cmbAnalyzers.SelectedItem == null)
{
return;
}
Analyzer analyzer = null;
try
{
// Trying to create type from executing assembly
Type analyzerType = (Type)analyzers[cmbAnalyzers.SelectedItem];
if (null == analyzerType)
{
// Trying to create type from Lucene.Net assembly
Assembly a = Assembly.GetAssembly(typeof(Lucene.Net.Analysis.Analyzer));
analyzerType = a.GetType((string)cmbAnalyzers.SelectedItem);
}
// Trying to create with default constructor
analyzer = (Analyzer)Activator.CreateInstance(analyzerType, Util.Version.LUCENE_30);
}
catch (Exception)
{}
if (null == analyzer)
{
ShowError("Couldn't instantiate analyzer - public zero-argument constructor required");
return;
}
txtOutput.Text = txtText.Text;
lstResults.BeginUpdate();
try
{
TokenStream ts = analyzer.TokenStream("Analyze", new StringReader(txtText.Text));
var token = ts.GetAttribute <ITermAttribute>();
var offset = ts.GetAttribute <IOffsetAttribute>();
lstResults.Items.Clear();
tokens.Clear();
while (ts.IncrementToken())
{
lstResults.Items.Add(token.Term);
var a = new OffsetAttribute();
a.SetOffset(offset.StartOffset, offset.EndOffset);
tokens.Add(a);
}
}
finally
{
lstResults.EndUpdate();
}
}
catch (Exception exc)
{
ShowError("Error analyzing: " + exc.Message);
}
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testBasic() throws Exception
public virtual void testBasic()
{
b = new SynonymMap.Builder(true);
add("a", "foo", true);
add("a b", "bar fee", true);
add("b c", "dog collar", true);
add("c d", "dog harness holder extras", true);
add("m c e", "dog barks loudly", false);
add("i j k", "feep", true);
add("e f", "foo bar", false);
add("e f", "baz bee", false);
add("z", "boo", false);
add("y", "bee", true);
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
verify("a b c", "a/bar b/fee c");
// syn output extends beyond input tokens
verify("x a b c d", "x a/bar b/fee c/dog d/harness holder extras");
verify("a b a", "a/bar b/fee a/foo");
// outputs that add to one another:
verify("c d c d", "c/dog d/harness c/holder/dog d/extras/harness holder extras");
// two outputs for same input
verify("e f", "foo/baz bar/bee");
// verify multi-word / single-output offsets:
verify("g i j k g", "g i/feep:7_3 j k g");
// mixed keepOrig true/false:
verify("a m c e x", "a/foo dog barks loudly x");
verify("c d m c e x", "c/dog d/harness holder/dog extras/barks loudly x");
assertTrue(tokensOut.CaptureCount > 0);
// no captureStates when no syns matched
verify("p q r s t", "p q r s t");
assertEquals(0, tokensOut.CaptureCount);
// no captureStates when only single-input syns, w/ no
// lookahead needed, matched
verify("p q z y t", "p q boo y/bee t");
assertEquals(0, tokensOut.CaptureCount);
}
/// <summary>
/// <para>Get the next token from the input stream.
/// </para>
/// <para>If the next token has <code>positionIncrement > 1</code>,
/// <code>positionIncrement - 1</code> <seealso cref="#fillerToken"/>s are
/// inserted first.
/// </para>
/// </summary>
/// <param name="target"> Where to put the new token; if null, a new instance is created. </param>
/// <returns> On success, the populated token; null otherwise </returns>
/// <exception cref="IOException"> if the input stream has a problem </exception>
private InputWindowToken getNextToken(InputWindowToken target)
{
InputWindowToken newTarget = target;
if (numFillerTokensToInsert > 0)
{
if (null == target)
{
newTarget = new InputWindowToken(this, nextInputStreamToken.cloneAttributes());
}
else
{
nextInputStreamToken.copyTo(target.attSource);
}
// A filler token occupies no space
newTarget.offsetAtt.setOffset(newTarget.offsetAtt.startOffset(), newTarget.offsetAtt.startOffset());
newTarget.termAtt.copyBuffer(fillerToken, 0, fillerToken.Length);
newTarget.isFiller = true;
--numFillerTokensToInsert;
}
else if (isNextInputStreamToken)
{
if (null == target)
{
newTarget = new InputWindowToken(this, nextInputStreamToken.cloneAttributes());
}
else
{
nextInputStreamToken.copyTo(target.attSource);
}
isNextInputStreamToken = false;
newTarget.isFiller = false;
}
else if (!exhausted)
{
if (input.incrementToken())
{
if (null == target)
{
newTarget = new InputWindowToken(this, cloneAttributes());
}
else
{
this.copyTo(target.attSource);
}
if (posIncrAtt.PositionIncrement > 1)
{
// Each output shingle must contain at least one input token,
// so no more than (maxShingleSize - 1) filler tokens will be inserted.
numFillerTokensToInsert = Math.Min(posIncrAtt.PositionIncrement - 1, maxShingleSize - 1);
// Save the current token as the next input stream token
if (null == nextInputStreamToken)
{
nextInputStreamToken = cloneAttributes();
}
else
{
this.copyTo(nextInputStreamToken);
}
isNextInputStreamToken = true;
// A filler token occupies no space
newTarget.offsetAtt.setOffset(offsetAtt.startOffset(), offsetAtt.startOffset());
newTarget.termAtt.copyBuffer(fillerToken, 0, fillerToken.Length);
newTarget.isFiller = true;
--numFillerTokensToInsert;
}
else
{
newTarget.isFiller = false;
}
}
else
{
exhausted = true;
input.end();
endState = captureState();
numFillerTokensToInsert = Math.Min(posIncrAtt.PositionIncrement, maxShingleSize - 1);
if (numFillerTokensToInsert > 0)
{
nextInputStreamToken = new AttributeSource(AttributeFactory);
nextInputStreamToken.addAttribute(typeof(CharTermAttribute));
OffsetAttribute newOffsetAtt = nextInputStreamToken.addAttribute(typeof(OffsetAttribute));
newOffsetAtt.setOffset(offsetAtt.endOffset(), offsetAtt.endOffset());
// Recurse/loop just once:
return(getNextToken(target));
}
else
{
newTarget = null;
}
}
}
else
{
newTarget = null;
}
return(newTarget);
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testRandom() throws Exception
public virtual void testRandom()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int alphabetSize = org.apache.lucene.util.TestUtil.nextInt(random(), 2, 7);
int alphabetSize = TestUtil.Next(random(), 2, 7);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int docLen = atLeast(3000);
int docLen = atLeast(3000);
//final int docLen = 50;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String document = getRandomString('a', alphabetSize, docLen);
string document = getRandomString('a', alphabetSize, docLen);
if (VERBOSE)
{
Console.WriteLine("TEST: doc=" + document);
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int numSyn = atLeast(5);
int numSyn = atLeast(5);
//final int numSyn = 2;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final java.util.Map<String,OneSyn> synMap = new java.util.HashMap<>();
IDictionary<string, OneSyn> synMap = new Dictionary<string, OneSyn>();
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final java.util.List<OneSyn> syns = new java.util.ArrayList<>();
IList<OneSyn> syns = new List<OneSyn>();
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final boolean dedup = random().nextBoolean();
bool dedup = random().nextBoolean();
if (VERBOSE)
{
Console.WriteLine(" dedup=" + dedup);
}
b = new SynonymMap.Builder(dedup);
for (int synIDX = 0;synIDX < numSyn;synIDX++)
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String synIn = getRandomString('a', alphabetSize, org.apache.lucene.util.TestUtil.nextInt(random(), 1, 5)).trim();
string synIn = getRandomString('a', alphabetSize, TestUtil.Next(random(), 1, 5)).Trim();
OneSyn s = synMap[synIn];
if (s == null)
{
s = new OneSyn();
s.@in = synIn;
syns.Add(s);
s.@out = new List<>();
synMap[synIn] = s;
s.keepOrig = random().nextBoolean();
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String synOut = getRandomString('0', 10, org.apache.lucene.util.TestUtil.nextInt(random(), 1, 5)).trim();
string synOut = getRandomString('0', 10, TestUtil.Next(random(), 1, 5)).Trim();
[email protected](synOut);
add(synIn, synOut, s.keepOrig);
if (VERBOSE)
{
Console.WriteLine(" syns[" + synIDX + "] = " + s.@in + " -> " + s.@out + " keepOrig=" + s.keepOrig);
}
}
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
if (dedup)
{
pruneDups(syns);
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String expected = slowSynMatcher(document, syns, 5);
string expected = slowSynMatcher(document, syns, 5);
if (VERBOSE)
{
Console.WriteLine("TEST: expected=" + expected);
}
verify(document, expected);
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testOutputHangsOffEnd() throws Exception
public virtual void testOutputHangsOffEnd()
{
b = new SynonymMap.Builder(true);
const bool keepOrig = false;
// b hangs off the end (no input token under it):
add("a", "a b", keepOrig);
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
// Make sure endOffset inherits from previous input token:
verify("a", "a b:1");
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testBasic2() throws Exception
public virtual void testBasic2()
{
b = new SynonymMap.Builder(true);
const bool keepOrig = false;
add("aaa", "aaaa1 aaaa2 aaaa3", keepOrig);
add("bbb", "bbbb1 bbbb2", keepOrig);
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
if (keepOrig)
{
verify("xyzzy bbb pot of gold", "xyzzy bbb/bbbb1 pot/bbbb2 of gold");
verify("xyzzy aaa pot of gold", "xyzzy aaa/aaaa1 pot/aaaa2 of/aaaa3 gold");
}
else
{
verify("xyzzy bbb pot of gold", "xyzzy bbbb1 pot/bbbb2 of gold");
verify("xyzzy aaa pot of gold", "xyzzy aaaa1 pot/aaaa2 of/aaaa3 gold");
}
}
public override void CopyTo(Attribute target)
{
OffsetAttribute t = (OffsetAttribute)target;
t.SetOffset(Start, End);
}
public InputWindowToken(ShingleFilter outerInstance, AttributeSource attSource)
{
this.outerInstance = outerInstance;
this.attSource = attSource;
this.termAtt = attSource.getAttribute(typeof(CharTermAttribute));
this.offsetAtt = attSource.getAttribute(typeof(OffsetAttribute));
}
