protected override void SetNextSentence(int sentenceStart, int sentenceEnd)
{
this.sentenceStart = sentenceStart;
this.sentenceEnd = sentenceEnd;
wrapper.SetText(m_buffer, sentenceStart, sentenceEnd - sentenceStart);
wordBreaker.SetText(new string(wrapper.Text, wrapper.Start, wrapper.Length));
}
protected override void SetNextSentence(int sentenceStart, int sentenceEnd)
{
// LUCENENET TODO: This class isn't passing thread safety checks.
// Adding locking and extra cloning of BreakIterator seems to help, but
// it is not a complete fix.
lock (syncLock)
{
this.sentenceStart = sentenceStart;
this.sentenceEnd = sentenceEnd;
wrapper.SetText(m_buffer, sentenceStart, sentenceEnd - sentenceStart);
wordBreaker.SetText(new string(wrapper.Text, wrapper.Start, wrapper.Length));
}
}
protected override void SetNextSentence(int sentenceStart, int sentenceEnd)
{
UninterruptableMonitor.Enter(syncLock);
try
{
this.sentenceStart = sentenceStart;
this.sentenceEnd = sentenceEnd;
wrapper.SetText(m_buffer, sentenceStart, sentenceEnd - sentenceStart);
wordBreaker.SetText(new string(wrapper.Text, wrapper.Start, wrapper.Length));
}
finally
{
UninterruptableMonitor.Exit(syncLock);
}
}
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 (!m_input.IncrementToken())
{
return(false);
}
if (termAtt.Length == 0 || !thaiPattern.IsMatch(string.Empty + termAtt[0]))
{
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 <ICharTermAttribute>();
clonedOffsetAtt = clonedToken.GetAttribute <IOffsetAttribute>();
}
else
{
this.CopyTo(clonedToken);
}
// reinit CharacterIterator
charIterator.SetText(clonedTermAtt.Buffer, 0, clonedTermAtt.Length);
breaker.SetText(new string(charIterator.Text, charIterator.Start, charIterator.Length));
int end2 = breaker.Next();
if (end2 != BreakIterator.Done)
{
termAtt.Length = end2;
if (hasIllegalOffsets)
{
offsetAtt.SetOffset(clonedOffsetAtt.StartOffset, clonedOffsetAtt.EndOffset);
}
else
{
offsetAtt.SetOffset(clonedOffsetAtt.StartOffset, clonedOffsetAtt.StartOffset + end2);
}
// position increment keeps as it is for first token
return(true);
}
return(false);
}