public override FunctionValues GetValues(IDictionary context, AtomicReaderContext readerContext)
{
FieldCache.Ints arr = cache.GetInts(readerContext.AtomicReader, field, parser, true);
Bits valid = cache.GetDocsWithField(readerContext.AtomicReader, field);
return(new IntDocValuesAnonymousInnerClassHelper(this, this, arr, valid));
}
public IntDocValuesAnonymousInnerClassHelper(EnumFieldSource outerInstance, EnumFieldSource @this, FieldCache.Ints arr, Bits valid) : base(@this)
{
this.outerInstance = outerInstance;
this.arr = arr;
this.valid = valid;
val = new MutableValueInt();
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: @Override public org.apache.lucene.queries.function.FunctionValues GetValues(java.util.Map context, org.apache.lucene.index.AtomicReaderContext readerContext) throws java.io.IOException
public override FunctionValues GetValues(IDictionary context, AtomicReaderContext readerContext)
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.search.FieldCache.Ints arr = cache.getInts(readerContext.reader(), field, parser, true);
FieldCache.Ints arr = cache.GetInts(readerContext.AtomicReader, field, parser, true);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.util.Bits valid = cache.getDocsWithField(readerContext.reader(), field);
Bits valid = cache.GetDocsWithField(readerContext.AtomicReader, field);
return(new IntDocValuesAnonymousInnerClassHelper(this, this, arr, valid));
}
public virtual void TestNumericFields()
{
foreach (string name in OldNames)
{
Directory dir = OldIndexDirs[name];
IndexReader reader = DirectoryReader.Open(dir);
IndexSearcher searcher = NewSearcher(reader);
for (int id = 10; id < 15; id++)
{
ScoreDoc[] hits = searcher.Search(NumericRangeQuery.NewIntRange("trieInt", 4, Convert.ToInt32(id), Convert.ToInt32(id), true, true), 100).ScoreDocs;
Assert.AreEqual(1, hits.Length, "wrong number of hits");
Document d = searcher.Doc(hits[0].Doc);
Assert.AreEqual(Convert.ToString(id), d.Get("id"));
hits = searcher.Search(NumericRangeQuery.NewLongRange("trieLong", 4, Convert.ToInt64(id), Convert.ToInt64(id), true, true), 100).ScoreDocs;
Assert.AreEqual(1, hits.Length, "wrong number of hits");
d = searcher.Doc(hits[0].Doc);
Assert.AreEqual(Convert.ToString(id), d.Get("id"));
}
// check that also lower-precision fields are ok
ScoreDoc[] hits_ = searcher.Search(NumericRangeQuery.NewIntRange("trieInt", 4, int.MinValue, int.MaxValue, false, false), 100).ScoreDocs;
Assert.AreEqual(34, hits_.Length, "wrong number of hits");
hits_ = searcher.Search(NumericRangeQuery.NewLongRange("trieLong", 4, long.MinValue, long.MaxValue, false, false), 100).ScoreDocs;
Assert.AreEqual(34, hits_.Length, "wrong number of hits");
// check decoding into field cache
FieldCache.Ints fci = FieldCache.DEFAULT.GetInts(SlowCompositeReaderWrapper.Wrap(searcher.IndexReader), "trieInt", false);
int maxDoc = searcher.IndexReader.MaxDoc;
for (int doc = 0; doc < maxDoc; doc++)
{
int val = fci.Get(doc);
Assert.IsTrue(val >= 0 && val < 35, "value in id bounds");
}
FieldCache.Longs fcl = FieldCache.DEFAULT.GetLongs(SlowCompositeReaderWrapper.Wrap(searcher.IndexReader), "trieLong", false);
for (int doc = 0; doc < maxDoc; doc++)
{
long val = fcl.Get(doc);
Assert.IsTrue(val >= 0L && val < 35L, "value in id bounds");
}
reader.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 = FieldInfo.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 IntField("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.ContainsKey(text))
{
actualTokens[text] = new Dictionary <int?, IList <Token> >();
}
IDictionary <int?, IList <Token> > postingsByDoc = actualTokens[text];
if (!postingsByDoc.ContainsKey(docCount))
{
postingsByDoc[docCount] = new List <Token>();
}
postingsByDoc[docCount].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.Reader;
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.Terms("content").Iterator(null);
DocsEnum docs = null;
DocsAndPositionsEnum docsAndPositions = null;
DocsAndPositionsEnum docsAndPositionsAndOffsets = null;
FieldCache.Ints docIDToID = FieldCache.DEFAULT.GetInts(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, DocsAndPositionsEnum.FLAG_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 TestIntersectRandom()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random(), dir, Similarity, TimeZone);
int numTerms = AtLeast(300);
//final int numTerms = 50;
HashSet <string> terms = new 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];
HashSet <BytesRef> termsSet = new 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.Reader;
w.Dispose();
// NOTE: intentional insanity!!
FieldCache.Ints docIDToID = FieldCache.DEFAULT.GetInts(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:
HashSet <string> acceptTerms = new HashSet <string>();
SortedSet <BytesRef> sortedAcceptTerms = new 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];
HashSet <BytesRef> acceptTermsSet = new 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, DocsEnum.FLAG_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 CustomScoreProviderAnonymousInnerClassHelper(CustomExternalQuery outerInstance, AtomicReaderContext context, FieldCache.Ints values) : base(context)
{
this.outerInstance = outerInstance;
this.context = context;
this.values = values;
}
protected override CustomScoreProvider GetCustomScoreProvider(AtomicReaderContext context)
{
FieldCache.Ints values = FieldCache.DEFAULT.GetInts(context.AtomicReader, INT_FIELD, false);
return(new CustomScoreProviderAnonymousInnerClassHelper(this, context, values));
}
public void TestNumericField()
{
Directory dir = NewDirectory();
var w = new RandomIndexWriter(Random(), dir);
var numDocs = AtLeast(500);
var answers = new object[numDocs];
FieldType.NumericType[] typeAnswers = new FieldType.NumericType[numDocs];
for (int id = 0; id < numDocs; id++)
{
Document doc = new Document();
Field nf;
Field sf;
object answer;
FieldType.NumericType typeAnswer;
if (Random().NextBoolean())
{
// float/double
if (Random().NextBoolean())
{
float f = Random().NextFloat();
answer = Convert.ToSingle(f);
nf = new FloatField("nf", f, Field.Store.NO);
sf = new StoredField("nf", f);
typeAnswer = FieldType.NumericType.FLOAT;
}
else
{
double d = Random().NextDouble();
answer = Convert.ToDouble(d);
nf = new DoubleField("nf", d, Field.Store.NO);
sf = new StoredField("nf", d);
typeAnswer = FieldType.NumericType.DOUBLE;
}
}
else
{
// int/long
if (Random().NextBoolean())
{
int i = Random().Next();
answer = Convert.ToInt32(i);
nf = new IntField("nf", i, Field.Store.NO);
sf = new StoredField("nf", i);
typeAnswer = FieldType.NumericType.INT;
}
else
{
long l = Random().NextLong();
answer = Convert.ToInt64(l);
nf = new LongField("nf", l, Field.Store.NO);
sf = new StoredField("nf", l);
typeAnswer = FieldType.NumericType.LONG;
}
}
doc.Add(nf);
doc.Add(sf);
answers[id] = answer;
typeAnswers[id] = typeAnswer;
FieldType ft = new FieldType(IntField.TYPE_STORED);
ft.NumericPrecisionStep = int.MaxValue;
doc.Add(new IntField("id", id, ft));
w.AddDocument(doc);
}
DirectoryReader r = w.Reader;
w.Dispose();
Assert.AreEqual(numDocs, r.NumDocs);
foreach (AtomicReaderContext ctx in r.Leaves)
{
AtomicReader sub = (AtomicReader)ctx.Reader;
FieldCache.Ints ids = FieldCache.DEFAULT.GetInts(sub, "id", false);
for (int docID = 0; docID < sub.NumDocs; docID++)
{
Document doc = sub.Document(docID);
Field f = (Field)doc.GetField("nf");
Assert.IsTrue(f is StoredField, "got f=" + f);
Assert.AreEqual(answers[ids.Get(docID)], f.NumericValue);
}
}
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.NextInt(Random(), 2, 10));
FieldCache.Ints docIDToID = FieldCache.DEFAULT.GetInts(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").Iterator(null);
int ord = 0;
while (allTE.Next() != null)
{
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.Iterator(null);
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.Next() == null)
{
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.Document = 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 void TestRandom()
{
int numberOfRuns = TestUtil.NextInt(Random(), 3, 6);
for (int iter = 0; iter < numberOfRuns; iter++)
{
if (VERBOSE)
{
Console.WriteLine(string.Format("TEST: iter={0} total={1}", iter, numberOfRuns));
}
int numDocs = TestUtil.NextInt(Random(), 100, 1000) * RANDOM_MULTIPLIER;
int numGroups = TestUtil.NextInt(Random(), 1, numDocs);
if (VERBOSE)
{
Console.WriteLine("TEST: numDocs=" + numDocs + " numGroups=" + numGroups);
}
List <BytesRef> groups = new List <BytesRef>();
for (int i = 0; i < numGroups; i++)
{
string randomValue;
do
{
// B/c of DV based impl we can't see the difference between an empty string and a null value.
// For that reason we don't generate empty string groups.
randomValue = TestUtil.RandomRealisticUnicodeString(Random());
} while ("".Equals(randomValue, StringComparison.Ordinal));
groups.Add(new BytesRef(randomValue));
}
string[] contentStrings = new string[TestUtil.NextInt(Random(), 2, 20)];
if (VERBOSE)
{
Console.WriteLine("TEST: create fake content");
}
for (int contentIDX = 0; contentIDX < contentStrings.Length; contentIDX++)
{
StringBuilder sb = new StringBuilder();
sb.append("real").append(Random().nextInt(3)).append(' ');
int fakeCount = Random().nextInt(10);
for (int fakeIDX = 0; fakeIDX < fakeCount; fakeIDX++)
{
sb.append("fake ");
}
contentStrings[contentIDX] = sb.toString();
if (VERBOSE)
{
Console.WriteLine(" content=" + sb.toString());
}
}
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
Random(),
dir,
NewIndexWriterConfig(TEST_VERSION_CURRENT,
new MockAnalyzer(Random())));
bool preFlex = "Lucene3x".Equals(w.w.Config.Codec.Name, StringComparison.Ordinal);
bool canUseIDV = !preFlex;
FieldInfo.DocValuesType_e valueType = vts[Random().nextInt(vts.Length)];
Document doc = new Document();
Document docNoGroup = new Document();
Field group = NewStringField("group", "", Field.Store.NO);
doc.Add(group);
Field valuesField = null;
if (canUseIDV)
{
switch (valueType)
{
case FieldInfo.DocValuesType_e.BINARY:
valuesField = new BinaryDocValuesField("group_dv", new BytesRef());
break;
case FieldInfo.DocValuesType_e.SORTED:
valuesField = new SortedDocValuesField("group_dv", new BytesRef());
break;
default:
fail("unhandled type");
break;
}
doc.Add(valuesField);
}
Field sort1 = NewStringField("sort1", "", Field.Store.NO);
doc.Add(sort1);
docNoGroup.Add(sort1);
Field sort2 = NewStringField("sort2", "", Field.Store.NO);
doc.Add(sort2);
docNoGroup.Add(sort2);
Field sort3 = NewStringField("sort3", "", Field.Store.NO);
doc.Add(sort3);
docNoGroup.Add(sort3);
Field content = NewTextField("content", "", Field.Store.NO);
doc.Add(content);
docNoGroup.Add(content);
IntField id = new IntField("id", 0, Field.Store.NO);
doc.Add(id);
docNoGroup.Add(id);
GroupDoc[] groupDocs = new GroupDoc[numDocs];
for (int i = 0; i < numDocs; i++)
{
BytesRef groupValue;
if (Random().nextInt(24) == 17)
{
// So we test the "doc doesn't have the group'd
// field" case:
groupValue = null;
}
else
{
groupValue = groups[Random().nextInt(groups.size())];
}
GroupDoc groupDoc = new GroupDoc(
i,
groupValue,
groups[Random().nextInt(groups.size())],
groups[Random().nextInt(groups.size())],
new BytesRef(string.Format(CultureInfo.InvariantCulture, "{0:D5}", i)),
contentStrings[Random().nextInt(contentStrings.Length)]
);
if (VERBOSE)
{
Console.WriteLine(" doc content=" + groupDoc.content + " id=" + i + " group=" + (groupDoc.group == null ? "null" : groupDoc.group.Utf8ToString()) + " sort1=" + groupDoc.sort1.Utf8ToString() + " sort2=" + groupDoc.sort2.Utf8ToString() + " sort3=" + groupDoc.sort3.Utf8ToString());
}
groupDocs[i] = groupDoc;
if (groupDoc.group != null)
{
group.StringValue = (groupDoc.group.Utf8ToString());
if (canUseIDV)
{
valuesField.BytesValue = (new BytesRef(groupDoc.group.Utf8ToString()));
}
}
sort1.StringValue = (groupDoc.sort1.Utf8ToString());
sort2.StringValue = (groupDoc.sort2.Utf8ToString());
sort3.StringValue = (groupDoc.sort3.Utf8ToString());
content.StringValue = (groupDoc.content);
id.IntValue = (groupDoc.id);
if (groupDoc.group == null)
{
w.AddDocument(docNoGroup);
}
else
{
w.AddDocument(doc);
}
}
DirectoryReader r = w.Reader;
w.Dispose();
// NOTE: intentional but temporary field cache insanity!
FieldCache.Ints docIdToFieldId = FieldCache.DEFAULT.GetInts(SlowCompositeReaderWrapper.Wrap(r), "id", false);
int[] fieldIdToDocID = new int[numDocs];
for (int i = 0; i < numDocs; i++)
{
int fieldId = docIdToFieldId.Get(i);
fieldIdToDocID[fieldId] = i;
}
try
{
IndexSearcher s = NewSearcher(r);
if (typeof(SlowCompositeReaderWrapper).IsAssignableFrom(s.IndexReader.GetType()))
{
canUseIDV = false;
}
else
{
canUseIDV = !preFlex;
}
for (int contentID = 0; contentID < 3; contentID++)
{
ScoreDoc[] hits = s.Search(new TermQuery(new Term("content", "real" + contentID)), numDocs).ScoreDocs;
foreach (ScoreDoc hit in hits)
{
GroupDoc gd = groupDocs[docIdToFieldId.Get(hit.Doc)];
assertTrue(gd.score == 0.0);
gd.score = hit.Score;
int docId = gd.id;
assertEquals(docId, docIdToFieldId.Get(hit.Doc));
}
}
foreach (GroupDoc gd in groupDocs)
{
assertTrue(gd.score != 0.0);
}
for (int searchIter = 0; searchIter < 100; searchIter++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: searchIter=" + searchIter);
}
string searchTerm = "real" + Random().nextInt(3);
bool sortByScoreOnly = Random().nextBoolean();
Sort sortWithinGroup = GetRandomSort(sortByScoreOnly);
AbstractAllGroupHeadsCollector allGroupHeadsCollector = CreateRandomCollector("group", sortWithinGroup, canUseIDV, valueType);
s.Search(new TermQuery(new Term("content", searchTerm)), allGroupHeadsCollector);
int[] expectedGroupHeads = CreateExpectedGroupHeads(searchTerm, groupDocs, sortWithinGroup, sortByScoreOnly, fieldIdToDocID);
int[] actualGroupHeads = allGroupHeadsCollector.RetrieveGroupHeads();
// The actual group heads contains Lucene ids. Need to change them into our id value.
for (int i = 0; i < actualGroupHeads.Length; i++)
{
actualGroupHeads[i] = docIdToFieldId.Get(actualGroupHeads[i]);
}
// Allows us the easily iterate and assert the actual and expected results.
Array.Sort(expectedGroupHeads);
Array.Sort(actualGroupHeads);
if (VERBOSE)
{
Console.WriteLine("Collector: " + allGroupHeadsCollector.GetType().Name);
Console.WriteLine("Sort within group: " + sortWithinGroup);
Console.WriteLine("Num group: " + numGroups);
Console.WriteLine("Num doc: " + numDocs);
Console.WriteLine("\n=== Expected: \n");
foreach (int expectedDocId in expectedGroupHeads)
{
GroupDoc expectedGroupDoc = groupDocs[expectedDocId];
string expectedGroup = expectedGroupDoc.group == null ? null : expectedGroupDoc.group.Utf8ToString();
Console.WriteLine(
string.Format(CultureInfo.InvariantCulture,
"Group:{0,10} score{1:0.0#######,5} Sort1:{2,10} Sort2:{3,10} Sort3:{4,10} doc:{5,10}",
expectedGroup, expectedGroupDoc.score, expectedGroupDoc.sort1.Utf8ToString(),
expectedGroupDoc.sort2.Utf8ToString(), expectedGroupDoc.sort3.Utf8ToString(), expectedDocId)
);
}
Console.WriteLine("\n=== Actual: \n");
foreach (int actualDocId in actualGroupHeads)
{
GroupDoc actualGroupDoc = groupDocs[actualDocId];
string actualGroup = actualGroupDoc.group == null ? null : actualGroupDoc.group.Utf8ToString();
Console.WriteLine(
string.Format(CultureInfo.InvariantCulture,
"Group:{0,10} score{1:0.0#######,5} Sort1:{2,10} Sort2:{3,10} Sort3:{4,10} doc:{5,10}",
actualGroup, actualGroupDoc.score, actualGroupDoc.sort1.Utf8ToString(),
actualGroupDoc.sort2.Utf8ToString(), actualGroupDoc.sort3.Utf8ToString(), actualDocId)
);
}
Console.WriteLine("\n===================================================================================");
}
assertArrayEquals(expectedGroupHeads, actualGroupHeads);
}
}
finally
{
QueryUtils.PurgeFieldCache(r);
}
r.Dispose();
dir.Dispose();
}
}
