public override Spans GetSpans(AtomicReaderContext context, IBits acceptDocs, IDictionary <Term, TermContext> termContexts)
{
TermState state;
if (!termContexts.TryGetValue(m_term, out TermContext termContext) || termContext == null)
{
// this happens with span-not query, as it doesn't include the NOT side in extractTerms()
// so we seek to the term now in this segment..., this sucks because its ugly mostly!
Fields fields = context.AtomicReader.Fields;
if (fields != null)
{
Terms terms = fields.GetTerms(m_term.Field);
if (terms != null)
{
TermsEnum termsEnum = terms.GetEnumerator();
if (termsEnum.SeekExact(m_term.Bytes))
{
state = termsEnum.GetTermState();
}
else
{
state = null;
}
}
else
{
state = null;
}
}
else
{
state = null;
}
}
else
{
state = termContext.Get(context.Ord);
}
if (state == null) // term is not present in that reader
{
return(TermSpans.EMPTY_TERM_SPANS);
}
TermsEnum termsEnum_ = context.AtomicReader.GetTerms(m_term.Field).GetEnumerator();
termsEnum_.SeekExact(m_term.Bytes, state);
DocsAndPositionsEnum postings = termsEnum_.DocsAndPositions(acceptDocs, null, DocsAndPositionsFlags.PAYLOADS);
if (postings != null)
{
return(new TermSpans(postings, m_term));
}
else
{
// term does exist, but has no positions
throw IllegalStateException.Create("field \"" + m_term.Field + "\" was indexed without position data; cannot run SpanTermQuery (term=" + m_term.Text + ")");
}
}
public override bool Collect(BytesRef bytes)
{
int e = terms.Add(bytes);
TermState state = termsEnum.GetTermState();
if (Debugging.AssertsEnabled)
{
Debugging.Assert(state != null);
}
if (e < 0)
{
// duplicate term: update docFreq
int pos = (-e) - 1;
array.termState[pos].Register(state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
if (Debugging.AssertsEnabled)
{
Debugging.Assert(array.boost[pos] == boostAtt.Boost, "boost should be equal in all segment TermsEnums");
}
}
else
{
// new entry: we populate the entry initially
array.boost[e] = boostAtt.Boost;
array.termState[e] = new TermContext(m_topReaderContext, state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
outerInstance.CheckMaxClauseCount(terms.Count);
}
return(true);
}
public override bool Collect(BytesRef bytes)
{
int e = terms.Add(bytes);
TermState state = termsEnum.GetTermState();
if (Debugging.AssertsEnabled)
{
Debugging.Assert(state != null);
}
if (e < 0)
{
// duplicate term: update docFreq
int pos = (-e) - 1;
array.termState[pos].Register(state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
// LUCENENET specific - compare bits rather than using equality operators to prevent these comparisons from failing in x86 in .NET Framework with optimizations enabled
if (Debugging.AssertsEnabled)
{
Debugging.Assert(NumericUtils.SingleToSortableInt32(array.boost[pos]) == NumericUtils.SingleToSortableInt32(boostAtt.Boost), "boost should be equal in all segment TermsEnums");
}
}
else
{
// new entry: we populate the entry initially
array.boost[e] = boostAtt.Boost;
array.termState[e] = new TermContext(m_topReaderContext, state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
outerInstance.CheckMaxClauseCount(terms.Count);
}
return(true);
}
public virtual void CollectTermContext(IndexReader reader, IList <AtomicReaderContext> leaves, TermContext[] contextArray, Term[] queryTerms)
{
TermsEnum termsEnum = null;
foreach (AtomicReaderContext context in leaves)
{
Fields fields = context.AtomicReader.Fields;
if (fields == null)
{
// reader has no fields
continue;
}
for (int i = 0; i < queryTerms.Length; i++)
{
Term term = queryTerms[i];
TermContext termContext = contextArray[i];
Terms terms = fields.GetTerms(term.Field);
if (terms == null)
{
// field does not exist
continue;
}
termsEnum = terms.GetIterator(termsEnum);
if (Debugging.AssertsEnabled)
{
Debugging.Assert(termsEnum != null);
}
if (termsEnum == TermsEnum.EMPTY)
{
continue;
}
if (termsEnum.SeekExact(term.Bytes))
{
if (termContext == null)
{
contextArray[i] = new TermContext(reader.Context, termsEnum.GetTermState(), context.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
}
else
{
termContext.Register(termsEnum.GetTermState(), context.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
}
}
}
}
}
public override bool Collect(BytesRef bytes)
{
float boost = boostAtt.Boost;
// make sure within a single seg we always collect
// terms in order
Debug.Assert(CompareToLastTerm(bytes));
//System.out.println("TTR.collect term=" + bytes.utf8ToString() + " boost=" + boost + " ord=" + readerContext.ord);
// ignore uncompetitive hits
if (stQueue.Count == maxSize)
{
ScoreTerm t = stQueue.Peek();
if (boost < t.Boost)
{
return(true);
}
if (boost == t.Boost && termComp.Compare(bytes, t.Bytes) > 0)
{
return(true);
}
}
TermState state = termsEnum.GetTermState();
Debug.Assert(state != null);
if (visitedTerms.TryGetValue(bytes, out ScoreTerm t2))
{
// if the term is already in the PQ, only update docFreq of term in PQ
Debug.Assert(t2.Boost == boost, "boost should be equal in all segment TermsEnums");
t2.TermState.Register(state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
}
else
{
// add new entry in PQ, we must clone the term, else it may get overwritten!
st.Bytes.CopyBytes(bytes);
st.Boost = boost;
visitedTerms[st.Bytes] = st;
Debug.Assert(st.TermState.DocFreq == 0);
st.TermState.Register(state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
stQueue.Add(st);
// possibly drop entries from queue
if (stQueue.Count > maxSize)
{
st = stQueue.Dequeue();
visitedTerms.Remove(st.Bytes);
st.TermState.Clear(); // reset the termstate!
}
else
{
st = new ScoreTerm(termComp, new TermContext(m_topReaderContext));
}
Debug.Assert(stQueue.Count <= maxSize, "the PQ size must be limited to maxSize");
// set maxBoostAtt with values to help FuzzyTermsEnum to optimize
if (stQueue.Count == maxSize)
{
t2 = stQueue.Peek();
maxBoostAtt.MaxNonCompetitiveBoost = t2.Boost;
maxBoostAtt.CompetitiveTerm = t2.Bytes;
}
}
return(true);
}
public override bool Collect(BytesRef bytes)
{
float boost = boostAtt.Boost;
// make sure within a single seg we always collect
// terms in order
if (Debugging.AssertsEnabled)
{
Debugging.Assert(CompareToLastTerm(bytes));
}
//System.out.println("TTR.collect term=" + bytes.utf8ToString() + " boost=" + boost + " ord=" + readerContext.ord);
// ignore uncompetitive hits
if (stQueue.Count == maxSize)
{
ScoreTerm t = stQueue.Peek();
// LUCENENET specific - compare bits rather than using equality operators to prevent these comparisons from failing in x86 in .NET Framework with optimizations enabled
if (NumericUtils.SingleToSortableInt32(boost) < NumericUtils.SingleToSortableInt32(t.Boost))
{
return(true);
}
// LUCENENET specific - compare bits rather than using equality operators to prevent these comparisons from failing in x86 in .NET Framework with optimizations enabled
if (NumericUtils.SingleToSortableInt32(boost) == NumericUtils.SingleToSortableInt32(t.Boost) && termComp.Compare(bytes, t.Bytes) > 0)
{
return(true);
}
}
TermState state = termsEnum.GetTermState();
if (Debugging.AssertsEnabled)
{
Debugging.Assert(state != null);
}
if (visitedTerms.TryGetValue(bytes, out ScoreTerm t2))
{
// if the term is already in the PQ, only update docFreq of term in PQ
// LUCENENET specific - compare bits rather than using equality operators to prevent these comparisons from failing in x86 in .NET Framework with optimizations enabled
if (Debugging.AssertsEnabled)
{
Debugging.Assert(NumericUtils.SingleToSortableInt32(t2.Boost) == NumericUtils.SingleToSortableInt32(boost), "boost should be equal in all segment TermsEnums");
}
t2.TermState.Register(state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
}
else
{
// add new entry in PQ, we must clone the term, else it may get overwritten!
st.Bytes.CopyBytes(bytes);
st.Boost = boost;
visitedTerms[st.Bytes] = st;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(st.TermState.DocFreq == 0);
}
st.TermState.Register(state, m_readerContext.Ord, termsEnum.DocFreq, termsEnum.TotalTermFreq);
stQueue.Add(st);
// possibly drop entries from queue
if (stQueue.Count > maxSize)
{
st = stQueue.Dequeue();
visitedTerms.Remove(st.Bytes);
st.TermState.Clear(); // reset the termstate!
}
else
{
st = new ScoreTerm(termComp, new TermContext(m_topReaderContext));
}
if (Debugging.AssertsEnabled)
{
Debugging.Assert(stQueue.Count <= maxSize, "the PQ size must be limited to maxSize");
}
// set maxBoostAtt with values to help FuzzyTermsEnum to optimize
if (stQueue.Count == maxSize)
{
t2 = stQueue.Peek();
maxBoostAtt.MaxNonCompetitiveBoost = t2.Boost;
maxBoostAtt.CompetitiveTerm = t2.Bytes;
}
}
return(true);
}