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();
}
public virtual void TestRandomStoredFields()
{
using Directory dir = NewDirectory();
Random rand = Random;
using RandomIndexWriter w = new RandomIndexWriter(rand, dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)).SetMaxBufferedDocs(TestUtil.NextInt32(rand, 5, 20)));
//w.w.setNoCFSRatio(0.0);
int docCount = AtLeast(200);
int fieldCount = TestUtil.NextInt32(rand, 1, 5);
IList <int> fieldIDs = new JCG.List <int>();
FieldType customType = new FieldType(TextField.TYPE_STORED);
customType.IsTokenized = false;
Field idField = NewField("id", "", customType);
for (int i = 0; i < fieldCount; i++)
{
fieldIDs.Add(i);
}
IDictionary <string, Document> docs = new Dictionary <string, Document>();
if (Verbose)
{
Console.WriteLine("TEST: build index docCount=" + docCount);
}
FieldType customType2 = new FieldType();
customType2.IsStored = true;
for (int i = 0; i < docCount; i++)
{
Document doc = new Document();
doc.Add(idField);
string id = "" + i;
idField.SetStringValue(id);
docs[id] = doc;
if (Verbose)
{
Console.WriteLine("TEST: add doc id=" + id);
}
foreach (int field in fieldIDs)
{
string s;
if (rand.Next(4) != 3)
{
s = TestUtil.RandomUnicodeString(rand, 1000);
doc.Add(NewField("f" + field, s, customType2));
}
else
{
s = null;
}
}
w.AddDocument(doc);
if (rand.Next(50) == 17)
{
// mixup binding of field name -> Number every so often
fieldIDs.Shuffle(Random);
}
if (rand.Next(5) == 3 && i > 0)
{
string delID = "" + rand.Next(i);
if (Verbose)
{
Console.WriteLine("TEST: delete doc id=" + delID);
}
w.DeleteDocuments(new Term("id", delID));
docs.Remove(delID);
}
}
if (Verbose)
{
Console.WriteLine("TEST: " + docs.Count + " docs in index; now load fields");
}
if (docs.Count > 0)
{
string[] idsList = docs.Keys.ToArray(/*new string[docs.Count]*/);
for (int x = 0; x < 2; x++)
{
using (IndexReader r = w.GetReader())
{
IndexSearcher s = NewSearcher(r);
if (Verbose)
{
Console.WriteLine("TEST: cycle x=" + x + " r=" + r);
}
int num = AtLeast(1000);
for (int iter = 0; iter < num; iter++)
{
string testID = idsList[rand.Next(idsList.Length)];
if (Verbose)
{
Console.WriteLine("TEST: test id=" + testID);
}
TopDocs hits = s.Search(new TermQuery(new Term("id", testID)), 1);
Assert.AreEqual(1, hits.TotalHits);
Document doc = r.Document(hits.ScoreDocs[0].Doc);
Document docExp = docs[testID];
for (int i = 0; i < fieldCount; i++)
{
assertEquals("doc " + testID + ", field f" + fieldCount + " is wrong", docExp.Get("f" + i), doc.Get("f" + i));
}
}
} // r.Dispose();
w.ForceMerge(1);
}
}
}
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) && (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, 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.GetIterator(null);
BytesRef termBR;
while ((termBR = termsEnum.Next()) != null)
{
int value = Convert.ToInt32(termBR.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 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.AreEqual("beer", termsEnum.Next().Utf8ToString());
Assert.AreEqual(0, termsEnum.Ord);
Assert.AreEqual("hello", termsEnum.Next().Utf8ToString());
Assert.AreEqual(1, termsEnum.Ord);
Assert.AreEqual("world", termsEnum.Next().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 virtual void Test2()
{
Random random = Random;
int NUM_DOCS = AtLeast(100);
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
random, dir);
bool allowDups = random.NextBoolean();
ISet <string> seen = new JCG.HashSet <string>();
if (Verbose)
{
Console.WriteLine("TEST: NUM_DOCS=" + NUM_DOCS + " allowDups=" + allowDups);
}
int numDocs = 0;
IList <BytesRef> docValues = new List <BytesRef>();
// TODO: deletions
while (numDocs < NUM_DOCS)
{
string s;
if (random.NextBoolean())
{
s = TestUtil.RandomSimpleString(random);
}
else
{
s = TestUtil.RandomUnicodeString(random);
}
BytesRef br = new BytesRef(s);
if (!allowDups)
{
if (seen.Contains(s))
{
continue;
}
seen.Add(s);
}
if (Verbose)
{
Console.WriteLine(" " + numDocs + ": s=" + s);
}
Document doc = new Document();
doc.Add(new SortedDocValuesField("stringdv", br));
doc.Add(new NumericDocValuesField("id", numDocs));
docValues.Add(br);
writer.AddDocument(doc);
numDocs++;
if (random.Next(40) == 17)
{
// force flush
writer.GetReader().Dispose();
}
}
writer.ForceMerge(1);
DirectoryReader r = writer.GetReader();
writer.Dispose();
AtomicReader sr = GetOnlySegmentReader(r);
long END_TIME = (J2N.Time.NanoTime() / J2N.Time.MillisecondsPerNanosecond) + (TestNightly ? 30 : 1); // LUCENENET: Use NanoTime() rather than CurrentTimeMilliseconds() for more accurate/reliable results
int NUM_THREADS = TestUtil.NextInt32(LuceneTestCase.Random, 1, 10);
ThreadJob[] threads = new ThreadJob[NUM_THREADS];
for (int thread = 0; thread < NUM_THREADS; thread++)
{
threads[thread] = new ThreadAnonymousClass2(random, docValues, sr, END_TIME);
threads[thread].Start();
}
foreach (ThreadJob thread in threads)
{
thread.Join();
}
r.Dispose();
dir.Dispose();
}
public virtual void Test()
{
Random random = new Random(Random.Next());
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(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
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();
List <BytesRef> terms = new List <BytesRef>();
TermsEnum termsEnum = MultiFields.GetTerms(r, "body").GetIterator(null);
BytesRef term;
while ((term = termsEnum.Next()) != null)
{
terms.Add(BytesRef.DeepCopyOf(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.Next() == null;
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 DoTestNumbers(bool withPayloads)
{
Directory dir = NewDirectory();
Analyzer analyzer = withPayloads ? (Analyzer) new MockPayloadAnalyzer() : new MockAnalyzer(Random);
iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer);
iwc.SetMergePolicy(NewLogMergePolicy()); // will rely on docids a bit for skipping
RandomIndexWriter w = new RandomIndexWriter(Random, dir, iwc);
FieldType ft = new FieldType(TextField.TYPE_STORED);
ft.IndexOptions = IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
if (Random.NextBoolean())
{
ft.StoreTermVectors = true;
ft.StoreTermVectorOffsets = Random.NextBoolean();
ft.StoreTermVectorPositions = Random.NextBoolean();
}
int numDocs = AtLeast(500);
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
doc.Add(new Field("numbers", English.Int32ToEnglish(i), ft));
doc.Add(new Field("oddeven", (i % 2) == 0 ? "even" : "odd", ft));
doc.Add(new StringField("id", "" + i, Field.Store.NO));
w.AddDocument(doc);
}
IndexReader reader = w.GetReader();
w.Dispose();
string[] terms = new string[] { "one", "two", "three", "four", "five", "six", "seven", "eight", "nine", "ten", "hundred" };
foreach (string term in terms)
{
DocsAndPositionsEnum dp = MultiFields.GetTermPositionsEnum(reader, null, "numbers", new BytesRef(term));
int doc;
while ((doc = dp.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
string storedNumbers = reader.Document(doc).Get("numbers");
int freq = dp.Freq;
for (int i = 0; i < freq; i++)
{
dp.NextPosition();
int start = dp.StartOffset;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(start >= 0);
}
int end = dp.EndOffset;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(end >= 0 && end >= start);
}
// check that the offsets correspond to the term in the src text
Assert.IsTrue(storedNumbers.Substring(start, end - start).Equals(term, StringComparison.Ordinal));
if (withPayloads)
{
// check that we have a payload and it starts with "pos"
Assert.IsNotNull(dp.GetPayload());
BytesRef payload = dp.GetPayload();
Assert.IsTrue(payload.Utf8ToString().StartsWith("pos:", StringComparison.Ordinal));
} // note: withPayloads=false doesnt necessarily mean we dont have them from MockAnalyzer!
}
}
}
// check we can skip correctly
int numSkippingTests = AtLeast(50);
for (int j = 0; j < numSkippingTests; j++)
{
int num = TestUtil.NextInt32(Random, 100, Math.Min(numDocs - 1, 999));
DocsAndPositionsEnum dp = MultiFields.GetTermPositionsEnum(reader, null, "numbers", new BytesRef("hundred"));
int doc = dp.Advance(num);
Assert.AreEqual(num, doc);
int freq = dp.Freq;
for (int i = 0; i < freq; i++)
{
string storedNumbers = reader.Document(doc).Get("numbers");
dp.NextPosition();
int start = dp.StartOffset;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(start >= 0);
}
int end = dp.EndOffset;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(end >= 0 && end >= start);
}
// check that the offsets correspond to the term in the src text
Assert.IsTrue(storedNumbers.Substring(start, end - start).Equals("hundred", StringComparison.Ordinal));
if (withPayloads)
{
// check that we have a payload and it starts with "pos"
Assert.IsNotNull(dp.GetPayload());
BytesRef payload = dp.GetPayload();
Assert.IsTrue(payload.Utf8ToString().StartsWith("pos:", StringComparison.Ordinal));
} // note: withPayloads=false doesnt necessarily mean we dont have them from MockAnalyzer!
}
}
// check that other fields (without offsets) work correctly
for (int i = 0; i < numDocs; i++)
{
DocsEnum dp = MultiFields.GetTermDocsEnum(reader, null, "id", new BytesRef("" + i), 0);
Assert.AreEqual(i, dp.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dp.NextDoc());
}
reader.Dispose();
dir.Dispose();
}
public virtual void TestRandomDocs()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random, dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)).SetMergePolicy(NewLogMergePolicy()));
int numDocs = AtLeast(49);
int max = 15678;
int term = Random.Next(max);
int[] freqInDoc = new int[numDocs];
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.OmitNorms = true;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
StringBuilder builder = new StringBuilder();
for (int j = 0; j < 199; j++)
{
int nextInt = Random.Next(max);
builder.Append(nextInt).Append(' ');
if (nextInt == term)
{
freqInDoc[i]++;
}
}
doc.Add(NewField(fieldName, builder.ToString(), customType));
writer.AddDocument(doc);
}
IndexReader reader = writer.GetReader();
writer.Dispose();
int num = AtLeast(13);
for (int i = 0; i < num; i++)
{
BytesRef bytes = new BytesRef("" + term);
IndexReaderContext topReaderContext = reader.Context;
foreach (AtomicReaderContext context in topReaderContext.Leaves)
{
int maxDoc = context.AtomicReader.MaxDoc;
DocsEnum docsEnum = TestUtil.Docs(Random, context.Reader, fieldName, bytes, null, null, DocsFlags.FREQS);
if (FindNext(freqInDoc, context.DocBase, context.DocBase + maxDoc) == int.MaxValue)
{
Assert.IsNull(docsEnum);
continue;
}
Assert.IsNotNull(docsEnum);
docsEnum.NextDoc();
for (int j = 0; j < maxDoc; j++)
{
if (freqInDoc[context.DocBase + j] != 0)
{
Assert.AreEqual(j, docsEnum.DocID);
Assert.AreEqual(docsEnum.Freq, freqInDoc[context.DocBase + j]);
if (i % 2 == 0 && Random.Next(10) == 0)
{
int next = FindNext(freqInDoc, context.DocBase + j + 1, context.DocBase + maxDoc) - context.DocBase;
int advancedTo = docsEnum.Advance(next);
if (next >= maxDoc)
{
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, advancedTo);
}
else
{
Assert.IsTrue(next >= advancedTo, "advanced to: " + advancedTo + " but should be <= " + next);
}
}
else
{
docsEnum.NextDoc();
}
}
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, docsEnum.DocID, "DocBase: " + context.DocBase + " maxDoc: " + maxDoc + " " + docsEnum.GetType());
}
}
reader.Dispose();
dir.Dispose();
}
public virtual void DoTestLongPostingsNoPositions(IndexOptions options)
{
// Don't use TestUtil.getTempDir so that we own the
// randomness (ie same seed will point to same dir):
Directory dir = NewFSDirectory(CreateTempDir("longpostings" + "." + Random.NextInt64()));
int NUM_DOCS = AtLeast(2000);
if (Verbose)
{
Console.WriteLine("TEST: NUM_DOCS=" + NUM_DOCS);
}
string s1 = GetRandomTerm(null);
string s2 = GetRandomTerm(s1);
if (Verbose)
{
Console.WriteLine("\nTEST: s1=" + s1 + " s2=" + s2);
/*
* for(int idx=0;idx<s1.length();idx++) {
* System.out.println(" s1 ch=0x" + Integer.toHexString(s1.charAt(idx)));
* }
* for(int idx=0;idx<s2.length();idx++) {
* System.out.println(" s2 ch=0x" + Integer.toHexString(s2.charAt(idx)));
* }
*/
}
FixedBitSet isS1 = new FixedBitSet(NUM_DOCS);
for (int idx = 0; idx < NUM_DOCS; idx++)
{
if (Random.NextBoolean())
{
isS1.Set(idx);
}
}
IndexReader r;
if (true)
{
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)).SetOpenMode(OpenMode.CREATE).SetMergePolicy(NewLogMergePolicy());
iwc.SetRAMBufferSizeMB(16.0 + 16.0 * Random.NextDouble());
iwc.SetMaxBufferedDocs(-1);
RandomIndexWriter riw = new RandomIndexWriter(Random, dir, iwc);
FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
ft.IndexOptions = options;
for (int idx = 0; idx < NUM_DOCS; idx++)
{
Document doc = new Document();
string s = isS1.Get(idx) ? s1 : s2;
Field f = NewField("field", s, ft);
int count = TestUtil.NextInt32(Random, 1, 4);
for (int ct = 0; ct < count; ct++)
{
doc.Add(f);
}
riw.AddDocument(doc);
}
r = riw.GetReader();
riw.Dispose();
}
else
{
#pragma warning disable 162
r = DirectoryReader.Open(dir);
#pragma warning restore 162
}
/*
* if (VERBOSE) {
* System.out.println("TEST: terms");
* TermEnum termEnum = r.Terms();
* while(termEnum.Next()) {
* System.out.println(" term=" + termEnum.Term() + " len=" + termEnum.Term().Text().length());
* Assert.IsTrue(termEnum.DocFreq() > 0);
* System.out.println(" s1?=" + (termEnum.Term().Text().equals(s1)) + " s1len=" + s1.length());
* System.out.println(" s2?=" + (termEnum.Term().Text().equals(s2)) + " s2len=" + s2.length());
* final String s = termEnum.Term().Text();
* for(int idx=0;idx<s.length();idx++) {
* System.out.println(" ch=0x" + Integer.toHexString(s.charAt(idx)));
* }
* }
* }
*/
Assert.AreEqual(NUM_DOCS, r.NumDocs);
Assert.IsTrue(r.DocFreq(new Term("field", s1)) > 0);
Assert.IsTrue(r.DocFreq(new Term("field", s2)) > 0);
int num = AtLeast(1000);
for (int iter = 0; iter < num; iter++)
{
string term;
bool doS1;
if (Random.NextBoolean())
{
term = s1;
doS1 = true;
}
else
{
term = s2;
doS1 = false;
}
if (Verbose)
{
Console.WriteLine("\nTEST: iter=" + iter + " doS1=" + doS1 + " term=" + term);
}
DocsEnum docs;
DocsEnum postings;
if (options == IndexOptions.DOCS_ONLY)
{
docs = TestUtil.Docs(Random, r, "field", new BytesRef(term), null, null, DocsFlags.NONE);
postings = null;
}
else
{
docs = postings = TestUtil.Docs(Random, r, "field", new BytesRef(term), null, null, DocsFlags.FREQS);
if (Debugging.AssertsEnabled)
{
Debugging.Assert(postings != null);
}
}
if (Debugging.AssertsEnabled)
{
Debugging.Assert(docs != null);
}
int docID = -1;
while (docID < DocIdSetIterator.NO_MORE_DOCS)
{
int what = Random.Next(3);
if (what == 0)
{
if (Verbose)
{
Console.WriteLine("TEST: docID=" + docID + "; do next()");
}
// nextDoc
int expected = docID + 1;
while (true)
{
if (expected == NUM_DOCS)
{
expected = int.MaxValue;
break;
}
else if (isS1.Get(expected) == doS1)
{
break;
}
else
{
expected++;
}
}
docID = docs.NextDoc();
if (Verbose)
{
Console.WriteLine(" got docID=" + docID);
}
Assert.AreEqual(expected, docID);
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
if (Random.Next(6) == 3 && postings != null)
{
int freq = postings.Freq;
Assert.IsTrue(freq >= 1 && freq <= 4);
}
}
else
{
// advance
int targetDocID;
if (docID == -1)
{
targetDocID = Random.Next(NUM_DOCS + 1);
}
else
{
targetDocID = docID + TestUtil.NextInt32(Random, 1, NUM_DOCS - docID);
}
if (Verbose)
{
Console.WriteLine("TEST: docID=" + docID + "; do advance(" + targetDocID + ")");
}
int expected = targetDocID;
while (true)
{
if (expected == NUM_DOCS)
{
expected = int.MaxValue;
break;
}
else if (isS1.Get(expected) == doS1)
{
break;
}
else
{
expected++;
}
}
docID = docs.Advance(targetDocID);
if (Verbose)
{
Console.WriteLine(" got docID=" + docID);
}
Assert.AreEqual(expected, docID);
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
if (Random.Next(6) == 3 && postings != null)
{
int freq = postings.Freq;
Assert.IsTrue(freq >= 1 && freq <= 4, "got invalid freq=" + freq);
}
}
}
}
r.Dispose();
dir.Dispose();
}
public virtual void TestStressAdvance_Mem()
{
for (int iter = 0; iter < 3; iter++)
{
if (Verbose)
{
Console.WriteLine("\nTEST: iter=" + iter);
}
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
ISet <int> aDocs = new JCG.HashSet <int>();
Documents.Document doc = new Documents.Document();
Field f = NewStringField("field", "", Field.Store.NO);
doc.Add(f);
Field idField = NewStringField("id", "", Field.Store.YES);
doc.Add(idField);
int num = AtLeast(4097);
if (Verbose)
{
Console.WriteLine("\nTEST: numDocs=" + num);
}
for (int id = 0; id < num; id++)
{
if (Random.Next(4) == 3)
{
f.SetStringValue("a");
aDocs.Add(id);
}
else
{
f.SetStringValue("b");
}
idField.SetStringValue("" + id);
w.AddDocument(doc);
if (Verbose)
{
Console.WriteLine("\nTEST: doc upto " + id);
}
}
w.ForceMerge(1);
IList <int> aDocIDs = new JCG.List <int>();
IList <int> bDocIDs = new JCG.List <int>();
DirectoryReader r = w.GetReader();
int[] idToDocID = new int[r.MaxDoc];
for (int docID = 0; docID < idToDocID.Length; docID++)
{
int id = Convert.ToInt32(r.Document(docID).Get("id"));
if (aDocs.Contains(id))
{
aDocIDs.Add(docID);
}
else
{
bDocIDs.Add(docID);
}
}
TermsEnum te = GetOnlySegmentReader(r).Fields.GetTerms("field").GetEnumerator();
DocsEnum de = null;
for (int iter2 = 0; iter2 < 10; iter2++)
{
if (Verbose)
{
Console.WriteLine("\nTEST: iter=" + iter + " iter2=" + iter2);
}
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, te.SeekCeil(new BytesRef("a")));
de = TestUtil.Docs(Random, te, null, de, DocsFlags.NONE);
TestOne(de, aDocIDs);
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, te.SeekCeil(new BytesRef("b")));
de = TestUtil.Docs(Random, te, null, de, DocsFlags.NONE);
TestOne(de, bDocIDs);
}
w.Dispose();
r.Dispose();
dir.Dispose();
}
}
public virtual void TestLongPostings_Mem()
{
// Don't use TestUtil.getTempDir so that we own the
// randomness (ie same seed will point to same dir):
Directory dir = NewFSDirectory(CreateTempDir("longpostings" + "." + Random.NextInt64()));
int NUM_DOCS = AtLeast(2000);
if (Verbose)
{
Console.WriteLine("TEST: NUM_DOCS=" + NUM_DOCS);
}
string s1 = GetRandomTerm(null);
string s2 = GetRandomTerm(s1);
if (Verbose)
{
Console.WriteLine("\nTEST: s1=" + s1 + " s2=" + s2);
/*
* for(int idx=0;idx<s1.length();idx++) {
* System.out.println(" s1 ch=0x" + Integer.toHexString(s1.charAt(idx)));
* }
* for(int idx=0;idx<s2.length();idx++) {
* System.out.println(" s2 ch=0x" + Integer.toHexString(s2.charAt(idx)));
* }
*/
}
FixedBitSet isS1 = new FixedBitSet(NUM_DOCS);
for (int idx = 0; idx < NUM_DOCS; idx++)
{
if (Random.NextBoolean())
{
isS1.Set(idx);
}
}
IndexReader r;
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)).SetOpenMode(OpenMode.CREATE).SetMergePolicy(NewLogMergePolicy());
iwc.SetRAMBufferSizeMB(16.0 + 16.0 * Random.NextDouble());
iwc.SetMaxBufferedDocs(-1);
RandomIndexWriter riw = new RandomIndexWriter(Random, dir, iwc);
for (int idx = 0; idx < NUM_DOCS; idx++)
{
Document doc = new Document();
string s = isS1.Get(idx) ? s1 : s2;
Field f = NewTextField("field", s, Field.Store.NO);
int count = TestUtil.NextInt32(Random, 1, 4);
for (int ct = 0; ct < count; ct++)
{
doc.Add(f);
}
riw.AddDocument(doc);
}
r = riw.GetReader();
riw.Dispose();
/*
* if (VERBOSE) {
* System.out.println("TEST: terms");
* TermEnum termEnum = r.Terms();
* while(termEnum.Next()) {
* System.out.println(" term=" + termEnum.Term() + " len=" + termEnum.Term().Text().length());
* Assert.IsTrue(termEnum.DocFreq() > 0);
* System.out.println(" s1?=" + (termEnum.Term().Text().equals(s1)) + " s1len=" + s1.length());
* System.out.println(" s2?=" + (termEnum.Term().Text().equals(s2)) + " s2len=" + s2.length());
* final String s = termEnum.Term().Text();
* for(int idx=0;idx<s.length();idx++) {
* System.out.println(" ch=0x" + Integer.toHexString(s.charAt(idx)));
* }
* }
* }
*/
Assert.AreEqual(NUM_DOCS, r.NumDocs);
Assert.IsTrue(r.DocFreq(new Term("field", s1)) > 0);
Assert.IsTrue(r.DocFreq(new Term("field", s2)) > 0);
int num = AtLeast(1000);
for (int iter = 0; iter < num; iter++)
{
string term;
bool doS1;
if (Random.NextBoolean())
{
term = s1;
doS1 = true;
}
else
{
term = s2;
doS1 = false;
}
if (Verbose)
{
Console.WriteLine("\nTEST: iter=" + iter + " doS1=" + doS1);
}
DocsAndPositionsEnum postings = MultiFields.GetTermPositionsEnum(r, null, "field", new BytesRef(term));
int docID = -1;
while (docID < DocIdSetIterator.NO_MORE_DOCS)
{
int what = Random.Next(3);
if (what == 0)
{
if (Verbose)
{
Console.WriteLine("TEST: docID=" + docID + "; do next()");
}
// nextDoc
int expected = docID + 1;
while (true)
{
if (expected == NUM_DOCS)
{
expected = int.MaxValue;
break;
}
else if (isS1.Get(expected) == doS1)
{
break;
}
else
{
expected++;
}
}
docID = postings.NextDoc();
if (Verbose)
{
Console.WriteLine(" got docID=" + docID);
}
Assert.AreEqual(expected, docID);
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
if (Random.Next(6) == 3)
{
int freq = postings.Freq;
Assert.IsTrue(freq >= 1 && freq <= 4);
for (int pos = 0; pos < freq; pos++)
{
Assert.AreEqual(pos, postings.NextPosition());
if (Random.NextBoolean())
{
var dummy = postings.GetPayload();
if (Random.NextBoolean())
{
dummy = postings.GetPayload(); // get it again
}
}
}
}
}
else
{
// advance
int targetDocID;
if (docID == -1)
{
targetDocID = Random.Next(NUM_DOCS + 1);
}
else
{
targetDocID = docID + TestUtil.NextInt32(Random, 1, NUM_DOCS - docID);
}
if (Verbose)
{
Console.WriteLine("TEST: docID=" + docID + "; do advance(" + targetDocID + ")");
}
int expected = targetDocID;
while (true)
{
if (expected == NUM_DOCS)
{
expected = int.MaxValue;
break;
}
else if (isS1.Get(expected) == doS1)
{
break;
}
else
{
expected++;
}
}
docID = postings.Advance(targetDocID);
if (Verbose)
{
Console.WriteLine(" got docID=" + docID);
}
Assert.AreEqual(expected, docID);
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
if (Random.Next(6) == 3)
{
int freq = postings.Freq;
Assert.IsTrue(freq >= 1 && freq <= 4);
for (int pos = 0; pos < freq; pos++)
{
Assert.AreEqual(pos, postings.NextPosition());
if (Random.NextBoolean())
{
var dummy = postings.GetPayload();
if (Random.NextBoolean())
{
dummy = postings.GetPayload(); // get it again
}
}
}
}
}
}
}
r.Dispose();
dir.Dispose();
}
public virtual void TestDocsWithField()
{
AssumeTrue("codec does not support docsWithField", DefaultCodecSupportsDocsWithField);
Directory dir = NewDirectory();
IndexWriterConfig iwc = NewIndexWriterConfig(Random, TEST_VERSION_CURRENT, null);
iwc.SetMergePolicy(NewLogMergePolicy());
RandomIndexWriter iw = new RandomIndexWriter(Random, dir, iwc);
int numDocs = AtLeast(500);
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
if (Random.Next(4) >= 0)
{
doc.Add(new NumericDocValuesField("numbers", Random.NextInt64()));
}
doc.Add(new NumericDocValuesField("numbersAlways", Random.NextInt64()));
iw.AddDocument(doc);
if (Random.Next(17) == 0)
{
iw.Commit();
}
}
DirectoryReader ir = iw.GetReader();
iw.ForceMerge(1);
DirectoryReader ir2 = iw.GetReader();
AtomicReader merged = GetOnlySegmentReader(ir2);
iw.Dispose();
IBits multi = MultiDocValues.GetDocsWithField(ir, "numbers");
IBits single = merged.GetDocsWithField("numbers");
if (multi == null)
{
Assert.IsNull(single);
}
else
{
Assert.AreEqual(single.Length, multi.Length);
for (int i = 0; i < numDocs; i++)
{
Assert.AreEqual(single.Get(i), multi.Get(i));
}
}
multi = MultiDocValues.GetDocsWithField(ir, "numbersAlways");
single = merged.GetDocsWithField("numbersAlways");
Assert.AreEqual(single.Length, multi.Length);
for (int i = 0; i < numDocs; i++)
{
Assert.AreEqual(single.Get(i), multi.Get(i));
}
ir.Dispose();
ir2.Dispose();
dir.Dispose();
}
public virtual void TestSortedSetWithDups()
{
AssumeTrue("codec does not support SORTED_SET", DefaultCodecSupportsSortedSet);
Directory dir = NewDirectory();
IndexWriterConfig iwc = NewIndexWriterConfig(Random, TEST_VERSION_CURRENT, null);
iwc.SetMergePolicy(NewLogMergePolicy());
RandomIndexWriter iw = new RandomIndexWriter(Random, dir, iwc);
int numDocs = AtLeast(500);
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
int numValues = Random.Next(5);
for (int j = 0; j < numValues; j++)
{
doc.Add(new SortedSetDocValuesField("bytes", new BytesRef(TestUtil.RandomSimpleString(Random, 2))));
}
iw.AddDocument(doc);
if (Random.Next(17) == 0)
{
iw.Commit();
}
}
DirectoryReader ir = iw.GetReader();
iw.ForceMerge(1);
DirectoryReader ir2 = iw.GetReader();
AtomicReader merged = GetOnlySegmentReader(ir2);
iw.Dispose();
SortedSetDocValues multi = MultiDocValues.GetSortedSetValues(ir, "bytes");
SortedSetDocValues single = merged.GetSortedSetDocValues("bytes");
if (multi == null)
{
Assert.IsNull(single);
}
else
{
Assert.AreEqual(single.ValueCount, multi.ValueCount);
BytesRef actual = new BytesRef();
BytesRef expected = new BytesRef();
// check values
for (long i = 0; i < single.ValueCount; i++)
{
single.LookupOrd(i, expected);
multi.LookupOrd(i, actual);
Assert.AreEqual(expected, actual);
}
// check ord list
for (int i = 0; i < numDocs; i++)
{
single.SetDocument(i);
List <long?> expectedList = new List <long?>();
long ord;
while ((ord = single.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
expectedList.Add(ord);
}
multi.SetDocument(i);
int upto = 0;
while ((ord = multi.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
Assert.AreEqual((long)expectedList[upto], ord);
upto++;
}
Assert.AreEqual(expectedList.Count, upto);
}
}
ir.Dispose();
ir2.Dispose();
dir.Dispose();
}
public virtual void TestArbitraryFields()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
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 IterableAnonymousInnerClassHelper(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
{
Debug.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.GetIterator(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.GetTerms(name) == 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 TestRandom()
{
// token -> docID -> tokens
IDictionary <string, IDictionary <int?, IList <Token> > > actualTokens = new Dictionary <string, IDictionary <int?, IList <Token> > >();
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random, dir, iwc);
int numDocs = AtLeast(20);
//final int numDocs = AtLeast(5);
FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
// TODO: randomize what IndexOptions we use; also test
// changing this up in one IW buffered segment...:
ft.IndexOptions = IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
if (Random.NextBoolean())
{
ft.StoreTermVectors = true;
ft.StoreTermVectorOffsets = Random.NextBoolean();
ft.StoreTermVectorPositions = Random.NextBoolean();
}
for (int docCount = 0; docCount < numDocs; docCount++)
{
Document doc = new Document();
doc.Add(new Int32Field("id", docCount, Field.Store.NO));
IList <Token> tokens = new List <Token>();
int numTokens = AtLeast(100);
//final int numTokens = AtLeast(20);
int pos = -1;
int offset = 0;
//System.out.println("doc id=" + docCount);
for (int tokenCount = 0; tokenCount < numTokens; tokenCount++)
{
string text;
if (Random.NextBoolean())
{
text = "a";
}
else if (Random.NextBoolean())
{
text = "b";
}
else if (Random.NextBoolean())
{
text = "c";
}
else
{
text = "d";
}
int posIncr = Random.NextBoolean() ? 1 : Random.Next(5);
if (tokenCount == 0 && posIncr == 0)
{
posIncr = 1;
}
int offIncr = Random.NextBoolean() ? 0 : Random.Next(5);
int tokenOffset = Random.Next(5);
Token token = MakeToken(text, posIncr, offset + offIncr, offset + offIncr + tokenOffset);
if (!actualTokens.TryGetValue(text, out IDictionary <int?, IList <Token> > postingsByDoc))
{
actualTokens[text] = postingsByDoc = new Dictionary <int?, IList <Token> >();
}
if (!postingsByDoc.TryGetValue(docCount, out IList <Token> postings))
{
postingsByDoc[docCount] = postings = new List <Token>();
}
postings.Add(token);
tokens.Add(token);
pos += posIncr;
// stuff abs position into type:
token.Type = "" + pos;
offset += offIncr + tokenOffset;
//System.out.println(" " + token + " posIncr=" + token.getPositionIncrement() + " pos=" + pos + " off=" + token.StartOffset + "/" + token.EndOffset + " (freq=" + postingsByDoc.Get(docCount).Size() + ")");
}
doc.Add(new Field("content", new CannedTokenStream(tokens.ToArray()), ft));
w.AddDocument(doc);
}
DirectoryReader r = w.GetReader();
w.Dispose();
string[] terms = new string[] { "a", "b", "c", "d" };
foreach (AtomicReaderContext ctx in r.Leaves)
{
// TODO: improve this
AtomicReader sub = (AtomicReader)ctx.Reader;
//System.out.println("\nsub=" + sub);
TermsEnum termsEnum = sub.Fields.GetTerms("content").GetIterator(null);
DocsEnum docs = null;
DocsAndPositionsEnum docsAndPositions = null;
DocsAndPositionsEnum docsAndPositionsAndOffsets = null;
FieldCache.Int32s docIDToID = FieldCache.DEFAULT.GetInt32s(sub, "id", false);
foreach (string term in terms)
{
//System.out.println(" term=" + term);
if (termsEnum.SeekExact(new BytesRef(term)))
{
docs = termsEnum.Docs(null, docs);
Assert.IsNotNull(docs);
int doc;
//System.out.println(" doc/freq");
while ((doc = docs.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
IList <Token> expected = actualTokens[term][docIDToID.Get(doc)];
//System.out.println(" doc=" + docIDToID.Get(doc) + " docID=" + doc + " " + expected.Size() + " freq");
Assert.IsNotNull(expected);
Assert.AreEqual(expected.Count, docs.Freq);
}
// explicitly exclude offsets here
docsAndPositions = termsEnum.DocsAndPositions(null, docsAndPositions, DocsAndPositionsFlags.PAYLOADS);
Assert.IsNotNull(docsAndPositions);
//System.out.println(" doc/freq/pos");
while ((doc = docsAndPositions.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
IList <Token> expected = actualTokens[term][docIDToID.Get(doc)];
//System.out.println(" doc=" + docIDToID.Get(doc) + " " + expected.Size() + " freq");
Assert.IsNotNull(expected);
Assert.AreEqual(expected.Count, docsAndPositions.Freq);
foreach (Token token in expected)
{
int pos = Convert.ToInt32(token.Type);
//System.out.println(" pos=" + pos);
Assert.AreEqual(pos, docsAndPositions.NextPosition());
}
}
docsAndPositionsAndOffsets = termsEnum.DocsAndPositions(null, docsAndPositions);
Assert.IsNotNull(docsAndPositionsAndOffsets);
//System.out.println(" doc/freq/pos/offs");
while ((doc = docsAndPositionsAndOffsets.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
IList <Token> expected = actualTokens[term][docIDToID.Get(doc)];
//System.out.println(" doc=" + docIDToID.Get(doc) + " " + expected.Size() + " freq");
Assert.IsNotNull(expected);
Assert.AreEqual(expected.Count, docsAndPositionsAndOffsets.Freq);
foreach (Token token in expected)
{
int pos = Convert.ToInt32(token.Type);
//System.out.println(" pos=" + pos);
Assert.AreEqual(pos, docsAndPositionsAndOffsets.NextPosition());
Assert.AreEqual(token.StartOffset, docsAndPositionsAndOffsets.StartOffset);
Assert.AreEqual(token.EndOffset, docsAndPositionsAndOffsets.EndOffset);
}
}
}
}
// TODO: test advance:
}
r.Dispose();
dir.Dispose();
}
public virtual void TestRandomPositions()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random, dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)).SetMergePolicy(NewLogMergePolicy()));
int numDocs = AtLeast(47);
int max = 1051;
int term = Random.Next(max);
int?[][] positionsInDoc = new int?[numDocs][];
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.OmitNorms = true;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
List <int?> positions = new List <int?>();
StringBuilder builder = new StringBuilder();
int num = AtLeast(131);
for (int j = 0; j < num; j++)
{
int nextInt = Random.Next(max);
builder.Append(nextInt).Append(" ");
if (nextInt == term)
{
positions.Add(Convert.ToInt32(j));
}
}
if (positions.Count == 0)
{
builder.Append(term);
positions.Add(num);
}
doc.Add(NewField(fieldName, builder.ToString(), customType));
positionsInDoc[i] = positions.ToArray();
writer.AddDocument(doc);
}
IndexReader reader = writer.GetReader();
writer.Dispose();
int num_ = AtLeast(13);
for (int i = 0; i < num_; i++)
{
BytesRef bytes = new BytesRef("" + term);
IndexReaderContext topReaderContext = reader.Context;
foreach (AtomicReaderContext atomicReaderContext in topReaderContext.Leaves)
{
DocsAndPositionsEnum docsAndPosEnum = GetDocsAndPositions((AtomicReader)atomicReaderContext.Reader, bytes, null);
Assert.IsNotNull(docsAndPosEnum);
int initDoc = 0;
int maxDoc = atomicReaderContext.Reader.MaxDoc;
// initially advance or do next doc
if (Random.NextBoolean())
{
initDoc = docsAndPosEnum.NextDoc();
}
else
{
initDoc = docsAndPosEnum.Advance(Random.Next(maxDoc));
}
// now run through the scorer and check if all positions are there...
do
{
int docID = docsAndPosEnum.DocID;
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
int?[] pos = positionsInDoc[atomicReaderContext.DocBase + docID];
Assert.AreEqual(pos.Length, docsAndPosEnum.Freq);
// number of positions read should be random - don't read all of them
// allways
int howMany = Random.Next(20) == 0 ? pos.Length - Random.Next(pos.Length) : pos.Length;
for (int j = 0; j < howMany; j++)
{
Assert.AreEqual(pos[j], docsAndPosEnum.NextPosition(), "iteration: " + i + " initDoc: " + initDoc + " doc: " + docID + " base: " + atomicReaderContext.DocBase + " positions: " + pos); /* TODO: + " usePayloads: "
+ usePayload*/
}
if (Random.Next(10) == 0) // once is a while advance
{
if (docsAndPosEnum.Advance(docID + 1 + Random.Next((maxDoc - docID))) == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
}
} while (docsAndPosEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
}
}
reader.Dispose();
dir.Dispose();
}
public virtual void TestRandom()
{
Directory dir = NewDirectory();
int NUM_TERMS = AtLeast(20);
ISet <BytesRef> terms = new JCG.HashSet <BytesRef>();
while (terms.Count < NUM_TERMS)
{
string s = TestUtil.RandomRealisticUnicodeString(Random);
//final String s = TestUtil.RandomSimpleString(random);
if (s.Length > 0)
{
terms.Add(new BytesRef(s));
}
}
BytesRef[] termsArray = terms.ToArray(/*new BytesRef[terms.Count]*/);
Array.Sort(termsArray);
int NUM_DOCS = AtLeast(100);
IndexWriterConfig conf = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random));
// Sometimes swap in codec that impls ord():
if (Random.Next(10) == 7)
{
// Make sure terms index has ords:
Codec codec = TestUtil.AlwaysPostingsFormat(PostingsFormat.ForName("Lucene41WithOrds"));
conf.SetCodec(codec);
}
RandomIndexWriter w = new RandomIndexWriter(Random, dir, conf);
int[][] idToOrds = new int[NUM_DOCS][];
ISet <int?> ordsForDocSet = new JCG.HashSet <int?>();
for (int id = 0; id < NUM_DOCS; id++)
{
Document doc = new Document();
doc.Add(new Int32Field("id", id, Field.Store.NO));
int termCount = TestUtil.NextInt32(Random, 0, 20 * RandomMultiplier);
while (ordsForDocSet.Count < termCount)
{
ordsForDocSet.Add(Random.Next(termsArray.Length));
}
int[] ordsForDoc = new int[termCount];
int upto = 0;
if (Verbose)
{
Console.WriteLine("TEST: doc id=" + id);
}
foreach (int ord in ordsForDocSet)
{
ordsForDoc[upto++] = ord;
Field field = NewStringField("field", termsArray[ord].Utf8ToString(), Field.Store.NO);
if (Verbose)
{
Console.WriteLine(" f=" + termsArray[ord].Utf8ToString());
}
doc.Add(field);
}
ordsForDocSet.Clear();
Array.Sort(ordsForDoc);
idToOrds[id] = ordsForDoc;
w.AddDocument(doc);
}
DirectoryReader r = w.GetReader();
w.Dispose();
if (Verbose)
{
Console.WriteLine("TEST: reader=" + r);
}
foreach (AtomicReaderContext ctx in r.Leaves)
{
if (Verbose)
{
Console.WriteLine("\nTEST: sub=" + ctx.Reader);
}
Verify((AtomicReader)ctx.Reader, idToOrds, termsArray, null);
}
// Also test top-level reader: its enum does not support
// ord, so this forces the OrdWrapper to run:
if (Verbose)
{
Console.WriteLine("TEST: top reader");
}
AtomicReader slowR = SlowCompositeReaderWrapper.Wrap(r);
Verify(slowR, idToOrds, termsArray, null);
FieldCache.DEFAULT.PurgeByCacheKey(slowR.CoreCacheKey);
r.Dispose();
dir.Dispose();
}
public virtual void TestLargeNumberOfPositions()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random, dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)));
int howMany = 1000;
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.OmitNorms = true;
for (int i = 0; i < 39; i++)
{
Document doc = new Document();
StringBuilder builder = new StringBuilder();
for (int j = 0; j < howMany; j++)
{
if (j % 2 == 0)
{
builder.Append("even ");
}
else
{
builder.Append("odd ");
}
}
doc.Add(NewField(fieldName, builder.ToString(), customType));
writer.AddDocument(doc);
}
// now do searches
IndexReader reader = writer.GetReader();
writer.Dispose();
int num = AtLeast(13);
for (int i = 0; i < num; i++)
{
BytesRef bytes = new BytesRef("even");
IndexReaderContext topReaderContext = reader.Context;
foreach (AtomicReaderContext atomicReaderContext in topReaderContext.Leaves)
{
DocsAndPositionsEnum docsAndPosEnum = GetDocsAndPositions((AtomicReader)atomicReaderContext.Reader, bytes, null);
Assert.IsNotNull(docsAndPosEnum);
int initDoc = 0;
int maxDoc = atomicReaderContext.Reader.MaxDoc;
// initially advance or do next doc
if (Random.NextBoolean())
{
initDoc = docsAndPosEnum.NextDoc();
}
else
{
initDoc = docsAndPosEnum.Advance(Random.Next(maxDoc));
}
string msg = "Iteration: " + i + " initDoc: " + initDoc; // TODO: + " payloads: " + usePayload;
Assert.AreEqual(howMany / 2, docsAndPosEnum.Freq);
for (int j = 0; j < howMany; j += 2)
{
Assert.AreEqual(j, docsAndPosEnum.NextPosition(), "position missmatch index: " + j + " with freq: " + docsAndPosEnum.Freq + " -- " + msg);
}
}
}
reader.Dispose();
dir.Dispose();
}
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("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 ThreadAnonymousClass(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 TestIntersectRandom()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
int numTerms = AtLeast(300);
//final int numTerms = 50;
ISet <string> terms = new JCG.HashSet <string>();
ICollection <string> pendingTerms = new 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 * RANDOM_MULTIPLIER; 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 == 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 == 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];
BytesRef actual = te.Next();
if (VERBOSE)
{
Console.WriteLine("TEST: next() expected=" + expected.Utf8ToString() + " actual=" + (actual == 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.IsNull(te.Next());
}
}
r.Dispose();
dir.Dispose();
}
public virtual void TestTermUTF16SortOrder()
{
Random rnd = Random;
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
rnd, dir);
Document d = new Document();
// Single segment
Field f = NewStringField("f", "", Field.Store.NO);
d.Add(f);
char[] chars = new char[2];
ISet <string> allTerms = new JCG.HashSet <string>();
int num = AtLeast(200);
for (int i = 0; i < num; i++)
{
string s;
if (rnd.NextBoolean())
{
// Single char
if (rnd.NextBoolean())
{
// Above surrogates
chars[0] = (char)GetInt(rnd, 1 + UnicodeUtil.UNI_SUR_LOW_END, 0xffff);
}
else
{
// Below surrogates
chars[0] = (char)GetInt(rnd, 0, UnicodeUtil.UNI_SUR_HIGH_START - 1);
}
s = new string(chars, 0, 1);
}
else
{
// Surrogate pair
chars[0] = (char)GetInt(rnd, UnicodeUtil.UNI_SUR_HIGH_START, UnicodeUtil.UNI_SUR_HIGH_END);
Assert.IsTrue(((int)chars[0]) >= UnicodeUtil.UNI_SUR_HIGH_START && ((int)chars[0]) <= UnicodeUtil.UNI_SUR_HIGH_END);
chars[1] = (char)GetInt(rnd, UnicodeUtil.UNI_SUR_LOW_START, UnicodeUtil.UNI_SUR_LOW_END);
s = new string(chars, 0, 2);
}
allTerms.Add(s);
f.SetStringValue(s);
writer.AddDocument(d);
if ((1 + i) % 42 == 0)
{
writer.Commit();
}
}
IndexReader r = writer.GetReader();
// Test each sub-segment
foreach (AtomicReaderContext ctx in r.Leaves)
{
CheckTermsOrder(ctx.Reader, allTerms, false);
}
CheckTermsOrder(r, allTerms, true);
// Test multi segment
r.Dispose();
writer.ForceMerge(1);
// Test single segment
r = writer.GetReader();
CheckTermsOrder(r, allTerms, true);
r.Dispose();
writer.Dispose();
dir.Dispose();
}
public virtual void TestIntersectStartTerm()
{
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", "abc", Field.Store.NO));
w.AddDocument(doc);
doc = new Document();
doc.Add(NewStringField("field", "abd", Field.Store.NO));
w.AddDocument(doc);
doc = new Document();
doc.Add(NewStringField("field", "acd", Field.Store.NO));
w.AddDocument(doc);
doc = new Document();
doc.Add(NewStringField("field", "bcd", 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(".*d", RegExpSyntax.NONE)).ToAutomaton();
CompiledAutomaton ca = new CompiledAutomaton(automaton, false, false);
TermsEnum te;
// should seek to startTerm
te = terms.Intersect(ca, new BytesRef("aad"));
Assert.AreEqual("abd", te.Next().Utf8ToString());
Assert.AreEqual(1, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.AreEqual("acd", te.Next().Utf8ToString());
Assert.AreEqual(2, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.AreEqual("bcd", te.Next().Utf8ToString());
Assert.AreEqual(3, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsNull(te.Next());
// should fail to find ceil label on second arc, rewind
te = terms.Intersect(ca, new BytesRef("add"));
Assert.AreEqual("bcd", te.Next().Utf8ToString());
Assert.AreEqual(3, te.Docs(null, null, DocsFlags.NONE).NextDoc());
Assert.IsNull(te.Next());
// should reach end
te = terms.Intersect(ca, new BytesRef("bcd"));
Assert.IsNull(te.Next());
te = terms.Intersect(ca, new BytesRef("ddd"));
Assert.IsNull(te.Next());
r.Dispose();
dir.Dispose();
}
public virtual void Test2()
{
Random random = Random;
int NUM_DOCS = AtLeast(100);
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
random, dir);
bool allowDups = random.NextBoolean();
HashSet <string> seen = new HashSet <string>();
if (VERBOSE)
{
Console.WriteLine("TEST: NUM_DOCS=" + NUM_DOCS + " allowDups=" + allowDups);
}
int numDocs = 0;
IList <BytesRef> docValues = new List <BytesRef>();
// TODO: deletions
while (numDocs < NUM_DOCS)
{
string s;
if (random.NextBoolean())
{
s = TestUtil.RandomSimpleString(random);
}
else
{
s = TestUtil.RandomUnicodeString(random);
}
BytesRef br = new BytesRef(s);
if (!allowDups)
{
if (seen.Contains(s))
{
continue;
}
seen.Add(s);
}
if (VERBOSE)
{
Console.WriteLine(" " + numDocs + ": s=" + s);
}
Document doc = new Document();
doc.Add(new SortedDocValuesField("stringdv", br));
doc.Add(new NumericDocValuesField("id", numDocs));
docValues.Add(br);
writer.AddDocument(doc);
numDocs++;
if (random.Next(40) == 17)
{
// force flush
writer.GetReader().Dispose();
}
}
writer.ForceMerge(1);
DirectoryReader r = writer.GetReader();
writer.Dispose();
AtomicReader sr = GetOnlySegmentReader(r);
long END_TIME = Environment.TickCount + (TEST_NIGHTLY ? 30 : 1);
int NUM_THREADS = TestUtil.NextInt32(LuceneTestCase.Random, 1, 10);
ThreadClass[] threads = new ThreadClass[NUM_THREADS];
for (int thread = 0; thread < NUM_THREADS; thread++)
{
threads[thread] = new ThreadAnonymousInnerClassHelper2(random, docValues, sr, END_TIME);
threads[thread].Start();
}
foreach (ThreadClass thread in threads)
{
thread.Join();
}
r.Dispose();
dir.Dispose();
}
public virtual void TestRandomWithPrefix()
{
Directory dir = NewDirectory();
ISet <string> prefixes = new JCG.HashSet <string>();
int numPrefix = TestUtil.NextInt32(Random, 2, 7);
if (VERBOSE)
{
Console.WriteLine("TEST: use " + numPrefix + " prefixes");
}
while (prefixes.Count < numPrefix)
{
prefixes.Add(TestUtil.RandomRealisticUnicodeString(Random));
//prefixes.Add(TestUtil.RandomSimpleString(random));
}
string[] prefixesArray = prefixes.ToArray(/*new string[prefixes.Count]*/);
int NUM_TERMS = AtLeast(20);
ISet <BytesRef> terms = new JCG.HashSet <BytesRef>();
while (terms.Count < NUM_TERMS)
{
string s = prefixesArray[Random.Next(prefixesArray.Length)] + TestUtil.RandomRealisticUnicodeString(Random);
//final String s = prefixesArray[random.nextInt(prefixesArray.Length)] + TestUtil.RandomSimpleString(random);
if (s.Length > 0)
{
terms.Add(new BytesRef(s));
}
}
BytesRef[] termsArray = terms.ToArray();
Array.Sort(termsArray);
int NUM_DOCS = AtLeast(100);
IndexWriterConfig conf = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random));
// Sometimes swap in codec that impls ord():
if (Random.Next(10) == 7)
{
Codec codec = TestUtil.AlwaysPostingsFormat(PostingsFormat.ForName("Lucene41WithOrds"));
conf.SetCodec(codec);
}
RandomIndexWriter w = new RandomIndexWriter(Random, dir, conf);
int[][] idToOrds = new int[NUM_DOCS][];
ISet <int?> ordsForDocSet = new JCG.HashSet <int?>();
for (int id = 0; id < NUM_DOCS; id++)
{
Document doc = new Document();
doc.Add(new Int32Field("id", id, Field.Store.NO));
int termCount = TestUtil.NextInt32(Random, 0, 20 * RANDOM_MULTIPLIER);
while (ordsForDocSet.Count < termCount)
{
ordsForDocSet.Add(Random.Next(termsArray.Length));
}
int[] ordsForDoc = new int[termCount];
int upto = 0;
if (VERBOSE)
{
Console.WriteLine("TEST: doc id=" + id);
}
foreach (int ord in ordsForDocSet)
{
ordsForDoc[upto++] = ord;
Field field = NewStringField("field", termsArray[ord].Utf8ToString(), Field.Store.NO);
if (VERBOSE)
{
Console.WriteLine(" f=" + termsArray[ord].Utf8ToString());
}
doc.Add(field);
}
ordsForDocSet.Clear();
Array.Sort(ordsForDoc);
idToOrds[id] = ordsForDoc;
w.AddDocument(doc);
}
DirectoryReader r = w.GetReader();
w.Dispose();
if (VERBOSE)
{
Console.WriteLine("TEST: reader=" + r);
}
AtomicReader slowR = SlowCompositeReaderWrapper.Wrap(r);
foreach (string prefix in prefixesArray)
{
BytesRef prefixRef = prefix == null ? null : new BytesRef(prefix);
int[][] idToOrdsPrefix = new int[NUM_DOCS][];
for (int id = 0; id < NUM_DOCS; id++)
{
int[] docOrds = idToOrds[id];
IList <int?> newOrds = new List <int?>();
foreach (int ord in idToOrds[id])
{
if (StringHelper.StartsWith(termsArray[ord], prefixRef))
{
newOrds.Add(ord);
}
}
int[] newOrdsArray = new int[newOrds.Count];
int upto = 0;
foreach (int ord in newOrds)
{
newOrdsArray[upto++] = ord;
}
idToOrdsPrefix[id] = newOrdsArray;
}
foreach (AtomicReaderContext ctx in r.Leaves)
{
if (VERBOSE)
{
Console.WriteLine("\nTEST: sub=" + ctx.Reader);
}
Verify((AtomicReader)ctx.Reader, idToOrdsPrefix, termsArray, prefixRef);
}
// Also test top-level reader: its enum does not support
// ord, so this forces the OrdWrapper to run:
if (VERBOSE)
{
Console.WriteLine("TEST: top reader");
}
Verify(slowR, idToOrdsPrefix, termsArray, prefixRef);
}
FieldCache.DEFAULT.PurgeByCacheKey(slowR.CoreCacheKey);
r.Dispose();
dir.Dispose();
}
public virtual void TestNumericField()
{
using Directory dir = NewDirectory();
DirectoryReader r = null;
try
{
var numDocs = AtLeast(500);
var answers = new Number[numDocs];
using (var w = new RandomIndexWriter(Random, dir))
{
NumericType[] typeAnswers = new NumericType[numDocs];
for (int id = 0; id < numDocs; id++)
{
Document doc = new Document();
Field nf;
Field sf;
Number answer;
NumericType typeAnswer;
if (Random.NextBoolean())
{
// float/double
if (Random.NextBoolean())
{
float f = Random.NextSingle();
answer = Single.GetInstance(f);
nf = new SingleField("nf", f, Field.Store.NO);
sf = new StoredField("nf", f);
typeAnswer = NumericType.SINGLE;
}
else
{
double d = Random.NextDouble();
answer = Double.GetInstance(d);
nf = new DoubleField("nf", d, Field.Store.NO);
sf = new StoredField("nf", d);
typeAnswer = NumericType.DOUBLE;
}
}
else
{
// int/long
if (Random.NextBoolean())
{
int i = Random.Next();
answer = Int32.GetInstance(i);
nf = new Int32Field("nf", i, Field.Store.NO);
sf = new StoredField("nf", i);
typeAnswer = NumericType.INT32;
}
else
{
long l = Random.NextInt64();
answer = Int64.GetInstance(l);
nf = new Int64Field("nf", l, Field.Store.NO);
sf = new StoredField("nf", l);
typeAnswer = NumericType.INT64;
}
}
doc.Add(nf);
doc.Add(sf);
answers[id] = answer;
typeAnswers[id] = typeAnswer;
FieldType ft = new FieldType(Int32Field.TYPE_STORED);
ft.NumericPrecisionStep = int.MaxValue;
doc.Add(new Int32Field("id", id, ft));
w.AddDocument(doc);
}
r = w.GetReader();
} // w.Dispose();
Assert.AreEqual(numDocs, r.NumDocs);
foreach (AtomicReaderContext ctx in r.Leaves)
{
AtomicReader sub = ctx.AtomicReader;
FieldCache.Int32s ids = FieldCache.DEFAULT.GetInt32s(sub, "id", false);
for (int docID = 0; docID < sub.NumDocs; docID++)
{
Document doc = sub.Document(docID);
Field f = doc.GetField <Field>("nf");
Assert.IsTrue(f is StoredField, "got f=" + f);
#pragma warning disable 612, 618
Assert.AreEqual(answers[ids.Get(docID)], f.GetNumericValue());
#pragma warning restore 612, 618
}
}
}
finally
{
r?.Dispose();
}
}
public override void Run()
{
try
{
while (operations > 0)
{
int oper = rand.Next(100);
if (oper < commitPercent)
{
if (numCommitting.IncrementAndGet() <= maxConcurrentCommits)
{
IDictionary <int, long> newCommittedModel;
long version;
DirectoryReader oldReader;
lock (outerInstance)
{
newCommittedModel = new Dictionary <int, long>(outerInstance.model); // take a snapshot
version = outerInstance.snapshotCount++;
oldReader = outerInstance.reader;
oldReader.IncRef(); // increment the reference since we will use this for reopening
}
DirectoryReader newReader;
if (rand.Next(100) < softCommitPercent)
{
// assertU(h.Commit("softCommit","true"));
if (Random.NextBoolean())
{
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": call writer.getReader");
}
newReader = writer.GetReader(true);
}
else
{
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": reopen reader=" + oldReader + " version=" + version);
}
newReader = DirectoryReader.OpenIfChanged(oldReader, writer.IndexWriter, true);
}
}
else
{
// assertU(commit());
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": commit+reopen reader=" + oldReader + " version=" + version);
}
writer.Commit();
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": now reopen after commit");
}
newReader = DirectoryReader.OpenIfChanged(oldReader);
}
// Code below assumes newReader comes w/
// extra ref:
if (newReader == null)
{
oldReader.IncRef();
newReader = oldReader;
}
oldReader.DecRef();
lock (outerInstance)
{
// install the new reader if it's newest (and check the current version since another reader may have already been installed)
//System.out.println(Thread.currentThread().getName() + ": newVersion=" + newReader.getVersion());
if (Debugging.AssertsEnabled)
{
Debugging.Assert(newReader.RefCount > 0);
}
if (Debugging.AssertsEnabled)
{
Debugging.Assert(outerInstance.reader.RefCount > 0);
}
if (newReader.Version > outerInstance.reader.Version)
{
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": install new reader=" + newReader);
}
outerInstance.reader.DecRef();
outerInstance.reader = newReader;
// Silly: forces fieldInfos to be
// loaded so we don't hit IOE on later
// reader.toString
newReader.ToString();
// install this snapshot only if it's newer than the current one
if (version >= outerInstance.committedModelClock)
{
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": install new model version=" + version);
}
outerInstance.committedModel = newCommittedModel;
outerInstance.committedModelClock = version;
}
else
{
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": skip install new model version=" + version);
}
}
}
else
{
// if the same reader, don't decRef.
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": skip install new reader=" + newReader);
}
newReader.DecRef();
}
}
}
numCommitting.DecrementAndGet();
}
else
{
int id = rand.Next(ndocs);
object sync = outerInstance.syncArr[id];
// set the lastId before we actually change it sometimes to try and
// uncover more race conditions between writing and reading
bool before = Random.NextBoolean();
if (before)
{
outerInstance.lastId = id;
}
// We can't concurrently update the same document and retain our invariants of increasing values
// since we can't guarantee what order the updates will be executed.
lock (sync)
{
long val = outerInstance.model[id];
long nextVal = Math.Abs(val) + 1;
if (oper < commitPercent + deletePercent)
{
// assertU("<delete><id>" + id + "</id></delete>");
// add tombstone first
if (tombstones)
{
Document d = new Document();
d.Add(NewStringField("id", "-" + Convert.ToString(id), Documents.Field.Store.YES));
d.Add(NewField(outerInstance.field, Convert.ToString(nextVal), storedOnlyType));
writer.UpdateDocument(new Term("id", "-" + Convert.ToString(id)), d);
}
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": term delDocs id:" + id + " nextVal=" + nextVal);
}
writer.DeleteDocuments(new Term("id", Convert.ToString(id)));
outerInstance.model[id] = -nextVal;
}
else if (oper < commitPercent + deletePercent + deleteByQueryPercent)
{
//assertU("<delete><query>id:" + id + "</query></delete>");
// add tombstone first
if (tombstones)
{
Document d = new Document();
d.Add(NewStringField("id", "-" + Convert.ToString(id), Documents.Field.Store.YES));
d.Add(NewField(outerInstance.field, Convert.ToString(nextVal), storedOnlyType));
writer.UpdateDocument(new Term("id", "-" + Convert.ToString(id)), d);
}
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": query delDocs id:" + id + " nextVal=" + nextVal);
}
writer.DeleteDocuments(new TermQuery(new Term("id", Convert.ToString(id))));
outerInstance.model[id] = -nextVal;
}
else
{
// assertU(adoc("id",Integer.toString(id), field, Long.toString(nextVal)));
Document d = new Document();
d.Add(NewStringField("id", Convert.ToString(id), Documents.Field.Store.YES));
d.Add(NewField(outerInstance.field, Convert.ToString(nextVal), storedOnlyType));
if (Verbose)
{
Console.WriteLine("TEST: " + Thread.CurrentThread.Name + ": u id:" + id + " val=" + nextVal);
}
writer.UpdateDocument(new Term("id", Convert.ToString(id)), d);
if (tombstones)
{
// remove tombstone after new addition (this should be optional?)
writer.DeleteDocuments(new Term("id", "-" + Convert.ToString(id)));
}
outerInstance.model[id] = nextVal;
}
}
if (!before)
{
outerInstance.lastId = id;
}
}
}
}
catch (Exception e)
{
Console.WriteLine(Thread.CurrentThread.Name + ": FAILED: unexpected exception");
Console.WriteLine(e.StackTrace);
throw new Exception(e.Message, e);
}
}
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();
}