public override void Load()
{
TermPositions tp = null;
byte[] payloadBuffer = new byte[4]; // four bytes for an int
try
{
tp = _reader.TermPositions(_sizeTerm);
if (tp == null)
{
return;
}
while (tp.Next())
{
if (tp.Freq > 0)
{
tp.NextPosition();
tp.GetPayload(payloadBuffer, 0);
int len = BytesToInt(payloadBuffer);
Allocate(tp.Doc, Math.Min(len, _maxItems), true);
}
}
}
finally
{
if (tp != null)
{
tp.Dispose();
}
}
}
protected virtual void LoadPayload(Term term)
{
byte[] payloadBuf = null;
TermPositions tp = _reader.TermPositions();
tp.Seek(term);
while (tp.Next())
{
if (tp.Freq > 0)
{
tp.NextPosition();
if (tp.IsPayloadAvailable)
{
int len = tp.PayloadLength;
payloadBuf = tp.GetPayload(payloadBuf, 0);
Add(tp.Doc, payloadBuf, len);
}
}
}
// save the last page
while (_curSlot < MAX_SLOTS)
{
_curPage[_curSlot++] = MISSING;
}
_list[_curPageNo] = CopyPage(new int[_curData]); // optimize the page to make getNumItems work
_curPage = null;
}
// TODO: Remove warning after API has been finalized
public override ICollection <byte[]> GetPayload()
{
byte[] bytes = new byte[internalPositions.PayloadLength];
bytes = internalPositions.GetPayload(bytes, 0);
var val = new System.Collections.Generic.List <byte[]>();
val.Add(bytes);
return(val);
}
public virtual void CheckSkipTo(TermPositions tp, int target, int maxCounter)
{
tp.SkipTo(target);
Assert.Greater(maxCounter, counter, "Too many bytes read: " + counter);
Assert.AreEqual(target, tp.Doc, "Wrong document " + tp.Doc + " after skipTo target " + target);
Assert.AreEqual(1, tp.Freq, "Frequency is not 1: " + tp.Freq);
tp.NextPosition();
byte[] b = new byte[1];
tp.GetPayload(b, 0);
Assert.AreEqual((byte)target, b[0], "Wrong payload for the target " + target + ": " + b[0]);
}
protected internal virtual void ProcessPayload(Similarity similarity)
{
if (positions.IsPayloadAvailable)
{
payload = positions.GetPayload(payload, 0);
payloadScore = Enclosing_Instance.Enclosing_Instance.function.CurrentScore(doc, Enclosing_Instance.Enclosing_Instance.internalTerm.Field, spans.Start(), spans.End(), payloadsSeen, payloadScore, similarity.ScorePayload(doc, Enclosing_Instance.Enclosing_Instance.internalTerm.Field, spans.Start(), spans.End(), payload, 0, positions.PayloadLength));
payloadsSeen++;
}
else
{
// zero out the payload?
}
}
/// <summary>Process postings from multiple segments all positioned on the
/// same term. Writes out merged entries into freqOutput and
/// the proxOutput streams.
///
/// </summary>
/// <param name="smis">array of segments
/// </param>
/// <param name="n">number of cells in the array actually occupied
/// </param>
/// <returns> number of documents across all segments where this term was found
/// </returns>
/// <throws> CorruptIndexException if the index is corrupt </throws>
/// <throws> IOException if there is a low-level IO error </throws>
private int AppendPostings(FormatPostingsTermsConsumer termsConsumer, SegmentMergeInfo[] smis, int n)
{
FormatPostingsDocsConsumer docConsumer = termsConsumer.AddTerm(smis[0].term.Text);
int df = 0;
for (int i = 0; i < n; i++)
{
SegmentMergeInfo smi = smis[i];
TermPositions postings = smi.GetPositions();
System.Diagnostics.Debug.Assert(postings != null);
int base_Renamed = smi.base_Renamed;
int[] docMap = smi.GetDocMap();
postings.Seek(smi.termEnum);
while (postings.Next())
{
df++;
int doc = postings.Doc;
if (docMap != null)
{
doc = docMap[doc]; // map around deletions
}
doc += base_Renamed; // convert to merged space
int freq = postings.Freq;
FormatPostingsPositionsConsumer posConsumer = docConsumer.AddDoc(doc, freq);
if (!omitTermFreqAndPositions)
{
for (int j = 0; j < freq; j++)
{
int position = postings.NextPosition();
int payloadLength = postings.PayloadLength;
if (payloadLength > 0)
{
if (payloadBuffer == null || payloadBuffer.Length < payloadLength)
{
payloadBuffer = new byte[payloadLength];
}
postings.GetPayload(payloadBuffer, 0);
}
posConsumer.AddPosition(position, payloadBuffer, 0, payloadLength);
}
posConsumer.Finish();
}
}
}
docConsumer.Finish();
return(df);
}
public virtual void TestThreadSafety()
{
rnd = NewRandom();
int numThreads = 5;
int numDocs = 50;
ByteArrayPool pool = new ByteArrayPool(numThreads, 5);
Directory dir = new RAMDirectory();
IndexWriter writer = new IndexWriter(dir, new WhitespaceAnalyzer(), IndexWriter.MaxFieldLength.LIMITED, null);
System.String field = "test";
ThreadClass[] ingesters = new ThreadClass[numThreads];
for (int i = 0; i < numThreads; i++)
{
ingesters[i] = new AnonymousClassThread(numDocs, field, pool, writer, this);
ingesters[i].Start();
}
for (int i = 0; i < numThreads; i++)
{
ingesters[i].Join();
}
writer.Close();
IndexReader reader = IndexReader.Open(dir, true, null);
TermEnum terms = reader.Terms(null);
while (terms.Next(null))
{
TermPositions tp = reader.TermPositions(terms.Term, null);
while (tp.Next(null))
{
int freq = tp.Freq;
for (int i = 0; i < freq; i++)
{
tp.NextPosition(null);
Assert.AreEqual(pool.BytesToString(tp.GetPayload(new byte[5], 0, null)), terms.Term.Text);
}
}
tp.Close();
}
terms.Close();
reader.Close();
Assert.AreEqual(pool.Size(), numThreads);
}
// builds an index with payloads in the given Directory and performs
// different tests to verify the payload encoding
private void PerformTest(Directory dir)
{
PayloadAnalyzer analyzer = new PayloadAnalyzer();
IndexWriter writer = new IndexWriter(dir, analyzer, true, IndexWriter.MaxFieldLength.LIMITED, null);
// should be in sync with value in TermInfosWriter
int skipInterval = 16;
int numTerms = 5;
System.String fieldName = "f1";
int numDocs = skipInterval + 1;
// create content for the test documents with just a few terms
Term[] terms = GenerateTerms(fieldName, numTerms);
System.Text.StringBuilder sb = new System.Text.StringBuilder();
for (int i = 0; i < terms.Length; i++)
{
sb.Append(terms[i].Text);
sb.Append(" ");
}
System.String content = sb.ToString();
int payloadDataLength = numTerms * numDocs * 2 + numTerms * numDocs * (numDocs - 1) / 2;
byte[] payloadData = GenerateRandomData(payloadDataLength);
Document d = new Document();
d.Add(new Field(fieldName, content, Field.Store.NO, Field.Index.ANALYZED));
// add the same document multiple times to have the same payload lengths for all
// occurrences within two consecutive skip intervals
int offset = 0;
for (int i = 0; i < 2 * numDocs; i++)
{
analyzer.SetPayloadData(fieldName, payloadData, offset, 1);
offset += numTerms;
writer.AddDocument(d, null);
}
// make sure we create more than one segment to test merging
writer.Commit(null);
// now we make sure to have different payload lengths next at the next skip point
for (int i = 0; i < numDocs; i++)
{
analyzer.SetPayloadData(fieldName, payloadData, offset, i);
offset += i * numTerms;
writer.AddDocument(d, null);
}
writer.Optimize(null);
// flush
writer.Close();
/*
* Verify the index
* first we test if all payloads are stored correctly
*/
IndexReader reader = IndexReader.Open(dir, true, null);
byte[] verifyPayloadData = new byte[payloadDataLength];
offset = 0;
TermPositions[] tps = new TermPositions[numTerms];
for (int i = 0; i < numTerms; i++)
{
tps[i] = reader.TermPositions(terms[i], null);
}
while (tps[0].Next(null))
{
for (int i = 1; i < numTerms; i++)
{
tps[i].Next(null);
}
int freq = tps[0].Freq;
for (int i = 0; i < freq; i++)
{
for (int j = 0; j < numTerms; j++)
{
tps[j].NextPosition(null);
tps[j].GetPayload(verifyPayloadData, offset, null);
offset += tps[j].PayloadLength;
}
}
}
for (int i = 0; i < numTerms; i++)
{
tps[i].Close();
}
AssertByteArrayEquals(payloadData, verifyPayloadData);
/*
* test lazy skipping
*/
TermPositions tp = reader.TermPositions(terms[0], null);
tp.Next(null);
tp.NextPosition(null);
// now we don't read this payload
tp.NextPosition(null);
Assert.AreEqual(1, tp.PayloadLength, "Wrong payload length.");
byte[] payload = tp.GetPayload(null, 0, null);
Assert.AreEqual(payload[0], payloadData[numTerms]);
tp.NextPosition(null);
// we don't read this payload and skip to a different document
tp.SkipTo(5, null);
tp.NextPosition(null);
Assert.AreEqual(1, tp.PayloadLength, "Wrong payload length.");
payload = tp.GetPayload(null, 0, null);
Assert.AreEqual(payload[0], payloadData[5 * numTerms]);
/*
* Test different lengths at skip points
*/
tp.Seek(terms[1], null);
tp.Next(null);
tp.NextPosition(null);
Assert.AreEqual(1, tp.PayloadLength, "Wrong payload length.");
tp.SkipTo(skipInterval - 1, null);
tp.NextPosition(null);
Assert.AreEqual(1, tp.PayloadLength, "Wrong payload length.");
tp.SkipTo(2 * skipInterval - 1, null);
tp.NextPosition(null);
Assert.AreEqual(1, tp.PayloadLength, "Wrong payload length.");
tp.SkipTo(3 * skipInterval - 1, null);
tp.NextPosition(null);
Assert.AreEqual(3 * skipInterval - 2 * numDocs - 1, tp.PayloadLength, "Wrong payload length.");
/*
* Test multiple call of getPayload()
*/
tp.GetPayload(null, 0, null);
// it is forbidden to call getPayload() more than once
// without calling nextPosition()
Assert.Throws <IOException>(() => tp.GetPayload(null, 0, null), "Expected exception not thrown");
reader.Close();
// test long payload
analyzer = new PayloadAnalyzer();
writer = new IndexWriter(dir, analyzer, true, IndexWriter.MaxFieldLength.LIMITED, null);
System.String singleTerm = "lucene";
d = new Document();
d.Add(new Field(fieldName, singleTerm, Field.Store.NO, Field.Index.ANALYZED));
// add a payload whose length is greater than the buffer size of BufferedIndexOutput
payloadData = GenerateRandomData(2000);
analyzer.SetPayloadData(fieldName, payloadData, 100, 1500);
writer.AddDocument(d, null);
writer.Optimize(null);
// flush
writer.Close();
reader = IndexReader.Open(dir, true, null);
tp = reader.TermPositions(new Term(fieldName, singleTerm), null);
tp.Next(null);
tp.NextPosition(null);
verifyPayloadData = new byte[tp.PayloadLength];
tp.GetPayload(verifyPayloadData, 0, null);
byte[] portion = new byte[1500];
Array.Copy(payloadData, 100, portion, 0, 1500);
AssertByteArrayEquals(portion, verifyPayloadData);
reader.Close();
}
public virtual void CheckSkipTo(TermPositions tp, int target, int maxCounter)
{
tp.SkipTo(target);
if (maxCounter < counter)
{
Assert.Fail("Too many bytes read: " + counter);
}
Assert.AreEqual(target, tp.Doc(), "Wrong document " + tp.Doc() + " after skipTo target " + target);
Assert.AreEqual(1, tp.Freq(), "Frequency is not 1: " + tp.Freq());
tp.NextPosition();
byte[] b = new byte[1];
tp.GetPayload(b, 0);
Assert.AreEqual((byte) target, b[0], "Wrong payload for the target " + target + ": " + b[0]);
}
/// <summary>Process postings from multiple segments all positioned on the
/// same term. Writes out merged entries into freqOutput and
/// the proxOutput streams.
///
/// </summary>
/// <param name="smis">array of segments
/// </param>
/// <param name="n">number of cells in the array actually occupied
/// </param>
/// <returns> number of documents across all segments where this term was found
/// </returns>
/// <throws> CorruptIndexException if the index is corrupt </throws>
/// <throws> IOException if there is a low-level IO error </throws>
private int AppendPostings(SegmentMergeInfo[] smis, int n)
{
int lastDoc = 0;
int df = 0; // number of docs w/ term
skipListWriter.ResetSkip();
bool storePayloads = fieldInfos.FieldInfo(smis[0].term.field).storePayloads;
int lastPayloadLength = -1; // ensures that we write the first length
for (int i = 0; i < n; i++)
{
SegmentMergeInfo smi = smis[i];
TermPositions postings = smi.GetPositions();
System.Diagnostics.Debug.Assert(postings != null);
int base_Renamed = smi.base_Renamed;
int[] docMap = smi.GetDocMap();
postings.Seek(smi.termEnum);
while (postings.Next())
{
int doc = postings.Doc();
if (docMap != null)
{
doc = docMap[doc]; // map around deletions
}
doc += base_Renamed; // convert to merged space
if (doc < 0 || (df > 0 && doc <= lastDoc))
{
throw new CorruptIndexException("docs out of order (" + doc + " <= " + lastDoc + " )");
}
df++;
if ((df % skipInterval) == 0)
{
skipListWriter.SetSkipData(lastDoc, storePayloads, lastPayloadLength);
skipListWriter.BufferSkip(df);
}
int docCode = (doc - lastDoc) << 1; // use low bit to flag freq=1
lastDoc = doc;
int freq = postings.Freq();
if (freq == 1)
{
freqOutput.WriteVInt(docCode | 1); // write doc & freq=1
}
else
{
freqOutput.WriteVInt(docCode); // write doc
freqOutput.WriteVInt(freq); // write frequency in doc
}
/** See {@link DocumentWriter#writePostings(Posting[], String) for
* documentation about the encoding of positions and payloads
*/
int lastPosition = 0; // write position deltas
for (int j = 0; j < freq; j++)
{
int position = postings.NextPosition();
int delta = position - lastPosition;
if (storePayloads)
{
int payloadLength = postings.GetPayloadLength();
if (payloadLength == lastPayloadLength)
{
proxOutput.WriteVInt(delta * 2);
}
else
{
proxOutput.WriteVInt(delta * 2 + 1);
proxOutput.WriteVInt(payloadLength);
lastPayloadLength = payloadLength;
}
if (payloadLength > 0)
{
if (payloadBuffer == null || payloadBuffer.Length < payloadLength)
{
payloadBuffer = new byte[payloadLength];
}
postings.GetPayload(payloadBuffer, 0);
proxOutput.WriteBytes(payloadBuffer, 0, payloadLength);
}
}
else
{
proxOutput.WriteVInt(delta);
}
lastPosition = position;
}
}
}
return(df);
}
