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);
}
/* run this to test if your JRE is buggy
* public void testSentenceInstanceJREBUG() {
* // we use the default locale, as its randomized by LuceneTestCase
* BreakIterator bi = BreakIterator.getSentenceInstance(Locale.getDefault());
* Segment ci = new Segment();
* for (int i = 0; i < 10000; i++) {
* char text[] = TestUtil.randomUnicodeString(random).toCharArray();
* ci.array = text;
* ci.offset = 0;
* ci.count = text.length;
* consume(bi, ci);
* }
* }
*/
private void DoTests(CharArrayIterator ci)
{
// basics
ci.SetText("testing".ToCharArray(), 0, "testing".Length);
assertEquals(0, ci.BeginIndex);
assertEquals(7, ci.EndIndex);
assertEquals(0, ci.Index);
assertEquals('t', ci.Current());
assertEquals('e', ci.Next());
assertEquals('g', ci.Last());
assertEquals('n', ci.Previous());
assertEquals('t', ci.First());
assertEquals(CharacterIterator.DONE, ci.Previous());
// first()
ci.SetText("testing".ToCharArray(), 0, "testing".Length);
ci.Next();
// Sets the position to getBeginIndex() and returns the character at that position.
assertEquals('t', ci.First());
assertEquals(ci.BeginIndex, ci.Index);
// or DONE if the text is empty
ci.SetText(new char[] { }, 0, 0);
assertEquals(CharacterIterator.DONE, ci.First());
// last()
ci.SetText("testing".ToCharArray(), 0, "testing".Length);
// Sets the position to getEndIndex()-1 (getEndIndex() if the text is empty)
// and returns the character at that position.
assertEquals('g', ci.Last());
assertEquals(ci.Index, ci.EndIndex - 1);
// or DONE if the text is empty
ci.SetText(new char[] { }, 0, 0);
assertEquals(CharacterIterator.DONE, ci.Last());
assertEquals(ci.EndIndex, ci.Index);
// current()
// Gets the character at the current position (as returned by getIndex()).
ci.SetText("testing".ToCharArray(), 0, "testing".Length);
assertEquals('t', ci.Current());
ci.Last();
ci.Next();
// or DONE if the current position is off the end of the text.
assertEquals(CharacterIterator.DONE, ci.Current());
// next()
ci.SetText("te".ToCharArray(), 0, 2);
// Increments the iterator's index by one and returns the character at the new index.
assertEquals('e', ci.Next());
assertEquals(1, ci.Index);
// or DONE if the new position is off the end of the text range.
assertEquals(CharacterIterator.DONE, ci.Next());
assertEquals(ci.EndIndex, ci.Index);
// setIndex()
ci.SetText("test".ToCharArray(), 0, "test".Length);
try
{
ci.SetIndex(5);
fail();
}
catch (Exception e)
{
assertTrue(e is System.ArgumentException);
}
// clone()
var text = "testing".ToCharArray();
ci.SetText(text, 0, text.Length);
ci.Next();
var ci2 = ci.Clone() as CharArrayIterator;
assertEquals(ci.Index, ci2.Index);
assertEquals(ci.Next(), ci2.Next());
assertEquals(ci.Last(), ci2.Last());
}
