internal override void AddTerm(int termID)
{
Debug.Assert(docState.TestPoint("TermVectorsTermsWriterPerField.addTerm start"));
TermVectorsPostingsArray postings = (TermVectorsPostingsArray)termsHashPerField.postingsArray;
postings.freqs[termID]++;
WriteProx(postings, termID);
}
internal override void CopyTo(ParallelPostingsArray toArray, int numToCopy)
{
Debug.Assert(toArray is TermVectorsPostingsArray);
TermVectorsPostingsArray to = (TermVectorsPostingsArray)toArray;
base.CopyTo(toArray, numToCopy);
Array.Copy(freqs, 0, to.freqs, 0, size);
Array.Copy(lastOffsets, 0, to.lastOffsets, 0, size);
Array.Copy(lastPositions, 0, to.lastPositions, 0, size);
}
internal override void NewTerm(int termID)
{
Debug.Assert(docState.TestPoint("TermVectorsTermsWriterPerField.newTerm start"));
TermVectorsPostingsArray postings = (TermVectorsPostingsArray)termsHashPerField.postingsArray;
postings.freqs[termID] = 1;
postings.lastOffsets[termID] = 0;
postings.lastPositions[termID] = 0;
WriteProx(postings, termID);
}
internal override void AddTerm(int termID)
{
// LUCENENET: .NET doesn't support asserts in release mode
if (Lucene.Net.Diagnostics.Debugging.AssertsEnabled)
{
docState.TestPoint("TermVectorsTermsWriterPerField.addTerm start");
}
TermVectorsPostingsArray postings = (TermVectorsPostingsArray)termsHashPerField.postingsArray;
postings.freqs[termID]++;
WriteProx(postings, termID);
}
internal void WriteProx(TermVectorsPostingsArray postings, int termID)
{
if (doVectorOffsets)
{
int startOffset = fieldState.Offset + offsetAttribute.StartOffset;
int endOffset = fieldState.Offset + offsetAttribute.EndOffset;
termsHashPerField.WriteVInt32(1, startOffset - postings.lastOffsets[termID]);
termsHashPerField.WriteVInt32(1, endOffset - startOffset);
postings.lastOffsets[termID] = endOffset;
}
if (doVectorPositions)
{
BytesRef payload;
if (payloadAttribute == null)
{
payload = null;
}
else
{
payload = payloadAttribute.Payload;
}
int pos = fieldState.Position - postings.lastPositions[termID];
if (payload != null && payload.Length > 0)
{
termsHashPerField.WriteVInt32(0, (pos << 1) | 1);
termsHashPerField.WriteVInt32(0, payload.Length);
termsHashPerField.WriteBytes(0, payload.Bytes, payload.Offset, payload.Length);
hasPayloads = true;
}
else
{
termsHashPerField.WriteVInt32(0, pos << 1);
}
postings.lastPositions[termID] = fieldState.Position;
}
}
internal void WriteProx(TermVectorsPostingsArray postings, int termID)
{
if (DoVectorOffsets)
{
int startOffset = FieldState.Offset_Renamed + OffsetAttribute.StartOffset();
int endOffset = FieldState.Offset_Renamed + OffsetAttribute.EndOffset();
TermsHashPerField.WriteVInt(1, startOffset - postings.LastOffsets[termID]);
TermsHashPerField.WriteVInt(1, endOffset - startOffset);
postings.LastOffsets[termID] = endOffset;
}
if (DoVectorPositions)
{
BytesRef payload;
if (PayloadAttribute == null)
{
payload = null;
}
else
{
payload = PayloadAttribute.Payload;
}
int pos = FieldState.Position_Renamed - postings.LastPositions[termID];
if (payload != null && payload.Length > 0)
{
TermsHashPerField.WriteVInt(0, (pos << 1) | 1);
TermsHashPerField.WriteVInt(0, payload.Length);
TermsHashPerField.WriteBytes(0, payload.Bytes, payload.Offset, payload.Length);
HasPayloads = true;
}
else
{
TermsHashPerField.WriteVInt(0, pos << 1);
}
postings.LastPositions[termID] = FieldState.Position_Renamed;
}
}
internal void FinishDocument()
{
Debug.Assert(docState.TestPoint("TermVectorsTermsWriterPerField.finish start"));
int numPostings = termsHashPerField.bytesHash.Count;
BytesRef flushTerm = termsWriter.flushTerm;
Debug.Assert(numPostings >= 0);
if (numPostings > maxNumPostings)
{
maxNumPostings = numPostings;
}
// this is called once, after inverting all occurrences
// of a given field in the doc. At this point we flush
// our hash into the DocWriter.
Debug.Assert(termsWriter.VectorFieldsInOrder(fieldInfo));
TermVectorsPostingsArray postings = (TermVectorsPostingsArray)termsHashPerField.postingsArray;
TermVectorsWriter tv = termsWriter.writer;
int[] termIDs = termsHashPerField.SortPostings(tv.Comparer);
tv.StartField(fieldInfo, numPostings, doVectorPositions, doVectorOffsets, hasPayloads);
ByteSliceReader posReader = doVectorPositions ? termsWriter.vectorSliceReaderPos : null;
ByteSliceReader offReader = doVectorOffsets ? termsWriter.vectorSliceReaderOff : null;
ByteBlockPool termBytePool = termsHashPerField.termBytePool;
for (int j = 0; j < numPostings; j++)
{
int termID = termIDs[j];
int freq = postings.freqs[termID];
// Get BytesRef
termBytePool.SetBytesRef(flushTerm, postings.textStarts[termID]);
tv.StartTerm(flushTerm, freq);
if (doVectorPositions || doVectorOffsets)
{
if (posReader != null)
{
termsHashPerField.InitReader(posReader, termID, 0);
}
if (offReader != null)
{
termsHashPerField.InitReader(offReader, termID, 1);
}
tv.AddProx(freq, posReader, offReader);
}
tv.FinishTerm();
}
tv.FinishField();
termsHashPerField.Reset();
fieldInfo.SetStoreTermVectors();
}
internal void WriteProx(TermVectorsPostingsArray postings, int termID)
{
if (DoVectorOffsets)
{
int startOffset = FieldState.Offset_Renamed + OffsetAttribute.StartOffset();
int endOffset = FieldState.Offset_Renamed + OffsetAttribute.EndOffset();
TermsHashPerField.WriteVInt(1, startOffset - postings.LastOffsets[termID]);
TermsHashPerField.WriteVInt(1, endOffset - startOffset);
postings.LastOffsets[termID] = endOffset;
}
if (DoVectorPositions)
{
BytesRef payload;
if (PayloadAttribute == null)
{
payload = null;
}
else
{
payload = PayloadAttribute.Payload;
}
int pos = FieldState.Position_Renamed - postings.LastPositions[termID];
if (payload != null && payload.Length > 0)
{
TermsHashPerField.WriteVInt(0, (pos << 1) | 1);
TermsHashPerField.WriteVInt(0, payload.Length);
TermsHashPerField.WriteBytes(0, payload.Bytes, payload.Offset, payload.Length);
HasPayloads = true;
}
else
{
TermsHashPerField.WriteVInt(0, pos << 1);
}
postings.LastPositions[termID] = FieldState.Position_Renamed;
}
}
