public virtual void Test()
{
Random random = new J2N.Randomizer(Random.NextInt64());
LineFileDocs docs = new LineFileDocs(random, DefaultCodecSupportsDocValues);
Directory d = NewDirectory();
MockAnalyzer analyzer = new MockAnalyzer(LuceneTestCase.Random);
analyzer.MaxTokenLength = TestUtil.NextInt32(LuceneTestCase.Random, 1, IndexWriter.MAX_TERM_LENGTH);
RandomIndexWriter w = new RandomIndexWriter(LuceneTestCase.Random, d, analyzer);
int numDocs = AtLeast(10);
for (int docCount = 0; docCount < numDocs; docCount++)
{
w.AddDocument(docs.NextDoc());
}
IndexReader r = w.GetReader();
w.Dispose();
JCG.List <BytesRef> terms = new JCG.List <BytesRef>();
TermsEnum termsEnum = MultiFields.GetTerms(r, "body").GetEnumerator();
while (termsEnum.MoveNext())
{
terms.Add(BytesRef.DeepCopyOf(termsEnum.Term));
}
if (Verbose)
{
Console.WriteLine("TEST: " + terms.Count + " terms");
}
int upto = -1;
int iters = AtLeast(200);
for (int iter = 0; iter < iters; iter++)
{
bool isEnd;
if (upto != -1 && LuceneTestCase.Random.NextBoolean())
{
// next
if (Verbose)
{
Console.WriteLine("TEST: iter next");
}
isEnd = termsEnum.MoveNext() == false;
upto++;
if (isEnd)
{
if (Verbose)
{
Console.WriteLine(" end");
}
Assert.AreEqual(upto, terms.Count);
upto = -1;
}
else
{
if (Verbose)
{
Console.WriteLine(" got term=" + termsEnum.Term.Utf8ToString() + " expected=" + terms[upto].Utf8ToString());
}
Assert.IsTrue(upto < terms.Count);
Assert.AreEqual(terms[upto], termsEnum.Term);
}
}
else
{
BytesRef target;
string exists;
if (LuceneTestCase.Random.NextBoolean())
{
// likely fake term
if (LuceneTestCase.Random.NextBoolean())
{
target = new BytesRef(TestUtil.RandomSimpleString(LuceneTestCase.Random));
}
else
{
target = new BytesRef(TestUtil.RandomRealisticUnicodeString(LuceneTestCase.Random));
}
exists = "likely not";
}
else
{
// real term
target = terms[LuceneTestCase.Random.Next(terms.Count)];
exists = "yes";
}
upto = terms.BinarySearch(target);
if (LuceneTestCase.Random.NextBoolean())
{
if (Verbose)
{
Console.WriteLine("TEST: iter seekCeil target=" + target.Utf8ToString() + " exists=" + exists);
}
// seekCeil
TermsEnum.SeekStatus status = termsEnum.SeekCeil(target);
if (Verbose)
{
Console.WriteLine(" got " + status);
}
if (upto < 0)
{
upto = -(upto + 1);
if (upto >= terms.Count)
{
Assert.AreEqual(TermsEnum.SeekStatus.END, status);
upto = -1;
}
else
{
Assert.AreEqual(TermsEnum.SeekStatus.NOT_FOUND, status);
Assert.AreEqual(terms[upto], termsEnum.Term);
}
}
else
{
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, status);
Assert.AreEqual(terms[upto], termsEnum.Term);
}
}
else
{
if (Verbose)
{
Console.WriteLine("TEST: iter seekExact target=" + target.Utf8ToString() + " exists=" + exists);
}
// seekExact
bool result = termsEnum.SeekExact(target);
if (Verbose)
{
Console.WriteLine(" got " + result);
}
if (upto < 0)
{
Assert.IsFalse(result);
upto = -1;
}
else
{
Assert.IsTrue(result);
Assert.AreEqual(target, termsEnum.Term);
}
}
}
}
r.Dispose();
d.Dispose();
docs.Dispose();
}
public virtual void TestIntersectEmptyString()
{
Directory dir = NewDirectory();
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random));
iwc.SetMergePolicy(new LogDocMergePolicy());
RandomIndexWriter w = new RandomIndexWriter(Random, dir, iwc);
Document doc = new Document();
doc.Add(NewStringField("field", "", Field.Store.NO));
doc.Add(NewStringField("field", "abc", Field.Store.NO));
w.AddDocument(doc);
doc = new Document();
// add empty string to both documents, so that singletonDocID == -1.
// For a FST-based term dict, we'll expect to see the first arc is
// flaged with HAS_FINAL_OUTPUT
doc.Add(NewStringField("field", "abc", Field.Store.NO));
doc.Add(NewStringField("field", "", Field.Store.NO));
w.AddDocument(doc);
w.ForceMerge(1);
DirectoryReader r = w.GetReader();
w.Dispose();
AtomicReader sub = GetOnlySegmentReader(r);
Terms terms = sub.Fields.GetTerms("field");
Automaton automaton = (new RegExp(".*", RegExpSyntax.NONE)).ToAutomaton(); // accept ALL
CompiledAutomaton ca = new CompiledAutomaton(automaton, false, false);
TermsEnum te = terms.Intersect(ca, null);
DocsEnum de;
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("", te.Term.Utf8ToString());
de = te.Docs(null, null, DocsFlags.NONE);
Assert.AreEqual(0, de.NextDoc());
Assert.AreEqual(1, de.NextDoc());
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("abc", te.Term.Utf8ToString());
de = te.Docs(null, null, DocsFlags.NONE);
Assert.AreEqual(0, de.NextDoc());
Assert.AreEqual(1, de.NextDoc());
Assert.IsFalse(te.MoveNext());
// pass empty string
te = terms.Intersect(ca, new BytesRef(""));
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("abc", te.Term.Utf8ToString());
de = te.Docs(null, null, DocsFlags.NONE);
Assert.AreEqual(0, de.NextDoc());
Assert.AreEqual(1, de.NextDoc());
Assert.IsFalse(te.MoveNext());
r.Dispose();
dir.Dispose();
}
public virtual void TestIntersectRandom()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random, dir);
int numTerms = AtLeast(300);
//final int numTerms = 50;
ISet <string> terms = new JCG.HashSet <string>();
ICollection <string> pendingTerms = new JCG.List <string>();
IDictionary <BytesRef, int> termToID = new Dictionary <BytesRef, int>();
int id = 0;
while (terms.Count != numTerms)
{
string s = RandomString;
if (!terms.Contains(s))
{
terms.Add(s);
pendingTerms.Add(s);
if (Random.Next(20) == 7)
{
AddDoc(w, pendingTerms, termToID, id++);
}
}
}
AddDoc(w, pendingTerms, termToID, id++);
BytesRef[] termsArray = new BytesRef[terms.Count];
ISet <BytesRef> termsSet = new JCG.HashSet <BytesRef>();
{
int upto = 0;
foreach (string s in terms)
{
BytesRef b = new BytesRef(s);
termsArray[upto++] = b;
termsSet.Add(b);
}
Array.Sort(termsArray);
}
if (Verbose)
{
Console.WriteLine("\nTEST: indexed terms (unicode order):");
foreach (BytesRef t in termsArray)
{
Console.WriteLine(" " + t.Utf8ToString() + " -> id:" + termToID[t]);
}
}
IndexReader r = w.GetReader();
w.Dispose();
// NOTE: intentional insanity!!
FieldCache.Int32s docIDToID = FieldCache.DEFAULT.GetInt32s(SlowCompositeReaderWrapper.Wrap(r), "id", false);
for (int iter = 0; iter < 10 * RandomMultiplier; iter++)
{
// TODO: can we also test infinite As here...?
// From the random terms, pick some ratio and compile an
// automaton:
ISet <string> acceptTerms = new JCG.HashSet <string>();
JCG.SortedSet <BytesRef> sortedAcceptTerms = new JCG.SortedSet <BytesRef>();
double keepPct = Random.NextDouble();
Automaton a;
if (iter == 0)
{
if (Verbose)
{
Console.WriteLine("\nTEST: empty automaton");
}
a = BasicAutomata.MakeEmpty();
}
else
{
if (Verbose)
{
Console.WriteLine("\nTEST: keepPct=" + keepPct);
}
foreach (string s in terms)
{
string s2;
if (Random.NextDouble() <= keepPct)
{
s2 = s;
}
else
{
s2 = RandomString;
}
acceptTerms.Add(s2);
sortedAcceptTerms.Add(new BytesRef(s2));
}
a = BasicAutomata.MakeStringUnion(sortedAcceptTerms);
}
if (Random.NextBoolean())
{
if (Verbose)
{
Console.WriteLine("TEST: reduce the automaton");
}
a.Reduce();
}
CompiledAutomaton c = new CompiledAutomaton(a, true, false);
BytesRef[] acceptTermsArray = new BytesRef[acceptTerms.Count];
ISet <BytesRef> acceptTermsSet = new JCG.HashSet <BytesRef>();
int upto = 0;
foreach (string s in acceptTerms)
{
BytesRef b = new BytesRef(s);
acceptTermsArray[upto++] = b;
acceptTermsSet.Add(b);
Assert.IsTrue(Accepts(c, b));
}
Array.Sort(acceptTermsArray);
if (Verbose)
{
Console.WriteLine("\nTEST: accept terms (unicode order):");
foreach (BytesRef t in acceptTermsArray)
{
Console.WriteLine(" " + t.Utf8ToString() + (termsSet.Contains(t) ? " (exists)" : ""));
}
Console.WriteLine(a.ToDot());
}
for (int iter2 = 0; iter2 < 100; iter2++)
{
BytesRef startTerm = acceptTermsArray.Length == 0 || Random.NextBoolean() ? null : acceptTermsArray[Random.Next(acceptTermsArray.Length)];
if (Verbose)
{
Console.WriteLine("\nTEST: iter2=" + iter2 + " startTerm=" + (startTerm is null ? "<null>" : startTerm.Utf8ToString()));
if (startTerm != null)
{
int state = c.RunAutomaton.InitialState;
for (int idx = 0; idx < startTerm.Length; idx++)
{
int label = startTerm.Bytes[startTerm.Offset + idx] & 0xff;
Console.WriteLine(" state=" + state + " label=" + label);
state = c.RunAutomaton.Step(state, label);
Assert.IsTrue(state != -1);
}
Console.WriteLine(" state=" + state);
}
}
TermsEnum te = MultiFields.GetTerms(r, "f").Intersect(c, startTerm);
int loc;
if (startTerm is null)
{
loc = 0;
}
else
{
loc = Array.BinarySearch(termsArray, BytesRef.DeepCopyOf(startTerm));
if (loc < 0)
{
loc = -(loc + 1);
}
else
{
// startTerm exists in index
loc++;
}
}
while (loc < termsArray.Length && !acceptTermsSet.Contains(termsArray[loc]))
{
loc++;
}
DocsEnum docsEnum = null;
while (loc < termsArray.Length)
{
BytesRef expected = termsArray[loc];
Assert.IsTrue(te.MoveNext());
BytesRef actual = te.Term;
if (Verbose)
{
Console.WriteLine("TEST: next() expected=" + expected.Utf8ToString() + " actual=" + (actual is null ? "null" : actual.Utf8ToString()));
}
Assert.AreEqual(expected, actual);
Assert.AreEqual(1, te.DocFreq);
docsEnum = TestUtil.Docs(Random, te, null, docsEnum, DocsFlags.NONE);
int docID = docsEnum.NextDoc();
Assert.IsTrue(docID != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(docIDToID.Get(docID), (int)termToID[expected]);
do
{
loc++;
} while (loc < termsArray.Length && !acceptTermsSet.Contains(termsArray[loc]));
}
Assert.IsFalse(te.MoveNext());
}
}
r.Dispose();
dir.Dispose();
}
public override void Run()
{
if (Verbose)
{
Console.WriteLine(Thread.CurrentThread.Name + ": launch search thread");
}
while (J2N.Time.NanoTime() / J2N.Time.MillisecondsPerNanosecond < stopTimeMS) // LUCENENET: Use NanoTime() rather than CurrentTimeMilliseconds() for more accurate/reliable results
{
try
{
IndexSearcher s = outerInstance.GetCurrentSearcher();
try
{
// Verify 1) IW is correctly setting
// diagnostics, and 2) segment warming for
// merged segments is actually happening:
foreach (AtomicReaderContext sub in s.IndexReader.Leaves)
{
SegmentReader segReader = (SegmentReader)sub.Reader;
IDictionary <string, string> diagnostics = segReader.SegmentInfo.Info.Diagnostics;
assertNotNull(diagnostics);
diagnostics.TryGetValue("source", out string source);
assertNotNull(source);
if (source.Equals("merge", StringComparison.Ordinal))
{
#if !FEATURE_CONDITIONALWEAKTABLE_ADDORUPDATE
UninterruptableMonitor.Enter(outerInstance.warmedLock);
try
{
#endif
assertTrue("sub reader " + sub + " wasn't warmed: warmed=" + outerInstance.warmed + " diagnostics=" + diagnostics + " si=" + segReader.SegmentInfo,
// LUCENENET: ConditionalWeakTable doesn't have ContainsKey, so we normalize to TryGetValue
!outerInstance.m_assertMergedSegmentsWarmed || outerInstance.warmed.TryGetValue(segReader.core, out BooleanRef _));
#if !FEATURE_CONDITIONALWEAKTABLE_ADDORUPDATE
}
finally
{
UninterruptableMonitor.Exit(outerInstance.warmedLock);
}
#endif
}
}
if (s.IndexReader.NumDocs > 0)
{
outerInstance.SmokeTestSearcher(s);
Fields fields = MultiFields.GetFields(s.IndexReader);
if (fields is null)
{
continue;
}
Terms terms = fields.GetTerms("body");
if (terms is null)
{
continue;
}
TermsEnum termsEnum = terms.GetEnumerator();
int seenTermCount = 0;
int shift;
int trigger;
if (totTermCount < 30)
{
shift = 0;
trigger = 1;
}
else
{
trigger = totTermCount / 30;
shift = Random.Next(trigger);
}
while (J2N.Time.NanoTime() / J2N.Time.MillisecondsPerNanosecond < stopTimeMS) // LUCENENET: Use NanoTime() rather than CurrentTimeMilliseconds() for more accurate/reliable results
{
if (!termsEnum.MoveNext())
{
totTermCount.Value = seenTermCount;
break;
}
seenTermCount++;
// search 30 terms
if ((seenTermCount + shift) % trigger == 0)
{
//if (VERBOSE) {
//System.out.println(Thread.currentThread().getName() + " now search body:" + term.Utf8ToString());
//}
totHits.AddAndGet(outerInstance.RunQuery(s, new TermQuery(new Term("body", termsEnum.Term))));
}
}
//if (VERBOSE) {
//System.out.println(Thread.currentThread().getName() + ": search done");
//}
}
}
finally
{
outerInstance.ReleaseSearcher(s);
}
}
catch (Exception t) when(t.IsThrowable())
{
Console.WriteLine(Thread.CurrentThread.Name + ": hit exc");
outerInstance.m_failed.Value = (true);
Console.WriteLine(t.ToString());
throw RuntimeException.Create(t);
}
}
}
public virtual void TestPositionReader()
{
TermVectorsReader reader = Codec.Default.TermVectorsFormat.VectorsReader(dir, seg.Info, fieldInfos, NewIOContext(Random));
//BytesRef[] terms; // LUCENENET NOTE: Not used in Lucene
Terms vector = reader.Get(0).GetTerms(testFields[0]);
Assert.IsNotNull(vector);
Assert.AreEqual(testTerms.Length, vector.Count);
TermsEnum termsEnum = vector.GetEnumerator();
DocsAndPositionsEnum dpEnum = null;
for (int i = 0; i < testTerms.Length; i++)
{
Assert.IsTrue(termsEnum.MoveNext());
BytesRef text = termsEnum.Term;
string term = text.Utf8ToString();
//System.out.println("Term: " + term);
Assert.AreEqual(testTerms[i], term);
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.IsNotNull(dpEnum);
int doc = dpEnum.DocID;
Assert.AreEqual(-1, doc);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(dpEnum.Freq, positions[i].Length);
for (int j = 0; j < positions[i].Length; j++)
{
Assert.AreEqual(positions[i][j], dpEnum.NextPosition());
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
doc = dpEnum.DocID;
Assert.AreEqual(-1, doc);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.IsNotNull(dpEnum);
Assert.AreEqual(dpEnum.Freq, positions[i].Length);
for (int j = 0; j < positions[i].Length; j++)
{
Assert.AreEqual(positions[i][j], dpEnum.NextPosition());
Assert.AreEqual(j * 10, dpEnum.StartOffset);
Assert.AreEqual(j * 10 + testTerms[i].Length, dpEnum.EndOffset);
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
}
Terms freqVector = reader.Get(0).GetTerms(testFields[1]); //no pos, no offset
Assert.IsNotNull(freqVector);
Assert.AreEqual(testTerms.Length, freqVector.Count);
termsEnum = freqVector.GetEnumerator();
Assert.IsNotNull(termsEnum);
for (int i = 0; i < testTerms.Length; i++)
{
Assert.IsTrue(termsEnum.MoveNext());
BytesRef text = termsEnum.Term;
string term = text.Utf8ToString();
//System.out.println("Term: " + term);
Assert.AreEqual(testTerms[i], term);
Assert.IsNotNull(termsEnum.Docs(null, null));
Assert.IsNull(termsEnum.DocsAndPositions(null, null)); // no pos
}
reader.Dispose();
}
public override bool MoveNext()
{
//System.out.println("FTE.next doSeek=" + doSeek);
//new Throwable().printStackTrace(System.out);
for (; ;)
{
// Seek or forward the iterator
if (doSeek)
{
doSeek = false;
BytesRef t = NextSeekTerm(actualTerm);
//System.out.println(" seek to t=" + (t == null ? "null" : t.utf8ToString()) + " tenum=" + tenum);
// Make sure we always seek forward:
if (Debugging.AssertsEnabled)
{
Debugging.Assert(actualTerm == null || t == null || Comparer.Compare(t, actualTerm) > 0, () => "curTerm=" + actualTerm + " seekTerm=" + t);
}
if (t == null || tenum.SeekCeil(t) == SeekStatus.END)
{
// no more terms to seek to or enum exhausted
//System.out.println(" return null");
return(false);
}
actualTerm = tenum.Term;
//System.out.println(" got term=" + actualTerm.utf8ToString());
}
else
{
if (tenum.MoveNext())
{
actualTerm = tenum.Term;
}
else
{
// enum exhausted
actualTerm = null;
return(false);
}
}
// check if term is accepted
switch (Accept(actualTerm))
{
case FilteredTermsEnum.AcceptStatus.YES_AND_SEEK:
doSeek = true;
// term accepted, but we need to seek so fall-through
goto case FilteredTermsEnum.AcceptStatus.YES;
case FilteredTermsEnum.AcceptStatus.YES:
// term accepted
return(true);
case FilteredTermsEnum.AcceptStatus.NO_AND_SEEK:
// invalid term, seek next time
doSeek = true;
break;
case FilteredTermsEnum.AcceptStatus.END:
// we are supposed to end the enum
return(false);
}
}
}
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)
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(false, "r1.NumDocs={0} vs r2.NumDocs={1}", r1.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.GetTerms(idField);
if (terms1 == null)
{
Assert.IsTrue(MultiFields.GetFields(r2) == null || MultiFields.GetFields(r2).GetTerms(idField) == null);
return;
}
TermsEnum termsEnum = terms1.GetEnumerator();
IBits liveDocs1 = MultiFields.GetLiveDocs(r1);
IBits 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):
IBits liveDocs = MultiFields.GetLiveDocs(r1);
DocsEnum docs = null;
while (termsEnum.MoveNext())
{
docs = TestUtil.Docs(Random, termsEnum, liveDocs, docs, DocsFlags.NONE);
while (docs.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
Assert.Fail("r1 is not empty but r2 is");
}
}
return;
}
Terms terms2 = fields.GetTerms(idField);
TermsEnum termsEnum2 = terms2.GetEnumerator();
DocsEnum termDocs1 = null;
DocsEnum termDocs2 = null;
while (termsEnum.MoveNext())
{
BytesRef term = termsEnum.Term;
//System.out.println("TEST: match id term=" + term);
termDocs1 = TestUtil.Docs(Random, termsEnum, liveDocs1, termDocs1, DocsFlags.NONE);
if (termsEnum2.SeekExact(term))
{
termDocs2 = TestUtil.Docs(Random, termsEnum2, liveDocs2, termDocs2, DocsFlags.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) when(t.IsThrowable())
{
Console.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2 + " term=" + term);
Console.WriteLine(" d1=" + r1.Document(id1));
Console.WriteLine(" d2=" + r2.Document(id2));
throw; // LUCENENET: CA2200: Rethrow to preserve stack details (https://docs.microsoft.com/en-us/visualstudio/code-quality/ca2200-rethrow-to-preserve-stack-details)
}
try
{
// verify term vectors are equivalent
VerifyEquals(r1.GetTermVectors(id1), r2.GetTermVectors(id2));
}
catch (Exception e) when(e.IsThrowable())
{
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.GetTerms(field);
Assert.IsNotNull(terms3);
TermsEnum termsEnum3 = terms3.GetEnumerator();
while (termsEnum3.MoveNext())
{
Console.WriteLine(" " + termsEnum3.Term.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, DocsFlags.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.GetTerms(field);
Assert.IsNotNull(terms3);
TermsEnum termsEnum3 = terms3.GetEnumerator();
while (termsEnum3.MoveNext())
{
Console.WriteLine(" " + termsEnum3.Term.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, DocsFlags.FREQS);
Assert.IsNotNull(dEnum);
Assert.IsTrue(dEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dEnum.Freq;
Console.WriteLine(" doc=" + dEnum.DocID + " freq=" + freq);
}
}
}
}
throw; // LUCENENET: CA2200: Rethrow to preserve stack details (https://docs.microsoft.com/en-us/visualstudio/code-quality/ca2200-rethrow-to-preserve-stack-details)
}
}
//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.GetTerms(field1);
if (terms == null)
{
continue;
}
termsEnum1 = terms.GetEnumerator();
}
if (!termsEnum1.MoveNext())
{
term1 = null;
// no more terms in this field
termsEnum1 = null;
continue;
}
term1 = termsEnum1.Term;
//System.out.println("TEST: term1=" + term1);
docs1 = TestUtil.Docs(Random, termsEnum1, liveDocs1, docs1, DocsFlags.FREQS);
while (docs1.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
int d = docs1.DocID;
int f = docs1.Freq;
info1[len1] = (((long)d) << 32) | (uint)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.GetTerms(field2);
if (terms == null)
{
continue;
}
termsEnum2 = terms.GetEnumerator();
}
if (!termsEnum2.MoveNext())
{
term2 = null;
// no more terms in this field
termsEnum2 = null;
continue;
}
term2 = termsEnum2.Term;
//System.out.println("TEST: term1=" + term1);
docs2 = TestUtil.Docs(Random, termsEnum2, liveDocs2, docs2, DocsFlags.FREQS);
while (docs2.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
int d = r2r1[docs2.DocID];
int f = docs2.Freq;
info2[len2] = (((long)d) << 32) | (uint)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=" + (info1[i].TripleShift(32)) + " f1=" + (info1[i] & int.MaxValue) + " d2=" + (info2[i].TripleShift(32)) + " f2=" + (info2[i] & int.MaxValue) + " field=" + field1 + " term=" + term1.Utf8ToString());
}
}
}
public override bool MoveNext()
{
return(m_input.MoveNext());
}
/// <summary>
/// Call this only once (if you subclass!) </summary>
protected virtual void Uninvert(AtomicReader reader, IBits liveDocs, BytesRef termPrefix)
{
FieldInfo info = reader.FieldInfos.FieldInfo(m_field);
if (info != null && info.HasDocValues)
{
throw IllegalStateException.Create("Type mismatch: " + m_field + " was indexed as " + info.DocValuesType);
}
//System.out.println("DTO uninvert field=" + field + " prefix=" + termPrefix);
long startTime = J2N.Time.NanoTime() / J2N.Time.MillisecondsPerNanosecond; // LUCENENET: Use NanoTime() rather than CurrentTimeMilliseconds() for more accurate/reliable results
m_prefix = termPrefix is null ? null : BytesRef.DeepCopyOf(termPrefix);
int maxDoc = reader.MaxDoc;
int[] index = new int[maxDoc]; // immediate term numbers, or the index into the byte[] representing the last number
int[] lastTerm = new int[maxDoc]; // last term we saw for this document
var bytes = new sbyte[maxDoc][]; // list of term numbers for the doc (delta encoded vInts)
Fields fields = reader.Fields;
if (fields is null)
{
// No terms
return;
}
Terms terms = fields.GetTerms(m_field);
if (terms is null)
{
// No terms
return;
}
TermsEnum te = terms.GetEnumerator();
BytesRef seekStart = termPrefix ?? new BytesRef();
//System.out.println("seekStart=" + seekStart.utf8ToString());
if (te.SeekCeil(seekStart) == TermsEnum.SeekStatus.END)
{
// No terms match
return;
}
// If we need our "term index wrapper", these will be
// init'd below:
IList <BytesRef> indexedTerms = null;
PagedBytes indexedTermsBytes = null;
bool testedOrd = false;
// we need a minimum of 9 bytes, but round up to 12 since the space would
// be wasted with most allocators anyway.
var tempArr = new sbyte[12];
//
// enumerate all terms, and build an intermediate form of the un-inverted field.
//
// During this intermediate form, every document has a (potential) byte[]
// and the int[maxDoc()] array either contains the termNumber list directly
// or the *end* offset of the termNumber list in it's byte array (for faster
// appending and faster creation of the final form).
//
// idea... if things are too large while building, we could do a range of docs
// at a time (but it would be a fair amount slower to build)
// could also do ranges in parallel to take advantage of multiple CPUs
// OPTIONAL: remap the largest df terms to the lowest 128 (single byte)
// values. this requires going over the field first to find the most
// frequent terms ahead of time.
int termNum = 0;
m_docsEnum = null;
// Loop begins with te positioned to first term (we call
// seek above):
for (; ;)
{
BytesRef t = te.Term;
if (t is null || (termPrefix != null && !StringHelper.StartsWith(t, termPrefix)))
{
break;
}
//System.out.println("visit term=" + t.utf8ToString() + " " + t + " termNum=" + termNum);
if (!testedOrd)
{
try
{
m_ordBase = (int)te.Ord;
//System.out.println("got ordBase=" + ordBase);
}
catch (Exception uoe) when(uoe.IsUnsupportedOperationException())
{
// Reader cannot provide ord support, so we wrap
// our own support by creating our own terms index:
indexedTerms = new JCG.List <BytesRef>();
indexedTermsBytes = new PagedBytes(15);
//System.out.println("NO ORDS");
}
testedOrd = true;
}
VisitTerm(te, termNum);
if (indexedTerms != null && (termNum & indexIntervalMask) == 0)
{
// Index this term
m_sizeOfIndexedStrings += t.Length;
BytesRef indexedTerm = new BytesRef();
indexedTermsBytes.Copy(t, indexedTerm);
// TODO: really should 1) strip off useless suffix,
// and 2) use FST not array/PagedBytes
indexedTerms.Add(indexedTerm);
}
int df = te.DocFreq;
if (df <= m_maxTermDocFreq)
{
m_docsEnum = te.Docs(liveDocs, m_docsEnum, DocsFlags.NONE);
// dF, but takes deletions into account
int actualDF = 0;
for (; ;)
{
int doc = m_docsEnum.NextDoc();
if (doc == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
//System.out.println(" chunk=" + chunk + " docs");
actualDF++;
m_termInstances++;
//System.out.println(" docID=" + doc);
// add TNUM_OFFSET to the term number to make room for special reserved values:
// 0 (end term) and 1 (index into byte array follows)
int delta = termNum - lastTerm[doc] + TNUM_OFFSET;
lastTerm[doc] = termNum;
int val = index[doc];
if ((val & 0xff) == 1)
{
// index into byte array (actually the end of
// the doc-specific byte[] when building)
int pos = val.TripleShift(8);
int ilen = VInt32Size(delta);
var arr = bytes[doc];
int newend = pos + ilen;
if (newend > arr.Length)
{
// We avoid a doubling strategy to lower memory usage.
// this faceting method isn't for docs with many terms.
// In hotspot, objects have 2 words of overhead, then fields, rounded up to a 64-bit boundary.
// TODO: figure out what array lengths we can round up to w/o actually using more memory
// (how much space does a byte[] take up? Is data preceded by a 32 bit length only?
// It should be safe to round up to the nearest 32 bits in any case.
int newLen = (newend + 3) & unchecked ((int)0xfffffffc); // 4 byte alignment
var newarr = new sbyte[newLen];
Array.Copy(arr, 0, newarr, 0, pos);
arr = newarr;
bytes[doc] = newarr;
}
pos = WriteInt32(delta, arr, pos);
index[doc] = (pos << 8) | 1; // update pointer to end index in byte[]
}
else
{
// OK, this int has data in it... find the end (a zero starting byte - not
// part of another number, hence not following a byte with the high bit set).
int ipos;
if (val == 0)
{
ipos = 0;
}
else if ((val & 0x0000ff80) == 0)
{
ipos = 1;
}
else if ((val & 0x00ff8000) == 0)
{
ipos = 2;
}
else if ((val & 0xff800000) == 0)
{
ipos = 3;
}
else
{
ipos = 4;
}
//System.out.println(" ipos=" + ipos);
int endPos = WriteInt32(delta, tempArr, ipos);
//System.out.println(" endpos=" + endPos);
if (endPos <= 4)
{
//System.out.println(" fits!");
// value will fit in the integer... move bytes back
for (int j = ipos; j < endPos; j++)
{
val |= (tempArr[j] & 0xff) << (j << 3);
}
index[doc] = val;
}
else
{
// value won't fit... move integer into byte[]
for (int j = 0; j < ipos; j++)
{
tempArr[j] = (sbyte)val;
val = val.TripleShift(8);
}
// point at the end index in the byte[]
index[doc] = (endPos << 8) | 1;
bytes[doc] = tempArr;
tempArr = new sbyte[12];
}
}
}
SetActualDocFreq(termNum, actualDF);
}
termNum++;
if (!te.MoveNext())
{
break;
}
}
m_numTermsInField = termNum;
long midPoint = J2N.Time.NanoTime() / J2N.Time.MillisecondsPerNanosecond; // LUCENENET: Use NanoTime() rather than CurrentTimeMilliseconds() for more accurate/reliable results
if (m_termInstances == 0)
{
// we didn't invert anything
// lower memory consumption.
m_tnums = null;
}
else
{
this.m_index = index;
//
// transform intermediate form into the final form, building a single byte[]
// at a time, and releasing the intermediate byte[]s as we go to avoid
// increasing the memory footprint.
//
for (int pass = 0; pass < 256; pass++)
{
var target = m_tnums[pass];
var pos = 0; // end in target;
if (target != null)
{
pos = target.Length;
}
else
{
target = new sbyte[4096];
}
// loop over documents, 0x00ppxxxx, 0x01ppxxxx, 0x02ppxxxx
// where pp is the pass (which array we are building), and xx is all values.
// each pass shares the same byte[] for termNumber lists.
for (int docbase = pass << 16; docbase < maxDoc; docbase += (1 << 24))
{
int lim = Math.Min(docbase + (1 << 16), maxDoc);
for (int doc = docbase; doc < lim; doc++)
{
//System.out.println(" pass="******" process docID=" + doc);
int val = index[doc];
if ((val & 0xff) == 1)
{
int len = val.TripleShift(8);
//System.out.println(" ptr pos=" + pos);
index[doc] = (pos << 8) | 1; // change index to point to start of array
if ((pos & 0xff000000) != 0)
{
// we only have 24 bits for the array index
throw IllegalStateException.Create("Too many values for UnInvertedField faceting on field " + m_field);
}
var arr = bytes[doc];
/*
* for(byte b : arr) {
* //System.out.println(" b=" + Integer.toHexString((int) b));
* }
*/
bytes[doc] = null; // IMPORTANT: allow GC to avoid OOM
if (target.Length <= pos + len)
{
int newlen = target.Length;
//* we don't have to worry about the array getting too large
// since the "pos" param will overflow first (only 24 bits available)
// if ((newlen<<1) <= 0) {
// // overflow...
// newlen = Integer.MAX_VALUE;
// if (newlen <= pos + len) {
// throw new SolrException(400,"Too many terms to uninvert field!");
// }
// } else {
// while (newlen <= pos + len) newlen<<=1; // doubling strategy
// }
//
while (newlen <= pos + len) // doubling strategy
{
newlen <<= 1;
}
var newtarget = new sbyte[newlen];
Array.Copy(target, 0, newtarget, 0, pos);
target = newtarget;
}
Array.Copy(arr, 0, target, pos, len);
pos += len + 1; // skip single byte at end and leave it 0 for terminator
}
}
}
// shrink array
if (pos < target.Length)
{
var newtarget = new sbyte[pos];
Array.Copy(target, 0, newtarget, 0, pos);
target = newtarget;
}
m_tnums[pass] = target;
if ((pass << 16) > maxDoc)
{
break;
}
}
}
if (indexedTerms != null)
{
m_indexedTermsArray = new BytesRef[indexedTerms.Count];
indexedTerms.CopyTo(m_indexedTermsArray, 0);
}
long endTime = J2N.Time.NanoTime() / J2N.Time.MillisecondsPerNanosecond; // LUCENENET: Use NanoTime() rather than CurrentTimeMilliseconds() for more accurate/reliable results
m_total_time = (int)(endTime - startTime);
m_phase1_time = (int)(midPoint - startTime);
}
public virtual void _run()
{
for (int iter = 0; iter < NUM_TEST_ITER; iter++)
{
FieldData field = fields[Random.Next(fields.Length)];
TermsEnum termsEnum = termsDict.GetTerms(field.fieldInfo.Name).GetEnumerator();
#pragma warning disable 612, 618
if (si.Codec is Lucene3xCodec)
#pragma warning restore 612, 618
{
// code below expects unicode sort order
continue;
}
int upto = 0;
// Test straight enum of the terms:
while (termsEnum.MoveNext())
{
BytesRef term = termsEnum.Term;
BytesRef expected = new BytesRef(field.terms[upto++].text2);
Assert.IsTrue(expected.BytesEquals(term), "expected=" + expected + " vs actual " + term);
}
Assert.AreEqual(upto, field.terms.Length);
// Test random seek:
TermData term2 = field.terms[Random.Next(field.terms.Length)];
TermsEnum.SeekStatus status = termsEnum.SeekCeil(new BytesRef(term2.text2));
Assert.AreEqual(status, TermsEnum.SeekStatus.FOUND);
Assert.AreEqual(term2.docs.Length, termsEnum.DocFreq);
if (field.omitTF)
{
this.VerifyDocs(term2.docs, term2.positions, TestUtil.Docs(Random, termsEnum, null, null, DocsFlags.NONE), false);
}
else
{
this.VerifyDocs(term2.docs, term2.positions, termsEnum.DocsAndPositions(null, null), true);
}
// Test random seek by ord:
int idx = Random.Next(field.terms.Length);
term2 = field.terms[idx];
bool success = false;
try
{
termsEnum.SeekExact(idx);
success = true;
}
catch (Exception uoe) when(uoe.IsUnsupportedOperationException())
{
// ok -- skip it
}
if (success)
{
Assert.AreEqual(status, TermsEnum.SeekStatus.FOUND);
Assert.IsTrue(termsEnum.Term.BytesEquals(new BytesRef(term2.text2)));
Assert.AreEqual(term2.docs.Length, termsEnum.DocFreq);
if (field.omitTF)
{
this.VerifyDocs(term2.docs, term2.positions, TestUtil.Docs(Random, termsEnum, null, null, DocsFlags.NONE), false);
}
else
{
this.VerifyDocs(term2.docs, term2.positions, termsEnum.DocsAndPositions(null, null), true);
}
}
// Test seek to non-existent terms:
if (Verbose)
{
Console.WriteLine("TEST: seek non-exist terms");
}
for (int i = 0; i < 100; i++)
{
string text2 = TestUtil.RandomUnicodeString(Random) + ".";
status = termsEnum.SeekCeil(new BytesRef(text2));
Assert.IsTrue(status == TermsEnum.SeekStatus.NOT_FOUND || status == TermsEnum.SeekStatus.END);
}
// Seek to each term, backwards:
if (Verbose)
{
Console.WriteLine("TEST: seek terms backwards");
}
for (int i = field.terms.Length - 1; i >= 0; i--)
{
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, termsEnum.SeekCeil(new BytesRef(field.terms[i].text2)), Thread.CurrentThread.Name + ": field=" + field.fieldInfo.Name + " term=" + field.terms[i].text2);
Assert.AreEqual(field.terms[i].docs.Length, termsEnum.DocFreq);
}
// Seek to each term by ord, backwards
for (int i = field.terms.Length - 1; i >= 0; i--)
{
try
{
termsEnum.SeekExact(i);
Assert.AreEqual(field.terms[i].docs.Length, termsEnum.DocFreq);
Assert.IsTrue(termsEnum.Term.BytesEquals(new BytesRef(field.terms[i].text2)));
}
catch (Exception uoe) when(uoe.IsUnsupportedOperationException())
{
}
}
// Seek to non-existent empty-string term
status = termsEnum.SeekCeil(new BytesRef(""));
Assert.IsNotNull(status);
//Assert.AreEqual(TermsEnum.SeekStatus.NOT_FOUND, status);
// Make sure we're now pointing to first term
Assert.IsTrue(termsEnum.Term.BytesEquals(new BytesRef(field.terms[0].text2)));
// Test docs enum
termsEnum.SeekCeil(new BytesRef(""));
upto = 0;
do
{
term2 = field.terms[upto];
if (Random.Next(3) == 1)
{
DocsEnum docs;
DocsEnum docsAndFreqs;
DocsAndPositionsEnum postings;
if (!field.omitTF)
{
postings = termsEnum.DocsAndPositions(null, null);
if (postings != null)
{
docs = docsAndFreqs = postings;
}
else
{
docs = docsAndFreqs = TestUtil.Docs(Random, termsEnum, null, null, DocsFlags.FREQS);
}
}
else
{
postings = null;
docsAndFreqs = null;
docs = TestUtil.Docs(Random, termsEnum, null, null, DocsFlags.NONE);
}
Assert.IsNotNull(docs);
int upto2 = -1;
bool ended = false;
while (upto2 < term2.docs.Length - 1)
{
// Maybe skip:
int left = term2.docs.Length - upto2;
int doc;
if (Random.Next(3) == 1 && left >= 1)
{
int inc = 1 + Random.Next(left - 1);
upto2 += inc;
if (Random.Next(2) == 1)
{
doc = docs.Advance(term2.docs[upto2]);
Assert.AreEqual(term2.docs[upto2], doc);
}
else
{
doc = docs.Advance(1 + term2.docs[upto2]);
if (doc == DocIdSetIterator.NO_MORE_DOCS)
{
// skipped past last doc
if (Debugging.AssertsEnabled)
{
Debugging.Assert(upto2 == term2.docs.Length - 1);
}
ended = true;
break;
}
else
{
// skipped to next doc
if (Debugging.AssertsEnabled)
{
Debugging.Assert(upto2 < term2.docs.Length - 1);
}
if (doc >= term2.docs[1 + upto2])
{
upto2++;
}
}
}
}
else
{
doc = docs.NextDoc();
Assert.IsTrue(doc != -1);
upto2++;
}
Assert.AreEqual(term2.docs[upto2], doc);
if (!field.omitTF)
{
Assert.AreEqual(term2.positions[upto2].Length, postings.Freq);
if (Random.Next(2) == 1)
{
this.VerifyPositions(term2.positions[upto2], postings);
}
}
}
if (!ended)
{
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, docs.NextDoc());
}
}
upto++;
} while (termsEnum.MoveNext());
Assert.AreEqual(upto, field.terms.Length);
}
}
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).GetTerms("field");
Assert.IsNotNull(vector);
TermsEnum termsEnum = vector.GetEnumerator();
Assert.IsTrue(termsEnum.MoveNext());
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.IsTrue(termsEnum.MoveNext());
Assert.AreEqual(new BytesRef("abcd"), termsEnum.Term);
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.IsFalse(termsEnum.MoveNext());
r.Dispose();
dir.Dispose();
}
public virtual void TestFixedPostings()
{
const int NUM_TERMS = 100;
TermData[] terms = new TermData[NUM_TERMS];
for (int i = 0; i < NUM_TERMS; i++)
{
int[] docs = new int[] { i };
string text = Convert.ToString(i);
terms[i] = new TermData(this, text, docs, null);
}
FieldInfos.Builder builder = new FieldInfos.Builder();
FieldData field = new FieldData(this, "field", builder, terms, true, false);
FieldData[] fields = new FieldData[] { field };
FieldInfos fieldInfos = builder.Finish();
// LUCENENET specific - BUG: we must wrap this in a using block in case anything in the below loop throws
using Directory dir = NewDirectory();
this.Write(fieldInfos, dir, fields, true);
Codec codec = Codec.Default;
SegmentInfo si = new SegmentInfo(dir, Constants.LUCENE_MAIN_VERSION, SEGMENT, 10000, false, codec, null);
// LUCENENET specific - BUG: we must wrap this in a using block in case anything in the below loop throws
using FieldsProducer reader = codec.PostingsFormat.FieldsProducer(new SegmentReadState(dir, si, fieldInfos, NewIOContext(Random), DirectoryReader.DEFAULT_TERMS_INDEX_DIVISOR));
IEnumerator <string> fieldsEnum = reader.GetEnumerator();
fieldsEnum.MoveNext();
string fieldName = fieldsEnum.Current;
Assert.IsNotNull(fieldName);
Terms terms2 = reader.GetTerms(fieldName);
Assert.IsNotNull(terms2);
TermsEnum termsEnum = terms2.GetEnumerator();
DocsEnum docsEnum = null;
for (int i = 0; i < NUM_TERMS; i++)
{
Assert.IsTrue(termsEnum.MoveNext());
BytesRef term = termsEnum.Term;
Assert.AreEqual(terms[i].text2, term.Utf8ToString());
// do this twice to stress test the codec's reuse, ie,
// make sure it properly fully resets (rewinds) its
// internal state:
for (int iter = 0; iter < 2; iter++)
{
docsEnum = TestUtil.Docs(Random, termsEnum, null, docsEnum, DocsFlags.NONE);
Assert.AreEqual(terms[i].docs[0], docsEnum.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, docsEnum.NextDoc());
}
}
Assert.IsFalse(termsEnum.MoveNext());
for (int i = 0; i < NUM_TERMS; i++)
{
Assert.AreEqual(termsEnum.SeekCeil(new BytesRef(terms[i].text2)), TermsEnum.SeekStatus.FOUND);
}
Assert.IsFalse(fieldsEnum.MoveNext());
}
public virtual void Test()
{
IList <string> postingsList = new List <string>();
int numTerms = AtLeast(300);
int maxTermsPerDoc = TestUtil.NextInt32(Random, 10, 20);
bool isSimpleText = "SimpleText".Equals(TestUtil.GetPostingsFormat("field"), StringComparison.Ordinal);
IndexWriterConfig iwc = NewIndexWriterConfig(Random, TEST_VERSION_CURRENT, new MockAnalyzer(Random));
if ((isSimpleText || iwc.MergePolicy is MockRandomMergePolicy) && (TestNightly || RandomMultiplier > 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, CultureInfo.InvariantCulture);
for (int j = 0; j < i; j++)
{
postingsList.Add(term);
}
}
postingsList.Shuffle(Random);
ConcurrentQueue <string> postings = new ConcurrentQueue <string>(postingsList);
Directory dir = NewFSDirectory(CreateTempDir("bagofpostings"));
RandomIndexWriter iw = new RandomIndexWriter(Random, dir, iwc);
int threadCount = TestUtil.NextInt32(Random, 1, 5);
if (Verbose)
{
Console.WriteLine("config: " + iw.IndexWriter.Config);
Console.WriteLine("threadCount=" + threadCount);
}
ThreadJob[] threads = new ThreadJob[threadCount];
CountdownEvent startingGun = new CountdownEvent(1);
for (int threadID = 0; threadID < threadCount; threadID++)
{
threads[threadID] = new ThreadAnonymousInnerClassHelper(this, maxTermsPerDoc, postings, iw, startingGun);
threads[threadID].Start();
}
startingGun.Signal();
foreach (ThreadJob t in threads)
{
t.Join();
}
iw.ForceMerge(1);
DirectoryReader ir = iw.GetReader();
Assert.AreEqual(1, ir.Leaves.Count);
AtomicReader air = (AtomicReader)ir.Leaves[0].Reader;
Terms terms = air.GetTerms("field");
// numTerms-1 because there cannot be a term 0 with 0 postings:
#pragma warning disable 612, 618
Assert.AreEqual(numTerms - 1, air.Fields.UniqueTermCount);
if (iwc.Codec is Lucene3xCodec == false)
#pragma warning restore 612, 618
{
Assert.AreEqual(numTerms - 1, terms.Count);
}
TermsEnum termsEnum = terms.GetEnumerator();
while (termsEnum.MoveNext())
{
int value = Convert.ToInt32(termsEnum.Term.Utf8ToString(), CultureInfo.InvariantCulture);
Assert.AreEqual(value, termsEnum.DocFreq);
// don't really need to check more than this, as CheckIndex
// will verify that docFreq == actual number of documents seen
// from a docsAndPositionsEnum.
}
ir.Dispose();
iw.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(new List <AtomicReader> {
reader1, reader2
}, si, (InfoStream)InfoStream.Default, mergedDir, IndexWriterConfig.DEFAULT_TERM_INDEX_INTERVAL, 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).GetTerms(DocHelper.TEXT_FIELD_2_KEY);
Assert.IsNotNull(vector);
Assert.AreEqual(3, vector.Count);
TermsEnum termsEnum = vector.GetEnumerator();
int i = 0;
while (termsEnum.MoveNext())
{
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, StringComparison.Ordinal) != -1);
Assert.IsTrue(DocHelper.FIELD_2_FREQS[i] == freq);
i++;
}
TestSegmentReader.CheckNorms(mergedReader);
mergedReader.Dispose();
}
private void TestRandomSeeks(IndexReader r, params string[] validTermStrings)
{
BytesRef[] validTerms = new BytesRef[validTermStrings.Length];
for (int termIDX = 0; termIDX < validTermStrings.Length; termIDX++)
{
validTerms[termIDX] = new BytesRef(validTermStrings[termIDX]);
}
Array.Sort(validTerms);
if (Verbose)
{
Console.WriteLine("TEST: " + validTerms.Length + " terms:");
foreach (BytesRef t in validTerms)
{
Console.WriteLine(" " + t.Utf8ToString() + " " + t);
}
}
TermsEnum te = MultiFields.GetTerms(r, FIELD).GetEnumerator();
int END_LOC = -validTerms.Length - 1;
IList <TermAndState> termStates = new JCG.List <TermAndState>();
for (int iter = 0; iter < 100 * RandomMultiplier; iter++)
{
BytesRef t;
int loc;
TermState termState;
if (Random.Next(6) == 4)
{
// pick term that doens't exist:
t = GetNonExistTerm(validTerms);
termState = null;
if (Verbose)
{
Console.WriteLine("\nTEST: invalid term=" + t.Utf8ToString());
}
loc = Array.BinarySearch(validTerms, t);
}
else if (termStates.Count != 0 && Random.Next(4) == 1)
{
TermAndState ts = termStates[Random.Next(termStates.Count)];
t = ts.Term;
loc = Array.BinarySearch(validTerms, t);
Assert.IsTrue(loc >= 0);
termState = ts.State;
if (Verbose)
{
Console.WriteLine("\nTEST: valid termState term=" + t.Utf8ToString());
}
}
else
{
// pick valid term
loc = Random.Next(validTerms.Length);
t = BytesRef.DeepCopyOf(validTerms[loc]);
termState = null;
if (Verbose)
{
Console.WriteLine("\nTEST: valid term=" + t.Utf8ToString());
}
}
// seekCeil or seekExact:
bool doSeekExact = Random.NextBoolean();
if (termState != null)
{
if (Verbose)
{
Console.WriteLine(" seekExact termState");
}
te.SeekExact(t, termState);
}
else if (doSeekExact)
{
if (Verbose)
{
Console.WriteLine(" seekExact");
}
Assert.AreEqual(loc >= 0, te.SeekExact(t));
}
else
{
if (Verbose)
{
Console.WriteLine(" seekCeil");
}
TermsEnum.SeekStatus result = te.SeekCeil(t);
if (Verbose)
{
Console.WriteLine(" got " + result);
}
if (loc >= 0)
{
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, result);
}
else if (loc == END_LOC)
{
Assert.AreEqual(TermsEnum.SeekStatus.END, result);
}
else
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(loc >= -validTerms.Length);
}
Assert.AreEqual(TermsEnum.SeekStatus.NOT_FOUND, result);
}
}
if (loc >= 0)
{
Assert.AreEqual(t, te.Term);
}
else if (doSeekExact)
{
// TermsEnum is unpositioned if seekExact returns false
continue;
}
else if (loc == END_LOC)
{
continue;
}
else
{
loc = -loc - 1;
Assert.AreEqual(validTerms[loc], te.Term);
}
// Do a bunch of next's after the seek
int numNext = Random.Next(validTerms.Length);
for (int nextCount = 0; nextCount < numNext; nextCount++)
{
if (Verbose)
{
Console.WriteLine("\nTEST: next loc=" + loc + " of " + validTerms.Length);
}
bool moved = te.MoveNext();
//BytesRef t2 = te.Term;
loc++;
if (loc == validTerms.Length)
{
//Assert.IsNull(t2); // LUCENENET specific - accessing the Term after MoveNext() returns false results in an assertion failure
Assert.IsFalse(moved);
break;
}
else
{
Assert.AreEqual(validTerms[loc], te.Term);
if (Random.Next(40) == 17 && termStates.Count < 100)
{
termStates.Add(new TermAndState(validTerms[loc], te.GetTermState()));
}
}
}
}
}
public virtual void Test()
{
IList <string> postingsList = new List <string>();
int numTerms = AtLeast(300);
int maxTermsPerDoc = TestUtil.NextInt32(Random, 10, 20);
bool isSimpleText = "SimpleText".Equals(TestUtil.GetPostingsFormat("field"), StringComparison.Ordinal);
IndexWriterConfig iwc = NewIndexWriterConfig(Random, TEST_VERSION_CURRENT, new MockAnalyzer(Random));
if ((isSimpleText || iwc.MergePolicy is MockRandomMergePolicy) && (TestNightly || RandomMultiplier > 1))
{
// Otherwise test can take way too long (> 2 hours)
//numTerms /= 2;
// LUCENENET specific - To keep this under the 1 hour free limit
// of Azure DevOps, this was reduced from /2 to /6.
numTerms /= 6;
}
if (Verbose)
{
Console.WriteLine("maxTermsPerDoc=" + maxTermsPerDoc);
Console.WriteLine("numTerms=" + numTerms);
}
for (int i = 0; i < numTerms; i++)
{
string term = Convert.ToString(i, CultureInfo.InvariantCulture);
for (int j = 0; j < i; j++)
{
postingsList.Add(term);
}
}
postingsList.Shuffle(Random);
ConcurrentQueue <string> postings = new ConcurrentQueue <string>(postingsList);
Directory dir = NewFSDirectory(CreateTempDir(GetFullMethodName()));
RandomIndexWriter iw = new RandomIndexWriter(Random, dir, iwc);
int threadCount = TestUtil.NextInt32(Random, 1, 5);
if (Verbose)
{
Console.WriteLine("config: " + iw.IndexWriter.Config);
Console.WriteLine("threadCount=" + threadCount);
}
Field prototype = NewTextField("field", "", Field.Store.NO);
FieldType fieldType = new FieldType(prototype.FieldType);
if (Random.NextBoolean())
{
fieldType.OmitNorms = true;
}
int options = Random.Next(3);
if (options == 0)
{
fieldType.IndexOptions = 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 = IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
}
// else just positions
ThreadJob[] threads = new ThreadJob[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 (ThreadJob t in threads)
{
t.Join();
}
iw.ForceMerge(1);
DirectoryReader ir = iw.GetReader();
Assert.AreEqual(1, ir.Leaves.Count);
AtomicReader air = (AtomicReader)ir.Leaves[0].Reader;
Terms terms = air.GetTerms("field");
// numTerms-1 because there cannot be a term 0 with 0 postings:
Assert.AreEqual(numTerms - 1, terms.Count);
TermsEnum termsEnum = terms.GetEnumerator();
while (termsEnum.MoveNext())
{
int value = Convert.ToInt32(termsEnum.Term.Utf8ToString(), CultureInfo.InvariantCulture);
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();
}
public virtual void TestIntersectBasic()
{
Directory dir = NewDirectory();
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random));
iwc.SetMergePolicy(new LogDocMergePolicy());
RandomIndexWriter w = new RandomIndexWriter(Random, dir, iwc);
Document doc = new Document();
doc.Add(NewTextField("field", "aaa", Field.Store.NO));
w.AddDocument(doc);
doc = new Document();
doc.Add(NewStringField("field", "bbb", Field.Store.NO));
w.AddDocument(doc);
doc = new Document();
doc.Add(NewTextField("field", "ccc", Field.Store.NO));
w.AddDocument(doc);
w.ForceMerge(1);
DirectoryReader r = w.GetReader();
w.Dispose();
AtomicReader sub = GetOnlySegmentReader(r);
Terms terms = sub.Fields.GetTerms("field");
Automaton automaton = (new RegExp(".*", RegExpSyntax.NONE)).ToAutomaton();
CompiledAutomaton ca = new CompiledAutomaton(automaton, false, false);
TermsEnum te = terms.Intersect(ca, null);
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("aaa", te.Term.Utf8ToString());
Assert.AreEqual(0, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("bbb", te.Term.Utf8ToString());
Assert.AreEqual(1, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("ccc", te.Term.Utf8ToString());
Assert.AreEqual(2, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsFalse(te.MoveNext());
te = terms.Intersect(ca, new BytesRef("abc"));
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("bbb", te.Term.Utf8ToString());
Assert.AreEqual(1, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("ccc", te.Term.Utf8ToString());
Assert.AreEqual(2, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsFalse(te.MoveNext());
te = terms.Intersect(ca, new BytesRef("aaa"));
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("bbb", te.Term.Utf8ToString());
Assert.AreEqual(1, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsTrue(te.MoveNext());
Assert.AreEqual("ccc", te.Term.Utf8ToString());
Assert.AreEqual(2, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsFalse(te.MoveNext());
r.Dispose();
dir.Dispose();
}
private void Verify(AtomicReader r, int[][] idToOrds, BytesRef[] termsArray, BytesRef prefixRef)
{
DocTermOrds dto = new DocTermOrds(r, r.LiveDocs, "field", prefixRef, int.MaxValue, TestUtil.NextInt32(Random, 2, 10));
FieldCache.Int32s docIDToID = FieldCache.DEFAULT.GetInt32s(r, "id", false);
/*
* for(int docID=0;docID<subR.MaxDoc;docID++) {
* System.out.println(" docID=" + docID + " id=" + docIDToID[docID]);
* }
*/
if (Verbose)
{
Console.WriteLine("TEST: verify prefix=" + (prefixRef == null ? "null" : prefixRef.Utf8ToString()));
Console.WriteLine("TEST: all TERMS:");
TermsEnum allTE = MultiFields.GetTerms(r, "field").GetEnumerator();
int ord = 0;
while (allTE.MoveNext())
{
Console.WriteLine(" ord=" + (ord++) + " term=" + allTE.Term.Utf8ToString());
}
}
//final TermsEnum te = subR.Fields.Terms("field").iterator();
TermsEnum te = dto.GetOrdTermsEnum(r);
if (dto.NumTerms == 0)
{
if (prefixRef == null)
{
Assert.IsNull(MultiFields.GetTerms(r, "field"));
}
else
{
Terms terms = MultiFields.GetTerms(r, "field");
if (terms != null)
{
TermsEnum termsEnum = terms.GetEnumerator();
TermsEnum.SeekStatus result = termsEnum.SeekCeil(prefixRef);
if (result != TermsEnum.SeekStatus.END)
{
Assert.IsFalse(StringHelper.StartsWith(termsEnum.Term, prefixRef), "term=" + termsEnum.Term.Utf8ToString() + " matches prefix=" + prefixRef.Utf8ToString());
}
else
{
// ok
}
}
else
{
// ok
}
}
return;
}
if (Verbose)
{
Console.WriteLine("TEST: TERMS:");
te.SeekExact(0);
while (true)
{
Console.WriteLine(" ord=" + te.Ord + " term=" + te.Term.Utf8ToString());
if (!te.MoveNext())
{
break;
}
}
}
SortedSetDocValues iter = dto.GetIterator(r);
for (int docID = 0; docID < r.MaxDoc; docID++)
{
if (Verbose)
{
Console.WriteLine("TEST: docID=" + docID + " of " + r.MaxDoc + " (id=" + docIDToID.Get(docID) + ")");
}
iter.SetDocument(docID);
int[] answers = idToOrds[docIDToID.Get(docID)];
int upto = 0;
long ord;
while ((ord = iter.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
te.SeekExact(ord);
BytesRef expected = termsArray[answers[upto++]];
if (Verbose)
{
Console.WriteLine(" exp=" + expected.Utf8ToString() + " actual=" + te.Term.Utf8ToString());
}
Assert.AreEqual(expected, te.Term, "expected=" + expected.Utf8ToString() + " actual=" + te.Term.Utf8ToString() + " ord=" + ord);
}
Assert.AreEqual(answers.Length, upto);
}
}
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.NextInt32(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 IterableAnonymousClass(this, fieldCount, finalDocCount, finalBaseCount));
}
IndexReader r = w.GetReader();
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)
{
IIndexableField 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.GetBinaryValue();
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
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(stringValue != null);
}
Assert.AreEqual(stringValue, f.GetStringValue());
}
}
if (indexed)
{
bool tv = counter % 2 == 1 && fieldID != 9;
if (tv)
{
Terms tfv = r.GetTermVectors(docID).GetTerms(name);
Assert.IsNotNull(tfv);
TermsEnum termsEnum = tfv.GetEnumerator();
Assert.IsTrue(termsEnum.MoveNext());
Assert.AreEqual(new BytesRef("" + counter), termsEnum.Term);
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.IsTrue(termsEnum.MoveNext());
Assert.AreEqual(new BytesRef("text"), termsEnum.Term);
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.IsFalse(termsEnum.MoveNext());
// TODO: offsets
}
else
{
Fields vectors = r.GetTermVectors(docID);
Assert.IsTrue(vectors is null || vectors.GetTerms(name) is null);
}
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(new Term("id", "" + id)), Occur.MUST);
bq.Add(new TermQuery(new Term(name, "text")), 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)), Occur.MUST);
bq.Add(new TermQuery(new Term(name, "" + counter)), 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 TestSortedTermsEnum()
{
Directory directory = NewDirectory();
Analyzer analyzer = new MockAnalyzer(Random);
IndexWriterConfig iwconfig = NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer);
iwconfig.SetMergePolicy(NewLogMergePolicy());
RandomIndexWriter iwriter = new RandomIndexWriter(Random, directory, iwconfig);
Document doc = new Document();
doc.Add(new StringField("field", "hello", Field.Store.NO));
iwriter.AddDocument(doc);
doc = new Document();
doc.Add(new StringField("field", "world", Field.Store.NO));
iwriter.AddDocument(doc);
doc = new Document();
doc.Add(new StringField("field", "beer", Field.Store.NO));
iwriter.AddDocument(doc);
iwriter.ForceMerge(1);
DirectoryReader ireader = iwriter.GetReader();
iwriter.Dispose();
AtomicReader ar = GetOnlySegmentReader(ireader);
SortedSetDocValues dv = FieldCache.DEFAULT.GetDocTermOrds(ar, "field");
Assert.AreEqual(3, dv.ValueCount);
TermsEnum termsEnum = dv.GetTermsEnum();
// next()
Assert.IsTrue(termsEnum.MoveNext());
Assert.AreEqual("beer", termsEnum.Term.Utf8ToString());
Assert.AreEqual(0, termsEnum.Ord);
Assert.IsTrue(termsEnum.MoveNext());
Assert.AreEqual("hello", termsEnum.Term.Utf8ToString());
Assert.AreEqual(1, termsEnum.Ord);
Assert.IsTrue(termsEnum.MoveNext());
Assert.AreEqual("world", termsEnum.Term.Utf8ToString());
Assert.AreEqual(2, termsEnum.Ord);
// seekCeil()
Assert.AreEqual(SeekStatus.NOT_FOUND, termsEnum.SeekCeil(new BytesRef("ha!")));
Assert.AreEqual("hello", termsEnum.Term.Utf8ToString());
Assert.AreEqual(1, termsEnum.Ord);
Assert.AreEqual(SeekStatus.FOUND, termsEnum.SeekCeil(new BytesRef("beer")));
Assert.AreEqual("beer", termsEnum.Term.Utf8ToString());
Assert.AreEqual(0, termsEnum.Ord);
Assert.AreEqual(SeekStatus.END, termsEnum.SeekCeil(new BytesRef("zzz")));
// seekExact()
Assert.IsTrue(termsEnum.SeekExact(new BytesRef("beer")));
Assert.AreEqual("beer", termsEnum.Term.Utf8ToString());
Assert.AreEqual(0, termsEnum.Ord);
Assert.IsTrue(termsEnum.SeekExact(new BytesRef("hello")));
Assert.AreEqual("hello", termsEnum.Term.Utf8ToString());
Assert.AreEqual(1, termsEnum.Ord);
Assert.IsTrue(termsEnum.SeekExact(new BytesRef("world")));
Assert.AreEqual("world", termsEnum.Term.Utf8ToString());
Assert.AreEqual(2, termsEnum.Ord);
Assert.IsFalse(termsEnum.SeekExact(new BytesRef("bogus")));
// seek(ord)
termsEnum.SeekExact(0);
Assert.AreEqual("beer", termsEnum.Term.Utf8ToString());
Assert.AreEqual(0, termsEnum.Ord);
termsEnum.SeekExact(1);
Assert.AreEqual("hello", termsEnum.Term.Utf8ToString());
Assert.AreEqual(1, termsEnum.Ord);
termsEnum.SeekExact(2);
Assert.AreEqual("world", termsEnum.Term.Utf8ToString());
Assert.AreEqual(2, termsEnum.Ord);
ireader.Dispose();
directory.Dispose();
}
public static void VerifyEquals(Fields d1, Fields d2)
{
if (d1 == null)
{
Assert.IsTrue(d2 == null || d2.Count == 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.GetTerms(field1);
Assert.IsNotNull(terms1);
TermsEnum termsEnum1 = terms1.GetEnumerator();
Terms terms2 = d2.GetTerms(field2);
Assert.IsNotNull(terms2);
TermsEnum termsEnum2 = terms2.GetEnumerator();
DocsAndPositionsEnum dpEnum1 = null;
DocsAndPositionsEnum dpEnum2 = null;
DocsEnum dEnum1 = null;
DocsEnum dEnum2 = null;
BytesRef term1;
while (termsEnum1.MoveNext())
{
term1 = termsEnum1.Term;
termsEnum2.MoveNext();
BytesRef term2 = termsEnum2.Term;
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, DocsFlags.FREQS);
dEnum2 = TestUtil.Docs(Random, termsEnum2, null, dEnum2, DocsFlags.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.IsFalse(termsEnum2.MoveNext());
}
Assert.IsFalse(fieldsEnum2.MoveNext());
}
protected virtual void AssertEquals(RandomTokenStream tk, FieldType ft, Terms terms)
{
Assert.AreEqual(1, terms.DocCount);
int termCount = new JCG.HashSet <string>(tk.terms).Count;
Assert.AreEqual((long)termCount, terms.Count); // LUCENENET specific - cast required because types don't match (xUnit checks this)
Assert.AreEqual((long)termCount, terms.SumDocFreq); // LUCENENET specific - cast required because types don't match (xUnit checks this)
Assert.AreEqual(ft.StoreTermVectorPositions, terms.HasPositions);
Assert.AreEqual(ft.StoreTermVectorOffsets, terms.HasOffsets);
Assert.AreEqual(ft.StoreTermVectorPayloads && tk.HasPayloads(), terms.HasPayloads);
ISet <BytesRef> uniqueTerms = new JCG.HashSet <BytesRef>();
foreach (string term in tk.freqs.Keys)
{
uniqueTerms.Add(new BytesRef(term));
}
BytesRef[] sortedTerms = uniqueTerms.ToArray(/*new BytesRef[0]*/);
Array.Sort(sortedTerms, terms.Comparer);
TermsEnum termsEnum = terms.GetEnumerator(Random.NextBoolean() ? null : this.termsEnum.Value);
this.termsEnum.Value = termsEnum;
for (int i = 0; i < sortedTerms.Length; ++i)
{
Assert.IsTrue(termsEnum.MoveNext());
Assert.AreEqual(sortedTerms[i], termsEnum.Term);
Assert.AreEqual(1, termsEnum.DocFreq);
FixedBitSet bits = new FixedBitSet(1);
DocsEnum docsEnum = termsEnum.Docs(bits, Random.NextBoolean() ? null : this.docsEnum.Value);
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsEnum.NextDoc());
bits.Set(0);
docsEnum = termsEnum.Docs(Random.NextBoolean() ? bits : null, Random.NextBoolean() ? null : docsEnum);
Assert.IsNotNull(docsEnum);
Assert.AreEqual(0, docsEnum.NextDoc());
Assert.AreEqual(0, docsEnum.DocID);
Assert.AreEqual(tk.freqs[termsEnum.Term.Utf8ToString()], docsEnum.Freq);
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsEnum.NextDoc());
this.docsEnum.Value = docsEnum;
bits.Clear(0);
DocsAndPositionsEnum docsAndPositionsEnum = termsEnum.DocsAndPositions(bits, Random.NextBoolean() ? null : this.docsAndPositionsEnum.Value);
Assert.AreEqual(ft.StoreTermVectorOffsets || ft.StoreTermVectorPositions, docsAndPositionsEnum != null);
if (docsAndPositionsEnum != null)
{
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsAndPositionsEnum.NextDoc());
}
bits.Set(0);
docsAndPositionsEnum = termsEnum.DocsAndPositions(Random.NextBoolean() ? bits : null, Random.NextBoolean() ? null : docsAndPositionsEnum);
Assert.AreEqual(ft.StoreTermVectorOffsets || ft.StoreTermVectorPositions, docsAndPositionsEnum != null);
if (terms.HasPositions || terms.HasOffsets)
{
Assert.AreEqual(0, docsAndPositionsEnum.NextDoc());
int freq = docsAndPositionsEnum.Freq;
Assert.AreEqual(tk.freqs[termsEnum.Term.Utf8ToString()], freq);
if (docsAndPositionsEnum != null)
{
for (int k = 0; k < freq; ++k)
{
int position = docsAndPositionsEnum.NextPosition();
ISet <int> indexes;
if (terms.HasPositions)
{
indexes = tk.positionToTerms[position];
Assert.IsNotNull(indexes);
}
else
{
indexes = tk.startOffsetToTerms[docsAndPositionsEnum.StartOffset];
Assert.IsNotNull(indexes);
}
if (terms.HasPositions)
{
bool foundPosition = false;
foreach (int index in indexes)
{
if (tk.termBytes[index].Equals(termsEnum.Term) && tk.positions[index] == position)
{
foundPosition = true;
break;
}
}
Assert.IsTrue(foundPosition);
}
if (terms.HasOffsets)
{
bool foundOffset = false;
foreach (int index in indexes)
{
if (tk.termBytes[index].Equals(termsEnum.Term) && tk.startOffsets[index] == docsAndPositionsEnum.StartOffset && tk.endOffsets[index] == docsAndPositionsEnum.EndOffset)
{
foundOffset = true;
break;
}
}
Assert.IsTrue(foundOffset);
}
if (terms.HasPayloads)
{
bool foundPayload = false;
foreach (int index in indexes)
{
if (tk.termBytes[index].Equals(termsEnum.Term) && Equals(tk.payloads[index], docsAndPositionsEnum.GetPayload()))
{
foundPayload = true;
break;
}
}
Assert.IsTrue(foundPayload);
}
}
try
{
docsAndPositionsEnum.NextPosition();
Assert.Fail();
}
catch (Exception e) when(e.IsException())
{
// ok
}
catch (Exception e) when(e.IsAssertionError())
{
// ok
}
}
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsAndPositionsEnum.NextDoc());
}
this.docsAndPositionsEnum.Value = docsAndPositionsEnum;
}
Assert.IsFalse(termsEnum.MoveNext());
for (int i = 0; i < 5; ++i)
{
if (Random.NextBoolean())
{
Assert.IsTrue(termsEnum.SeekExact(RandomPicks.RandomFrom(Random, tk.termBytes)));
}
else
{
Assert.AreEqual(SeekStatus.FOUND, termsEnum.SeekCeil(RandomPicks.RandomFrom(Random, tk.termBytes)));
}
}
}