/// <summary>
/// Returns the number of documents containing the term
/// <code>t</code>. this method returns 0 if the term or
/// field does not exists. this method does not take into
/// account deleted documents that have not yet been merged
/// away.
/// </summary>
public override sealed long TotalTermFreq(Term term)
{
Fields fields = Fields;
if (fields == null)
{
return(0);
}
Terms terms = fields.Terms(term.Field());
if (terms == null)
{
return(0);
}
TermsEnum termsEnum = terms.Iterator(null);
if (termsEnum.SeekExact(term.Bytes()))
{
return(termsEnum.TotalTermFreq());
}
else
{
return(0);
}
}
public virtual void TestEndOffsetPositionWithCachingTokenFilter()
{
Directory dir = NewDirectory();
Analyzer analyzer = new MockAnalyzer(Random());
IndexWriter w = new IndexWriter(dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer));
Document doc = new Document();
IOException priorException = null;
TokenStream stream = analyzer.TokenStream("field", new StringReader("abcd "));
try
{
stream.Reset(); // TODO: weird to reset before wrapping with CachingTokenFilter... correct?
TokenStream cachedStream = new CachingTokenFilter(stream);
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.StoreTermVectors = true;
customType.StoreTermVectorPositions = true;
customType.StoreTermVectorOffsets = true;
Field f = new Field("field", cachedStream, customType);
doc.Add(f);
doc.Add(f);
w.AddDocument(doc);
}
catch (IOException e)
{
priorException = e;
}
finally
{
IOUtils.CloseWhileHandlingException(priorException, stream);
}
w.Dispose();
IndexReader r = DirectoryReader.Open(dir);
TermsEnum termsEnum = r.GetTermVectors(0).Terms("field").Iterator(null);
Assert.IsNotNull(termsEnum.Next());
DocsAndPositionsEnum dpEnum = termsEnum.DocsAndPositions(null, null);
Assert.AreEqual(2, termsEnum.TotalTermFreq());
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(0, dpEnum.StartOffset());
Assert.AreEqual(4, dpEnum.EndOffset());
dpEnum.NextPosition();
Assert.AreEqual(8, dpEnum.StartOffset());
Assert.AreEqual(12, dpEnum.EndOffset());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
r.Dispose();
dir.Dispose();
}
public virtual void TestEndOffsetPositionStandardEmptyField2()
{
Directory dir = NewDirectory();
IndexWriter w = new IndexWriter(dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random())));
Document doc = new Document();
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.StoreTermVectors = true;
customType.StoreTermVectorPositions = true;
customType.StoreTermVectorOffsets = true;
Field f = NewField("field", "abcd", customType);
doc.Add(f);
doc.Add(NewField("field", "", customType));
Field f2 = NewField("field", "crunch", customType);
doc.Add(f2);
w.AddDocument(doc);
w.Dispose();
IndexReader r = DirectoryReader.Open(dir);
TermsEnum termsEnum = r.GetTermVectors(0).Terms("field").Iterator(null);
Assert.IsNotNull(termsEnum.Next());
DocsAndPositionsEnum dpEnum = termsEnum.DocsAndPositions(null, null);
Assert.AreEqual(1, (int)termsEnum.TotalTermFreq());
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(0, dpEnum.StartOffset());
Assert.AreEqual(4, dpEnum.EndOffset());
Assert.IsNotNull(termsEnum.Next());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(6, dpEnum.StartOffset());
Assert.AreEqual(12, dpEnum.EndOffset());
r.Dispose();
dir.Dispose();
}
public virtual void TestTermVectors()
{
Terms result = Reader.GetTermVectors(0).Terms(DocHelper.TEXT_FIELD_2_KEY);
Assert.IsNotNull(result);
Assert.AreEqual(3, result.Size());
TermsEnum termsEnum = result.Iterator(null);
while (termsEnum.Next() != null)
{
string term = termsEnum.Term().Utf8ToString();
int freq = (int)termsEnum.TotalTermFreq();
Assert.IsTrue(DocHelper.FIELD_2_TEXT.IndexOf(term) != -1);
Assert.IsTrue(freq > 0);
}
Fields results = Reader.GetTermVectors(0);
Assert.IsTrue(results != null);
Assert.AreEqual(3, results.Size, "We do not have 3 term freq vectors");
}
/// <summary>
/// checks term-level statistics
/// </summary>
public void AssertTermStatsEquals(string info, TermsEnum leftTermsEnum, TermsEnum rightTermsEnum)
{
Assert.AreEqual(leftTermsEnum.DocFreq(), rightTermsEnum.DocFreq(), info);
if (leftTermsEnum.TotalTermFreq() != -1 && rightTermsEnum.TotalTermFreq() != -1)
{
Assert.AreEqual(leftTermsEnum.TotalTermFreq(), rightTermsEnum.TotalTermFreq(), info);
}
}
public override long TotalTermFreq()
{
return(TermsEnum.TotalTermFreq());
}
public virtual void Test()
{
IList <string> postingsList = new List <string>();
int numTerms = AtLeast(300);
int maxTermsPerDoc = TestUtil.NextInt(Random(), 10, 20);
bool isSimpleText = "SimpleText".Equals(TestUtil.GetPostingsFormat("field"));
IndexWriterConfig iwc = NewIndexWriterConfig(Random(), TEST_VERSION_CURRENT, new MockAnalyzer(Random()));
if ((isSimpleText || iwc.MergePolicy is MockRandomMergePolicy) && (TEST_NIGHTLY || RANDOM_MULTIPLIER > 1))
{
// Otherwise test can take way too long (> 2 hours)
numTerms /= 2;
}
if (VERBOSE)
{
Console.WriteLine("maxTermsPerDoc=" + maxTermsPerDoc);
Console.WriteLine("numTerms=" + numTerms);
}
for (int i = 0; i < numTerms; i++)
{
string term = Convert.ToString(i);
for (int j = 0; j < i; j++)
{
postingsList.Add(term);
}
}
postingsList = CollectionsHelper.Shuffle(postingsList);
ConcurrentQueue <string> postings = new ConcurrentQueue <string>(postingsList);
Directory dir = NewFSDirectory(CreateTempDir(GetFullMethodName()));
RandomIndexWriter iw = new RandomIndexWriter(Random(), dir, iwc);
int threadCount = TestUtil.NextInt(Random(), 1, 5);
if (VERBOSE)
{
Console.WriteLine("config: " + iw.w.Config);
Console.WriteLine("threadCount=" + threadCount);
}
Field prototype = NewTextField("field", "", Field.Store.NO);
FieldType fieldType = new FieldType((FieldType)prototype.FieldType);
if (Random().NextBoolean())
{
fieldType.OmitNorms = true;
}
int options = Random().Next(3);
if (options == 0)
{
fieldType.IndexOptions = FieldInfo.IndexOptions.DOCS_AND_FREQS; // we dont actually need positions
fieldType.StoreTermVectors = true; // but enforce term vectors when we do this so we check SOMETHING
}
else if (options == 1 && !DoesntSupportOffsets.Contains(TestUtil.GetPostingsFormat("field")))
{
fieldType.IndexOptions = FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
}
// else just positions
ThreadClass[] threads = new ThreadClass[threadCount];
CountdownEvent startingGun = new CountdownEvent(1);
for (int threadID = 0; threadID < threadCount; threadID++)
{
Random threadRandom = new Random(Random().Next());
Document document = new Document();
Field field = new Field("field", "", fieldType);
document.Add(field);
threads[threadID] = new ThreadAnonymousInnerClassHelper(this, numTerms, maxTermsPerDoc, postings, iw, startingGun, threadRandom, document, field);
threads[threadID].Start();
}
startingGun.Signal();
foreach (ThreadClass t in threads)
{
t.Join();
}
iw.ForceMerge(1);
DirectoryReader ir = iw.Reader;
Assert.AreEqual(1, ir.Leaves.Count);
AtomicReader air = (AtomicReader)ir.Leaves[0].Reader;
Terms terms = air.Terms("field");
// numTerms-1 because there cannot be a term 0 with 0 postings:
Assert.AreEqual(numTerms - 1, terms.Size());
TermsEnum termsEnum = terms.Iterator(null);
BytesRef termBR;
while ((termBR = termsEnum.Next()) != null)
{
int value = Convert.ToInt32(termBR.Utf8ToString());
Assert.AreEqual(value, termsEnum.TotalTermFreq());
// don't really need to check more than this, as CheckIndex
// will verify that totalTermFreq == total number of positions seen
// from a docsAndPositionsEnum.
}
ir.Dispose();
iw.Dispose();
dir.Dispose();
}
/// <summary>
/// checks term-level statistics
/// </summary>
public virtual void AssertTermStats(TermsEnum leftTermsEnum, TermsEnum rightTermsEnum)
{
Assert.AreEqual(leftTermsEnum.DocFreq(), rightTermsEnum.DocFreq());
if (leftTermsEnum.TotalTermFreq() != -1 && rightTermsEnum.TotalTermFreq() != -1)
{
Assert.AreEqual(leftTermsEnum.TotalTermFreq(), rightTermsEnum.TotalTermFreq());
}
}
public virtual void TestDoubleOffsetCounting()
{
Directory dir = NewDirectory();
IndexWriter w = new IndexWriter(dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random())));
Document doc = new Document();
FieldType customType = new FieldType(StringField.TYPE_NOT_STORED);
customType.StoreTermVectors = true;
customType.StoreTermVectorPositions = true;
customType.StoreTermVectorOffsets = true;
Field f = NewField("field", "abcd", customType);
doc.Add(f);
doc.Add(f);
Field f2 = NewField("field", "", customType);
doc.Add(f2);
doc.Add(f);
w.AddDocument(doc);
w.Dispose();
IndexReader r = DirectoryReader.Open(dir);
Terms vector = r.GetTermVectors(0).Terms("field");
Assert.IsNotNull(vector);
TermsEnum termsEnum = vector.Iterator(null);
Assert.IsNotNull(termsEnum.Next());
Assert.AreEqual("", termsEnum.Term().Utf8ToString());
// Token "" occurred once
Assert.AreEqual(1, termsEnum.TotalTermFreq());
DocsAndPositionsEnum dpEnum = termsEnum.DocsAndPositions(null, null);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(8, dpEnum.StartOffset());
Assert.AreEqual(8, dpEnum.EndOffset());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
// Token "abcd" occurred three times
Assert.AreEqual(new BytesRef("abcd"), termsEnum.Next());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.AreEqual(3, termsEnum.TotalTermFreq());
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(0, dpEnum.StartOffset());
Assert.AreEqual(4, dpEnum.EndOffset());
dpEnum.NextPosition();
Assert.AreEqual(4, dpEnum.StartOffset());
Assert.AreEqual(8, dpEnum.EndOffset());
dpEnum.NextPosition();
Assert.AreEqual(8, dpEnum.StartOffset());
Assert.AreEqual(12, dpEnum.EndOffset());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
Assert.IsNull(termsEnum.Next());
r.Dispose();
dir.Dispose();
}
/// <summary>
/// Creates a <seealso cref="TermContext"/> from a top-level <seealso cref="IndexReaderContext"/> and the
/// given <seealso cref="Term"/>. this method will lookup the given term in all context's leaf readers
/// and register each of the readers containing the term in the returned <seealso cref="TermContext"/>
/// using the leaf reader's ordinal.
/// <p>
/// Note: the given context must be a top-level context.
/// </summary>
public static TermContext Build(IndexReaderContext context, Term term)
{
Debug.Assert(context != null && context.IsTopLevel);
string field = term.Field();
BytesRef bytes = term.Bytes();
TermContext perReaderTermState = new TermContext(context);
//if (DEBUG) System.out.println("prts.build term=" + term);
foreach (AtomicReaderContext ctx in context.Leaves)
{
//if (DEBUG) System.out.println(" r=" + leaves[i].reader);
Fields fields = ctx.AtomicReader.Fields;
if (fields != null)
{
Terms terms = fields.Terms(field);
if (terms != null)
{
TermsEnum termsEnum = terms.Iterator(null);
if (termsEnum.SeekExact(bytes))
{
TermState termState = termsEnum.TermState();
//if (DEBUG) System.out.println(" found");
perReaderTermState.Register(termState, ctx.Ord, termsEnum.DocFreq(), termsEnum.TotalTermFreq());
}
}
}
}
return(perReaderTermState);
}
public virtual void TestArbitraryFields()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random(), dir);
int NUM_DOCS = AtLeast(27);
if (VERBOSE)
{
Console.WriteLine("TEST: " + NUM_DOCS + " docs");
}
int[] fieldsPerDoc = new int[NUM_DOCS];
int baseCount = 0;
for (int docCount = 0; docCount < NUM_DOCS; docCount++)
{
int fieldCount = TestUtil.NextInt(Random(), 1, 17);
fieldsPerDoc[docCount] = fieldCount - 1;
int finalDocCount = docCount;
if (VERBOSE)
{
Console.WriteLine("TEST: " + fieldCount + " fields in doc " + docCount);
}
int finalBaseCount = baseCount;
baseCount += fieldCount - 1;
w.AddDocument(new IterableAnonymousInnerClassHelper(this, fieldCount, finalDocCount, finalBaseCount));
}
IndexReader r = w.Reader;
w.Dispose();
IndexSearcher s = NewSearcher(r);
int counter = 0;
for (int id = 0; id < NUM_DOCS; id++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: verify doc id=" + id + " (" + fieldsPerDoc[id] + " fields) counter=" + counter);
}
TopDocs hits = s.Search(new TermQuery(new Term("id", "" + id)), 1);
Assert.AreEqual(1, hits.TotalHits);
int docID = hits.ScoreDocs[0].Doc;
Document doc = s.Doc(docID);
int endCounter = counter + fieldsPerDoc[id];
while (counter < endCounter)
{
string name = "f" + counter;
int fieldID = counter % 10;
bool stored = (counter & 1) == 0 || fieldID == 3;
bool binary = fieldID == 3;
bool indexed = fieldID != 3;
string stringValue;
if (fieldID != 3 && fieldID != 9)
{
stringValue = "text " + counter;
}
else
{
stringValue = null;
}
// stored:
if (stored)
{
IndexableField f = doc.GetField(name);
Assert.IsNotNull(f, "doc " + id + " doesn't have field f" + counter);
if (binary)
{
Assert.IsNotNull(f, "doc " + id + " doesn't have field f" + counter);
BytesRef b = f.BinaryValue();
Assert.IsNotNull(b);
Assert.AreEqual(10, b.Length);
for (int idx = 0; idx < 10; idx++)
{
Assert.AreEqual((byte)(idx + counter), b.Bytes[b.Offset + idx]);
}
}
else
{
Debug.Assert(stringValue != null);
Assert.AreEqual(stringValue, f.StringValue);
}
}
if (indexed)
{
bool tv = counter % 2 == 1 && fieldID != 9;
if (tv)
{
Terms tfv = r.GetTermVectors(docID).Terms(name);
Assert.IsNotNull(tfv);
TermsEnum termsEnum = tfv.Iterator(null);
Assert.AreEqual(new BytesRef("" + counter), termsEnum.Next());
Assert.AreEqual(1, termsEnum.TotalTermFreq());
DocsAndPositionsEnum dpEnum = termsEnum.DocsAndPositions(null, null);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(1, dpEnum.Freq());
Assert.AreEqual(1, dpEnum.NextPosition());
Assert.AreEqual(new BytesRef("text"), termsEnum.Next());
Assert.AreEqual(1, termsEnum.TotalTermFreq());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(1, dpEnum.Freq());
Assert.AreEqual(0, dpEnum.NextPosition());
Assert.IsNull(termsEnum.Next());
// TODO: offsets
}
else
{
Fields vectors = r.GetTermVectors(docID);
Assert.IsTrue(vectors == null || vectors.Terms(name) == null);
}
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(new Term("id", "" + id)), BooleanClause.Occur.MUST);
bq.Add(new TermQuery(new Term(name, "text")), BooleanClause.Occur.MUST);
TopDocs hits2 = s.Search(bq, 1);
Assert.AreEqual(1, hits2.TotalHits);
Assert.AreEqual(docID, hits2.ScoreDocs[0].Doc);
bq = new BooleanQuery();
bq.Add(new TermQuery(new Term("id", "" + id)), BooleanClause.Occur.MUST);
bq.Add(new TermQuery(new Term(name, "" + counter)), BooleanClause.Occur.MUST);
TopDocs hits3 = s.Search(bq, 1);
Assert.AreEqual(1, hits3.TotalHits);
Assert.AreEqual(docID, hits3.ScoreDocs[0].Doc);
}
counter++;
}
}
r.Dispose();
dir.Dispose();
}
public virtual void TestMerge()
{
Codec codec = Codec.Default;
SegmentInfo si = new SegmentInfo(MergedDir, Constants.LUCENE_MAIN_VERSION, MergedSegment, -1, false, codec, null);
SegmentMerger merger = new SegmentMerger(Arrays.AsList <AtomicReader>(Reader1, Reader2), si, InfoStream.Default, MergedDir, IndexWriterConfig.DEFAULT_TERM_INDEX_INTERVAL, MergeState.CheckAbort.NONE, new FieldInfos.FieldNumbers(), NewIOContext(Random()), true);
MergeState mergeState = merger.Merge();
int docsMerged = mergeState.SegmentInfo.DocCount;
Assert.IsTrue(docsMerged == 2);
//Should be able to open a new SegmentReader against the new directory
SegmentReader mergedReader = new SegmentReader(new SegmentCommitInfo(new SegmentInfo(MergedDir, Constants.LUCENE_MAIN_VERSION, MergedSegment, docsMerged, false, codec, null), 0, -1L, -1L), DirectoryReader.DEFAULT_TERMS_INDEX_DIVISOR, NewIOContext(Random()));
Assert.IsTrue(mergedReader != null);
Assert.IsTrue(mergedReader.NumDocs == 2);
Document newDoc1 = mergedReader.Document(0);
Assert.IsTrue(newDoc1 != null);
//There are 2 unstored fields on the document
Assert.IsTrue(DocHelper.NumFields(newDoc1) == DocHelper.NumFields(Doc1) - DocHelper.Unstored.Count);
Document newDoc2 = mergedReader.Document(1);
Assert.IsTrue(newDoc2 != null);
Assert.IsTrue(DocHelper.NumFields(newDoc2) == DocHelper.NumFields(Doc2) - DocHelper.Unstored.Count);
DocsEnum termDocs = TestUtil.Docs(Random(), mergedReader, DocHelper.TEXT_FIELD_2_KEY, new BytesRef("field"), MultiFields.GetLiveDocs(mergedReader), null, 0);
Assert.IsTrue(termDocs != null);
Assert.IsTrue(termDocs.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int tvCount = 0;
foreach (FieldInfo fieldInfo in mergedReader.FieldInfos)
{
if (fieldInfo.HasVectors())
{
tvCount++;
}
}
//System.out.println("stored size: " + stored.Size());
Assert.AreEqual(3, tvCount, "We do not have 3 fields that were indexed with term vector");
Terms vector = mergedReader.GetTermVectors(0).Terms(DocHelper.TEXT_FIELD_2_KEY);
Assert.IsNotNull(vector);
Assert.AreEqual(3, vector.Size());
TermsEnum termsEnum = vector.Iterator(null);
int i = 0;
while (termsEnum.Next() != null)
{
string term = termsEnum.Term().Utf8ToString();
int freq = (int)termsEnum.TotalTermFreq();
//System.out.println("Term: " + term + " Freq: " + freq);
Assert.IsTrue(DocHelper.FIELD_2_TEXT.IndexOf(term) != -1);
Assert.IsTrue(DocHelper.FIELD_2_FREQS[i] == freq);
i++;
}
TestSegmentReader.CheckNorms(mergedReader);
mergedReader.Dispose();
}
public static void VerifyEquals(Fields d1, Fields d2)
{
if (d1 == null)
{
Assert.IsTrue(d2 == null || d2.Size == 0);
return;
}
Assert.IsTrue(d2 != null);
IEnumerator <string> fieldsEnum2 = d2.GetEnumerator();
foreach (string field1 in d1)
{
fieldsEnum2.MoveNext();
string field2 = fieldsEnum2.Current;
Assert.AreEqual(field1, field2);
Terms terms1 = d1.Terms(field1);
Assert.IsNotNull(terms1);
TermsEnum termsEnum1 = terms1.Iterator(null);
Terms terms2 = d2.Terms(field2);
Assert.IsNotNull(terms2);
TermsEnum termsEnum2 = terms2.Iterator(null);
DocsAndPositionsEnum dpEnum1 = null;
DocsAndPositionsEnum dpEnum2 = null;
DocsEnum dEnum1 = null;
DocsEnum dEnum2 = null;
BytesRef term1;
while ((term1 = termsEnum1.Next()) != null)
{
BytesRef term2 = termsEnum2.Next();
Assert.AreEqual(term1, term2);
Assert.AreEqual(termsEnum1.TotalTermFreq(), termsEnum2.TotalTermFreq());
dpEnum1 = termsEnum1.DocsAndPositions(null, dpEnum1);
dpEnum2 = termsEnum2.DocsAndPositions(null, dpEnum2);
if (dpEnum1 != null)
{
Assert.IsNotNull(dpEnum2);
int docID1 = dpEnum1.NextDoc();
dpEnum2.NextDoc();
// docIDs are not supposed to be equal
//int docID2 = dpEnum2.NextDoc();
//Assert.AreEqual(docID1, docID2);
Assert.IsTrue(docID1 != DocIdSetIterator.NO_MORE_DOCS);
int freq1 = dpEnum1.Freq();
int freq2 = dpEnum2.Freq();
Assert.AreEqual(freq1, freq2);
IOffsetAttribute offsetAtt1 = dpEnum1.Attributes().HasAttribute <IOffsetAttribute>() ? dpEnum1.Attributes().GetAttribute <IOffsetAttribute>() : null;
IOffsetAttribute offsetAtt2 = dpEnum2.Attributes().HasAttribute <IOffsetAttribute>() ? dpEnum2.Attributes().GetAttribute <IOffsetAttribute>() : null;
if (offsetAtt1 != null)
{
Assert.IsNotNull(offsetAtt2);
}
else
{
Assert.IsNull(offsetAtt2);
}
for (int posUpto = 0; posUpto < freq1; posUpto++)
{
int pos1 = dpEnum1.NextPosition();
int pos2 = dpEnum2.NextPosition();
Assert.AreEqual(pos1, pos2);
if (offsetAtt1 != null)
{
Assert.AreEqual(offsetAtt1.StartOffset(), offsetAtt2.StartOffset());
Assert.AreEqual(offsetAtt1.EndOffset(), offsetAtt2.EndOffset());
}
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum1.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum2.NextDoc());
}
else
{
dEnum1 = TestUtil.Docs(Random(), termsEnum1, null, dEnum1, DocsEnum.FLAG_FREQS);
dEnum2 = TestUtil.Docs(Random(), termsEnum2, null, dEnum2, DocsEnum.FLAG_FREQS);
Assert.IsNotNull(dEnum1);
Assert.IsNotNull(dEnum2);
int docID1 = dEnum1.NextDoc();
dEnum2.NextDoc();
// docIDs are not supposed to be equal
//int docID2 = dEnum2.NextDoc();
//Assert.AreEqual(docID1, docID2);
Assert.IsTrue(docID1 != DocIdSetIterator.NO_MORE_DOCS);
int freq1 = dEnum1.Freq();
int freq2 = dEnum2.Freq();
Assert.AreEqual(freq1, freq2);
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dEnum1.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dEnum2.NextDoc());
}
}
Assert.IsNull(termsEnum2.Next());
}
Assert.IsFalse(fieldsEnum2.MoveNext());
}
public virtual void VerifyEquals(DirectoryReader r1, DirectoryReader r2, string idField)
{
if (VERBOSE)
{
Console.WriteLine("\nr1 docs:");
PrintDocs(r1);
Console.WriteLine("\nr2 docs:");
PrintDocs(r2);
}
if (r1.NumDocs != r2.NumDocs)
{
Debug.Assert(false, "r1.NumDocs=" + r1.NumDocs + " vs r2.NumDocs=" + r2.NumDocs);
}
bool hasDeletes = !(r1.MaxDoc == r2.MaxDoc && r1.NumDocs == r1.MaxDoc);
int[] r2r1 = new int[r2.MaxDoc]; // r2 id to r1 id mapping
// create mapping from id2 space to id2 based on idField
Fields f1 = MultiFields.GetFields(r1);
if (f1 == null)
{
// make sure r2 is empty
Assert.IsNull(MultiFields.GetFields(r2));
return;
}
Terms terms1 = f1.Terms(idField);
if (terms1 == null)
{
Assert.IsTrue(MultiFields.GetFields(r2) == null || MultiFields.GetFields(r2).Terms(idField) == null);
return;
}
TermsEnum termsEnum = terms1.Iterator(null);
Bits liveDocs1 = MultiFields.GetLiveDocs(r1);
Bits liveDocs2 = MultiFields.GetLiveDocs(r2);
Fields fields = MultiFields.GetFields(r2);
if (fields == null)
{
// make sure r1 is in fact empty (eg has only all
// deleted docs):
Bits liveDocs = MultiFields.GetLiveDocs(r1);
DocsEnum docs = null;
while (termsEnum.Next() != null)
{
docs = TestUtil.Docs(Random(), termsEnum, liveDocs, docs, DocsEnum.FLAG_NONE);
while (docs.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
Assert.Fail("r1 is not empty but r2 is");
}
}
return;
}
Terms terms2 = fields.Terms(idField);
TermsEnum termsEnum2 = terms2.Iterator(null);
DocsEnum termDocs1 = null;
DocsEnum termDocs2 = null;
while (true)
{
BytesRef term = termsEnum.Next();
//System.out.println("TEST: match id term=" + term);
if (term == null)
{
break;
}
termDocs1 = TestUtil.Docs(Random(), termsEnum, liveDocs1, termDocs1, DocsEnum.FLAG_NONE);
if (termsEnum2.SeekExact(term))
{
termDocs2 = TestUtil.Docs(Random(), termsEnum2, liveDocs2, termDocs2, DocsEnum.FLAG_NONE);
}
else
{
termDocs2 = null;
}
if (termDocs1.NextDoc() == DocIdSetIterator.NO_MORE_DOCS)
{
// this doc is deleted and wasn't replaced
Assert.IsTrue(termDocs2 == null || termDocs2.NextDoc() == DocIdSetIterator.NO_MORE_DOCS);
continue;
}
int id1 = termDocs1.DocID();
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, termDocs1.NextDoc());
Assert.IsTrue(termDocs2.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int id2 = termDocs2.DocID();
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, termDocs2.NextDoc());
r2r1[id2] = id1;
// verify stored fields are equivalent
try
{
VerifyEquals(r1.Document(id1), r2.Document(id2));
}
catch (Exception t)
{
Console.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2 + " term=" + term);
Console.WriteLine(" d1=" + r1.Document(id1));
Console.WriteLine(" d2=" + r2.Document(id2));
throw t;
}
try
{
// verify term vectors are equivalent
VerifyEquals(r1.GetTermVectors(id1), r2.GetTermVectors(id2));
}
catch (Exception e)
{
Console.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2);
Fields tv1 = r1.GetTermVectors(id1);
Console.WriteLine(" d1=" + tv1);
if (tv1 != null)
{
DocsAndPositionsEnum dpEnum = null;
DocsEnum dEnum = null;
foreach (string field in tv1)
{
Console.WriteLine(" " + field + ":");
Terms terms3 = tv1.Terms(field);
Assert.IsNotNull(terms3);
TermsEnum termsEnum3 = terms3.Iterator(null);
BytesRef term2;
while ((term2 = termsEnum3.Next()) != null)
{
Console.WriteLine(" " + term2.Utf8ToString() + ": freq=" + termsEnum3.TotalTermFreq());
dpEnum = termsEnum3.DocsAndPositions(null, dpEnum);
if (dpEnum != null)
{
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dpEnum.Freq();
Console.WriteLine(" doc=" + dpEnum.DocID() + " freq=" + freq);
for (int posUpto = 0; posUpto < freq; posUpto++)
{
Console.WriteLine(" pos=" + dpEnum.NextPosition());
}
}
else
{
dEnum = TestUtil.Docs(Random(), termsEnum3, null, dEnum, DocsEnum.FLAG_FREQS);
Assert.IsNotNull(dEnum);
Assert.IsTrue(dEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dEnum.Freq();
Console.WriteLine(" doc=" + dEnum.DocID() + " freq=" + freq);
}
}
}
}
Fields tv2 = r2.GetTermVectors(id2);
Console.WriteLine(" d2=" + tv2);
if (tv2 != null)
{
DocsAndPositionsEnum dpEnum = null;
DocsEnum dEnum = null;
foreach (string field in tv2)
{
Console.WriteLine(" " + field + ":");
Terms terms3 = tv2.Terms(field);
Assert.IsNotNull(terms3);
TermsEnum termsEnum3 = terms3.Iterator(null);
BytesRef term2;
while ((term2 = termsEnum3.Next()) != null)
{
Console.WriteLine(" " + term2.Utf8ToString() + ": freq=" + termsEnum3.TotalTermFreq());
dpEnum = termsEnum3.DocsAndPositions(null, dpEnum);
if (dpEnum != null)
{
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dpEnum.Freq();
Console.WriteLine(" doc=" + dpEnum.DocID() + " freq=" + freq);
for (int posUpto = 0; posUpto < freq; posUpto++)
{
Console.WriteLine(" pos=" + dpEnum.NextPosition());
}
}
else
{
dEnum = TestUtil.Docs(Random(), termsEnum3, null, dEnum, DocsEnum.FLAG_FREQS);
Assert.IsNotNull(dEnum);
Assert.IsTrue(dEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dEnum.Freq();
Console.WriteLine(" doc=" + dEnum.DocID() + " freq=" + freq);
}
}
}
}
throw e;
}
}
//System.out.println("TEST: done match id");
// Verify postings
//System.out.println("TEST: create te1");
Fields fields1 = MultiFields.GetFields(r1);
IEnumerator <string> fields1Enum = fields1.GetEnumerator();
Fields fields2 = MultiFields.GetFields(r2);
IEnumerator <string> fields2Enum = fields2.GetEnumerator();
string field1 = null, field2 = null;
TermsEnum termsEnum1 = null;
termsEnum2 = null;
DocsEnum docs1 = null, docs2 = null;
// pack both doc and freq into single element for easy sorting
long[] info1 = new long[r1.NumDocs];
long[] info2 = new long[r2.NumDocs];
for (; ;)
{
BytesRef term1 = null, term2 = null;
// iterate until we get some docs
int len1;
for (; ;)
{
len1 = 0;
if (termsEnum1 == null)
{
if (!fields1Enum.MoveNext())
{
break;
}
field1 = fields1Enum.Current;
Terms terms = fields1.Terms(field1);
if (terms == null)
{
continue;
}
termsEnum1 = terms.Iterator(null);
}
term1 = termsEnum1.Next();
if (term1 == null)
{
// no more terms in this field
termsEnum1 = null;
continue;
}
//System.out.println("TEST: term1=" + term1);
docs1 = TestUtil.Docs(Random(), termsEnum1, liveDocs1, docs1, DocsEnum.FLAG_FREQS);
while (docs1.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
int d = docs1.DocID();
int f = docs1.Freq();
info1[len1] = (((long)d) << 32) | f;
len1++;
}
if (len1 > 0)
{
break;
}
}
// iterate until we get some docs
int len2;
for (; ;)
{
len2 = 0;
if (termsEnum2 == null)
{
if (!fields2Enum.MoveNext())
{
break;
}
field2 = fields2Enum.Current;
Terms terms = fields2.Terms(field2);
if (terms == null)
{
continue;
}
termsEnum2 = terms.Iterator(null);
}
term2 = termsEnum2.Next();
if (term2 == null)
{
// no more terms in this field
termsEnum2 = null;
continue;
}
//System.out.println("TEST: term1=" + term1);
docs2 = TestUtil.Docs(Random(), termsEnum2, liveDocs2, docs2, DocsEnum.FLAG_FREQS);
while (docs2.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
int d = r2r1[docs2.DocID()];
int f = docs2.Freq();
info2[len2] = (((long)d) << 32) | f;
len2++;
}
if (len2 > 0)
{
break;
}
}
Assert.AreEqual(len1, len2);
if (len1 == 0) // no more terms
{
break;
}
Assert.AreEqual(field1, field2);
Assert.IsTrue(term1.BytesEquals(term2));
if (!hasDeletes)
{
Assert.AreEqual(termsEnum1.DocFreq(), termsEnum2.DocFreq());
}
Assert.AreEqual(term1, term2, "len1=" + len1 + " len2=" + len2 + " deletes?=" + hasDeletes);
// sort info2 to get it into ascending docid
Array.Sort(info2, 0, len2);
// now compare
for (int i = 0; i < len1; i++)
{
Assert.AreEqual(info1[i], info2[i], "i=" + i + " len=" + len1 + " d1=" + ((long)((ulong)info1[i] >> 32)) + " f1=" + (info1[i] & int.MaxValue) + " d2=" + ((long)((ulong)info2[i] >> 32)) + " f2=" + (info2[i] & int.MaxValue) + " field=" + field1 + " term=" + term1.Utf8ToString());
}
}
}
public override long TotalTermFreq()
{
return(@in.TotalTermFreq());
}