private bool TermNotInReader(AtomicReader reader, Term term)
{
// only called from assert
//System.out.println("TQ.termNotInReader reader=" + reader + " term=" + field + ":" + bytes.utf8ToString());
return(reader.DocFreq(term) == 0);
}
/// <summary>
/// Returns the a <see cref="DocIdSetIterator"/> representing the Boolean composition
/// of the filters that have been added.
/// </summary>
public override DocIdSet GetDocIdSet(AtomicReaderContext context, IBits acceptDocs)
{
FixedBitSet res = null;
AtomicReader reader = context.AtomicReader;
bool hasShouldClauses = false;
foreach (FilterClause fc in clauses)
{
if (fc.Occur == Occur.SHOULD)
{
hasShouldClauses = true;
DocIdSetIterator disi = GetDISI(fc.Filter, context);
if (disi == null)
{
continue;
}
if (res == null)
{
res = new FixedBitSet(reader.MaxDoc);
}
res.Or(disi);
}
}
if (hasShouldClauses && res == null)
{
return(null);
}
foreach (FilterClause fc in clauses)
{
if (fc.Occur == Occur.MUST_NOT)
{
if (res == null)
{
Debug.Assert(!hasShouldClauses);
res = new FixedBitSet(reader.MaxDoc);
res.Set(0, reader.MaxDoc); // NOTE: may set bits on deleted docs
}
DocIdSetIterator disi = GetDISI(fc.Filter, context);
if (disi != null)
{
res.AndNot(disi);
}
}
}
foreach (FilterClause fc in clauses)
{
if (fc.Occur == Occur.MUST)
{
DocIdSetIterator disi = GetDISI(fc.Filter, context);
if (disi == null)
{
return(null); // no documents can match
}
if (res == null)
{
res = new FixedBitSet(reader.MaxDoc);
res.Or(disi);
}
else
{
res.And(disi);
}
}
}
return(BitsFilteredDocIdSet.Wrap(res, acceptDocs));
}
public override Scorer GetScorer(AtomicReaderContext context, IBits acceptDocs)
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(outerInstance.terms.Count > 0);
}
AtomicReader reader = context.AtomicReader;
IBits liveDocs = acceptDocs;
PostingsAndFreq[] postingsFreqs = new PostingsAndFreq[outerInstance.terms.Count];
Terms fieldTerms = reader.GetTerms(outerInstance.field);
if (fieldTerms == null)
{
return(null);
}
// Reuse single TermsEnum below:
TermsEnum te = fieldTerms.GetEnumerator();
for (int i = 0; i < outerInstance.terms.Count; i++)
{
Term t = outerInstance.terms[i];
TermState state = states[i].Get(context.Ord);
if (state == null) // term doesnt exist in this segment
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(TermNotInReader(reader, t), "no termstate found but term exists in reader");
}
return(null);
}
te.SeekExact(t.Bytes, state);
DocsAndPositionsEnum postingsEnum = te.DocsAndPositions(liveDocs, null, DocsAndPositionsFlags.NONE);
// PhraseQuery on a field that did not index
// positions.
if (postingsEnum == null)
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(te.SeekExact(t.Bytes), "termstate found but no term exists in reader");
}
// term does exist, but has no positions
throw IllegalStateException.Create("field \"" + t.Field + "\" was indexed without position data; cannot run PhraseQuery (term=" + t.Text + ")");
}
postingsFreqs[i] = new PostingsAndFreq(postingsEnum, te.DocFreq, outerInstance.positions[i], t);
}
// sort by increasing docFreq order
if (outerInstance.slop == 0)
{
ArrayUtil.TimSort(postingsFreqs);
}
if (outerInstance.slop == 0) // optimize exact case
{
ExactPhraseScorer s = new ExactPhraseScorer(this, postingsFreqs, similarity.GetSimScorer(stats, context));
if (s.noDocs)
{
return(null);
}
else
{
return(s);
}
}
else
{
return(new SloppyPhraseScorer(this, postingsFreqs, outerInstance.slop, similarity.GetSimScorer(stats, context)));
}
}
public DocIdSetAnonymousInnerClassHelper(FilterAnonymousInnerClassHelper3 outerInstance, bool nullBitset, AtomicReader reader, BitArray bitSet)
{
this.OuterInstance = outerInstance;
this.NullBitset = nullBitset;
this.Reader = reader;
this.BitSet = bitSet;
}
internal readonly Sorter.DocMap docMap; // pkg-protected to avoid synthetic accessor methods
#endregion Fields
#region Constructors
//JAVA TO C# CONVERTER WARNING: 'final' parameters are not available in .NET:
//ORIGINAL LINE: private SortingAtomicReader(final org.apache.lucene.index.AtomicReader in, final Sorter.DocMap docMap)
private SortingAtomicReader(AtomicReader @in, Sorter.DocMap docMap)
: base(@in)
{
this.docMap = docMap;
}
public CollectorAnonymousInnerClassHelper(Query q, IndexSearcher s, IList<AtomicReaderContext> readerContextArray,
int skip_op, int[] order, int[] opidx, int[] lastDoc, float maxDiff, AtomicReader[] lastReader,
Similarity similarity)
{
this.q = q;
this.s = s;
this.ReaderContextArray = readerContextArray;
this.Skip_op = skip_op;
this.Order = order;
this.Opidx = opidx;
this.LastDoc = lastDoc;
this.MaxDiff = maxDiff;
this.LastReader = lastReader;
this.Similarity = similarity;
}
private IEnumerable<long?> GetMergeSortedSetDocToOrdCountEnumerable(AtomicReader[] readers, SortedSetDocValues[] dvs)
{
int readerUpto = -1;
int docIDUpto = 0;
AtomicReader currentReader = null;
Bits currentLiveDocs = null;
while (true)
{
if (readerUpto == readers.Length)
{
yield break;
}
if (currentReader == null || docIDUpto == currentReader.MaxDoc)
{
readerUpto++;
if (readerUpto < readers.Length)
{
currentReader = readers[readerUpto];
currentLiveDocs = currentReader.LiveDocs;
}
docIDUpto = 0;
continue;
}
if (currentLiveDocs == null || currentLiveDocs.Get(docIDUpto))
{
SortedSetDocValues dv = dvs[readerUpto];
dv.Document = docIDUpto;
long value = 0;
while (dv.NextOrd() != SortedSetDocValues.NO_MORE_ORDS)
{
value++;
}
docIDUpto++;
yield return value;
continue;
}
docIDUpto++;
}
}
public override DocComparer GetComparer(AtomicReader reader, int docbase)
{
m_luceneComparer.SetNextReader(reader.AtomicContext);
return(new LuceneCustomDocComparer(m_luceneComparer));
}
internal DelegatingAtomicReader(AtomicReader reader) : base(reader)
{
}
private int CopyFieldsWithDeletions(MergeState mergeState, AtomicReader reader, Lucene40StoredFieldsReader matchingFieldsReader, int[] rawDocLengths)
{
int docCount = 0;
int maxDoc = reader.MaxDoc();
Bits liveDocs = reader.LiveDocs;
Debug.Assert(liveDocs != null);
if (matchingFieldsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
for (int j = 0; j < maxDoc; )
{
if (!liveDocs.Get(j))
{
// skip deleted docs
++j;
continue;
}
// We can optimize this case (doing a bulk byte copy) since the field
// numbers are identical
int start = j, numDocs = 0;
do
{
j++;
numDocs++;
if (j >= maxDoc)
{
break;
}
if (!liveDocs.Get(j))
{
j++;
break;
}
} while (numDocs < MAX_RAW_MERGE_DOCS);
IndexInput stream = matchingFieldsReader.RawDocs(rawDocLengths, start, numDocs);
AddRawDocuments(stream, rawDocLengths, numDocs);
docCount += numDocs;
mergeState.checkAbort.Work(300 * numDocs);
}
}
else
{
for (int j = 0; j < maxDoc; j++)
{
if (!liveDocs.Get(j))
{
// skip deleted docs
continue;
}
// TODO: this could be more efficient using
// FieldVisitor instead of loading/writing entire
// doc; ie we just have to renumber the field number
// on the fly?
// NOTE: it's very important to first assign to doc then pass it to
// fieldsWriter.addDocument; see LUCENE-1282
Document doc = reader.Document(j);
AddDocument(doc, mergeState.FieldInfos);
docCount++;
mergeState.checkAbort.Work(300);
}
}
return docCount;
}
private int CopyFieldsNoDeletions(MergeState mergeState, AtomicReader reader, Lucene40StoredFieldsReader matchingFieldsReader, int[] rawDocLengths)
{
int maxDoc = reader.MaxDoc();
int docCount = 0;
if (matchingFieldsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
while (docCount < maxDoc)
{
int len = Math.Min(MAX_RAW_MERGE_DOCS, maxDoc - docCount);
IndexInput stream = matchingFieldsReader.RawDocs(rawDocLengths, docCount, len);
AddRawDocuments(stream, rawDocLengths, len);
docCount += len;
mergeState.checkAbort.Work(300 * len);
}
}
else
{
for (; docCount < maxDoc; docCount++)
{
// NOTE: it's very important to first assign to doc then pass it to
// fieldsWriter.addDocument; see LUCENE-1282
Document doc = reader.Document(docCount);
AddDocument(doc, mergeState.FieldInfos);
mergeState.checkAbort.Work(300);
}
}
return docCount;
}
/// <summary>Train the classifier using the underlying Lucene index</summary>
/// <param name="analyzer">the analyzer used to tokenize / filter the unseen text</param>
/// <param name="atomicReader">the reader to use to access the Lucene index</param>
/// <param name="classFieldName">the name of the field containing the class assigned to documents</param>
/// <param name="query">the query to filter which documents use for training</param>
/// <param name="textFieldName">the name of the field used to compare documents</param>
public virtual void Train(AtomicReader atomicReader, string textFieldName, string classFieldName, Analyzer analyzer, Query query)
{
Train(atomicReader, new string[] { textFieldName }, classFieldName, analyzer, query);
}
/// <summary>
/// Train the classifier using the underlying Lucene index
/// </summary>
/// <param name="analyzer"> the analyzer used to tokenize / filter the unseen text</param>
/// <param name="atomicReader">the reader to use to access the Lucene index</param>
/// <param name="classFieldName">the name of the field containing the class assigned to documents</param>
/// <param name="textFieldName">the name of the field used to compare documents</param>
public virtual void Train(AtomicReader atomicReader, string textFieldName, string classFieldName, Analyzer analyzer)
{
Train(atomicReader, textFieldName, classFieldName, analyzer, null);
}
public void BeforeClassSorterTestBase()
{
dir = NewDirectory();
int numDocs = AtLeast(20);
CreateIndex(dir, numDocs, Random());
reader = SlowCompositeReaderWrapper.Wrap(DirectoryReader.Open(dir));
}
public void Test()
{
RandomIndexWriter writer;
DirectoryReader indexReader;
int numParents = AtLeast(200);
IndexWriterConfig cfg = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random));
cfg.SetMergePolicy(NewLogMergePolicy());
using (writer = new RandomIndexWriter(Random, NewDirectory(), cfg))
{
Document parentDoc = new Document();
NumericDocValuesField parentVal = new NumericDocValuesField("parent_val", 0L);
parentDoc.Add(parentVal);
StringField parent = new StringField("parent", "true", Field.Store.YES);
parentDoc.Add(parent);
for (int i = 0; i < numParents; ++i)
{
List <Document> documents = new List <Document>();
int numChildren = Random.nextInt(10);
for (int j = 0; j < numChildren; ++j)
{
Document childDoc = new Document();
childDoc.Add(new NumericDocValuesField("child_val", Random.nextInt(5)));
documents.Add(childDoc);
}
parentVal.SetInt64Value(Random.nextInt(50));
documents.Add(parentDoc);
writer.AddDocuments(documents);
}
writer.ForceMerge(1);
indexReader = writer.GetReader();
}
AtomicReader reader = GetOnlySegmentReader(indexReader);
Filter parentsFilter = new FixedBitSetCachingWrapperFilter(new QueryWrapperFilter(new TermQuery(new Term("parent", "true"))));
FixedBitSet parentBits = (FixedBitSet)parentsFilter.GetDocIdSet(reader.AtomicContext, null);
NumericDocValues parentValues = reader.GetNumericDocValues("parent_val");
NumericDocValues childValues = reader.GetNumericDocValues("child_val");
Sort parentSort = new Sort(new SortField("parent_val", SortFieldType.INT64));
Sort childSort = new Sort(new SortField("child_val", SortFieldType.INT64));
Sort sort = new Sort(new SortField("custom", new BlockJoinComparerSource(parentsFilter, parentSort, childSort)));
Sorter sorter = new Sorter(sort);
Sorter.DocMap docMap = sorter.Sort(reader);
assertEquals(reader.MaxDoc, docMap.Count);
int[] children = new int[1];
int numChildren2 = 0;
int previousParent = -1;
for (int i = 0; i < docMap.Count; ++i)
{
int oldID = docMap.NewToOld(i);
if (parentBits.Get(oldID))
{
// check that we have the right children
for (int j = 0; j < numChildren2; ++j)
{
assertEquals(oldID, parentBits.NextSetBit(children[j]));
}
// check that children are sorted
for (int j = 1; j < numChildren2; ++j)
{
int doc1 = children[j - 1];
int doc2 = children[j];
if (childValues.Get(doc1) == childValues.Get(doc2))
{
assertTrue(doc1 < doc2); // sort is stable
}
else
{
assertTrue(childValues.Get(doc1) < childValues.Get(doc2));
}
}
// check that parents are sorted
if (previousParent != -1)
{
if (parentValues.Get(previousParent) == parentValues.Get(oldID))
{
assertTrue(previousParent < oldID);
}
else
{
assertTrue(parentValues.Get(previousParent) < parentValues.Get(oldID));
}
}
// reset
previousParent = oldID;
numChildren2 = 0;
}
else
{
children = ArrayUtil.Grow(children, numChildren2 + 1);
children[numChildren2++] = oldID;
}
}
indexReader.Dispose();
writer.IndexWriter.Directory.Dispose();
}
public DocIdSetAnonymousInnerClassHelper(FilterAnonymousInnerClassHelper3 outerInstance, bool nullBitset, AtomicReader reader, BitSet bitSet)
{
this.outerInstance = outerInstance;
this.nullBitset = nullBitset;
this.reader = reader;
this.bitSet = bitSet;
}
public override void BeforeClass()
{
base.BeforeClass();
NUM_DOCS = AtLeast(500);
NUM_ORDS = AtLeast(2);
Directory = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random(), Directory, NewIndexWriterConfig(Random(), TEST_VERSION_CURRENT, new MockAnalyzer(Random()), Similarity, TimeZone).SetMergePolicy(NewLogMergePolicy()));
long theLong = long.MaxValue;
double theDouble = double.MaxValue;
sbyte theByte = sbyte.MaxValue;
short theShort = short.MaxValue;
int theInt = int.MaxValue;
float theFloat = float.MaxValue;
UnicodeStrings = new string[NUM_DOCS];
//MultiValued = new BytesRef[NUM_DOCS, NUM_ORDS];
MultiValued = RectangularArrays.ReturnRectangularArray <BytesRef>(NUM_DOCS, NUM_ORDS);
if (VERBOSE)
{
Console.WriteLine("TEST: setUp");
}
for (int i = 0; i < NUM_DOCS; i++)
{
Document doc = new Document();
doc.Add(NewStringField("theLong", (theLong--).ToString(CultureInfo.InvariantCulture), Field.Store.NO));
doc.Add(NewStringField("theDouble", (theDouble--).ToString("R", CultureInfo.InvariantCulture), Field.Store.NO));
doc.Add(NewStringField("theByte", (theByte--).ToString(CultureInfo.InvariantCulture), Field.Store.NO));
doc.Add(NewStringField("theShort", (theShort--).ToString(CultureInfo.InvariantCulture), Field.Store.NO));
doc.Add(NewStringField("theInt", (theInt--).ToString(CultureInfo.InvariantCulture), Field.Store.NO));
doc.Add(NewStringField("theFloat", (theFloat--).ToString("R", CultureInfo.InvariantCulture), Field.Store.NO));
if (i % 2 == 0)
{
doc.Add(NewStringField("sparse", (i).ToString(CultureInfo.InvariantCulture), Field.Store.NO));
}
if (i % 2 == 0)
{
doc.Add(new Int32Field("numInt", i, Field.Store.NO));
}
// sometimes skip the field:
if (Random().Next(40) != 17)
{
UnicodeStrings[i] = GenerateString(i);
doc.Add(NewStringField("theRandomUnicodeString", UnicodeStrings[i], Field.Store.YES));
}
// sometimes skip the field:
if (Random().Next(10) != 8)
{
for (int j = 0; j < NUM_ORDS; j++)
{
string newValue = GenerateString(i);
MultiValued[i][j] = new BytesRef(newValue);
doc.Add(NewStringField("theRandomUnicodeMultiValuedField", newValue, Field.Store.YES));
}
Array.Sort(MultiValued[i]);
}
writer.AddDocument(doc);
}
IndexReader r = writer.Reader;
Reader = SlowCompositeReaderWrapper.Wrap(r);
writer.Dispose();
}
private IDictionary <int, object> HighlightField(string field, string[] contents, BreakIterator bi, BytesRef[] terms, int[] docids, IList <AtomicReaderContext> leaves, int maxPassages, Query query)
{
IDictionary <int, object> highlights = new Dictionary <int, object>();
PassageFormatter fieldFormatter = GetFormatter(field);
if (fieldFormatter == null)
{
throw new NullReferenceException("PassageFormatter cannot be null");
}
// check if we should do any multiterm processing
Analyzer analyzer = GetIndexAnalyzer(field);
CharacterRunAutomaton[] automata = new CharacterRunAutomaton[0];
if (analyzer != null)
{
automata = MultiTermHighlighting.ExtractAutomata(query, field);
}
// resize 'terms', where the last term is the multiterm matcher
if (automata.Length > 0)
{
BytesRef[] newTerms = new BytesRef[terms.Length + 1];
System.Array.Copy(terms, 0, newTerms, 0, terms.Length);
terms = newTerms;
}
// we are processing in increasing docid order, so we only need to reinitialize stuff on segment changes
// otherwise, we will just advance() existing enums to the new document in the same segment.
DocsAndPositionsEnum[] postings = null;
TermsEnum termsEnum = null;
int lastLeaf = -1;
for (int i = 0; i < docids.Length; i++)
{
string content = contents[i];
if (content.Length == 0)
{
continue; // nothing to do
}
bi.SetText(content);
int doc = docids[i];
int leaf = ReaderUtil.SubIndex(doc, leaves);
AtomicReaderContext subContext = leaves[leaf];
AtomicReader r = subContext.AtomicReader;
Debug.Assert(leaf >= lastLeaf); // increasing order
// if the segment has changed, we must initialize new enums.
if (leaf != lastLeaf)
{
Terms t = r.GetTerms(field);
if (t != null)
{
termsEnum = t.GetIterator(null);
postings = new DocsAndPositionsEnum[terms.Length];
}
}
if (termsEnum == null)
{
continue; // no terms for this field, nothing to do
}
// if there are multi-term matches, we have to initialize the "fake" enum for each document
if (automata.Length > 0)
{
DocsAndPositionsEnum dp = MultiTermHighlighting.GetDocsEnum(analyzer.GetTokenStream(field, content), automata);
dp.Advance(doc - subContext.DocBase);
postings[terms.Length - 1] = dp; // last term is the multiterm matcher
}
Passage[] passages = HighlightDoc(field, terms, content.Length, bi, doc - subContext.DocBase, termsEnum, postings, maxPassages);
if (passages.Length == 0)
{
// no passages were returned, so ask for a default summary
passages = GetEmptyHighlight(field, bi, maxPassages);
}
if (passages.Length > 0)
{
highlights[doc] = fieldFormatter.Format(passages, content);
}
lastLeaf = leaf;
}
return(highlights);
}
public CollectorAnonymousInnerClassHelper2(Query q, IndexSearcher s, float maxDiff, int[] lastDoc, AtomicReader[] lastReader, IList<AtomicReaderContext> context, Similarity similarity)
{
this.q = q;
this.s = s;
this.MaxDiff = maxDiff;
this.LastDoc = lastDoc;
this.LastReader = lastReader;
this.Context = context;
this.Similarity = similarity;
}
private int CopyVectorsWithDeletions(MergeState mergeState, Lucene40TermVectorsReader matchingVectorsReader, AtomicReader reader, int[] rawDocLengths, int[] rawDocLengths2)
{
int maxDoc = reader.MaxDoc;
IBits liveDocs = reader.LiveDocs;
int totalNumDocs = 0;
if (matchingVectorsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
for (int docNum = 0; docNum < maxDoc;)
{
if (!liveDocs.Get(docNum))
{
// skip deleted docs
++docNum;
continue;
}
// We can optimize this case (doing a bulk byte copy) since the field
// numbers are identical
int start = docNum, numDocs = 0;
do
{
docNum++;
numDocs++;
if (docNum >= maxDoc)
{
break;
}
if (!liveDocs.Get(docNum))
{
docNum++;
break;
}
} while (numDocs < MAX_RAW_MERGE_DOCS);
matchingVectorsReader.RawDocs(rawDocLengths, rawDocLengths2, start, numDocs);
AddRawDocuments(matchingVectorsReader, rawDocLengths, rawDocLengths2, numDocs);
totalNumDocs += numDocs;
mergeState.CheckAbort.Work(300 * numDocs);
}
}
else
{
for (int docNum = 0; docNum < maxDoc; docNum++)
{
if (!liveDocs.Get(docNum))
{
// skip deleted docs
continue;
}
// NOTE: it's very important to first assign to vectors then pass it to
// termVectorsWriter.addAllDocVectors; see LUCENE-1282
Fields vectors = reader.GetTermVectors(docNum);
AddAllDocVectors(vectors, mergeState);
totalNumDocs++;
mergeState.CheckAbort.Work(300);
}
}
return(totalNumDocs);
}
/// <summary>
/// Returns <c>true</c> if the given <paramref name="reader"/> is sorted by the specified <paramref name="sort"/>.
/// </summary>
public static bool IsSorted(AtomicReader reader, Sort sort)
{
if (reader is SegmentReader)
{
SegmentReader segReader = (SegmentReader)reader;
IDictionary<string, string> diagnostics = segReader.SegmentInfo.Info.Diagnostics;
var diagnosticsSort = diagnostics.ContainsKey(SORTER_ID_PROP) ? diagnostics[SORTER_ID_PROP] : null;
if (diagnostics != null && sort.ToString().Equals(diagnosticsSort))
{
return true;
}
}
return false;
}
private int CopyVectorsNoDeletions(MergeState mergeState, Lucene40TermVectorsReader matchingVectorsReader, AtomicReader reader, int[] rawDocLengths, int[] rawDocLengths2)
{
int maxDoc = reader.MaxDoc;
if (matchingVectorsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
int docCount = 0;
while (docCount < maxDoc)
{
int len = Math.Min(MAX_RAW_MERGE_DOCS, maxDoc - docCount);
matchingVectorsReader.RawDocs(rawDocLengths, rawDocLengths2, docCount, len);
AddRawDocuments(matchingVectorsReader, rawDocLengths, rawDocLengths2, len);
docCount += len;
mergeState.CheckAbort.Work(300 * len);
}
}
else
{
for (int docNum = 0; docNum < maxDoc; docNum++)
{
// NOTE: it's very important to first assign to vectors then pass it to
// termVectorsWriter.addAllDocVectors; see LUCENE-1282
Fields vectors = reader.GetTermVectors(docNum);
AddAllDocVectors(vectors, mergeState);
mergeState.CheckAbort.Work(300);
}
}
return(maxDoc);
}
/// <summary>
/// Return a sorted view of <code>reader</code> according to the order
/// defined by <code>sort</code>. If the reader is already sorted, this
/// method might return the reader as-is.
/// </summary>
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public static org.apache.lucene.index.AtomicReader wrap(org.apache.lucene.index.AtomicReader reader, org.apache.lucene.search.Sort sort) throws java.io.IOException
public static AtomicReader wrap(AtomicReader reader, Sort sort)
{
return wrap(reader, (new Sorter(sort)).sort(reader));
}
public override DocComparer GetComparer(AtomicReader reader, int docbase)
{
SortedDocValues values = FieldCache.DEFAULT.GetTermsIndex(reader, this.m_field);
return(new StringOrdDocComparer(values));
}
private IEnumerable <long?> GetMergeSortedSetOrdsEnumerable(AtomicReader[] readers, SortedSetDocValues[] dvs, OrdinalMap map)
{
int readerUpto = -1;
int docIDUpto = 0;
AtomicReader currentReader = null;
IBits currentLiveDocs = null;
var ords = new long[8];
int ordUpto = 0;
int ordLength = 0;
while (true)
{
if (readerUpto == readers.Length)
{
yield break;
}
if (ordUpto < ordLength)
{
var value = ords[ordUpto];
ordUpto++;
yield return(value);
continue;
}
if (currentReader == null || docIDUpto == currentReader.MaxDoc)
{
readerUpto++;
if (readerUpto < readers.Length)
{
currentReader = readers[readerUpto];
currentLiveDocs = currentReader.LiveDocs;
}
docIDUpto = 0;
continue;
}
if (currentLiveDocs == null || currentLiveDocs.Get(docIDUpto))
{
Debug.Assert(docIDUpto < currentReader.MaxDoc);
SortedSetDocValues dv = dvs[readerUpto];
dv.SetDocument(docIDUpto);
ordUpto = ordLength = 0;
long ord;
while ((ord = dv.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
if (ordLength == ords.Length)
{
ords = ArrayUtil.Grow(ords, ordLength + 1);
}
ords[ordLength] = map.GetGlobalOrd(readerUpto, ord);
ordLength++;
}
docIDUpto++;
continue;
}
docIDUpto++;
}
}
public abstract DocComparer GetComparer(AtomicReader reader, int docbase);
private static bool TermNotInReader(AtomicReader reader, Term term) // LUCENENET: CA1822: Mark members as static
{
// only called from assert
//System.out.println("TQ.termNotInReader reader=" + reader + " term=" + field + ":" + bytes.utf8ToString());
return(reader.DocFreq(term) == 0);
}
public override DocComparer GetComparer(AtomicReader reader, int docbase)
{
FieldCache.Doubles values = FieldCache.DEFAULT.GetDoubles(reader, this.m_field, true);
return(new DoubleDocComparer(values));
}
// only called from assert
private bool TermNotInReader(AtomicReader reader, Term term)
{
return(reader.DocFreq(term) == 0);
}
public override DocComparer GetComparer(AtomicReader reader, int docbase)
{
return(new RelevanceDocComparer());
}
public override AtomicReader Wrap(AtomicReader reader)
{
return new AssertingAtomicReader(reader);
}
public override DocComparer GetComparer(AtomicReader reader, int docbase)
{
return(new DocIdDocComparer(docbase));
}
private int CopyFieldsWithDeletions(MergeState mergeState, AtomicReader reader, Lucene40StoredFieldsReader matchingFieldsReader, int[] rawDocLengths)
{
int docCount = 0;
int maxDoc = reader.MaxDoc;
IBits liveDocs = reader.LiveDocs;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(liveDocs != null);
}
if (matchingFieldsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
for (int j = 0; j < maxDoc;)
{
if (!liveDocs.Get(j))
{
// skip deleted docs
++j;
continue;
}
// We can optimize this case (doing a bulk byte copy) since the field
// numbers are identical
int start = j, numDocs = 0;
do
{
j++;
numDocs++;
if (j >= maxDoc)
{
break;
}
if (!liveDocs.Get(j))
{
j++;
break;
}
} while (numDocs < MAX_RAW_MERGE_DOCS);
IndexInput stream = matchingFieldsReader.RawDocs(rawDocLengths, start, numDocs);
AddRawDocuments(stream, rawDocLengths, numDocs);
docCount += numDocs;
mergeState.CheckAbort.Work(300 * numDocs);
}
}
else
{
for (int j = 0; j < maxDoc; j++)
{
if (!liveDocs.Get(j))
{
// skip deleted docs
continue;
}
// TODO: this could be more efficient using
// FieldVisitor instead of loading/writing entire
// doc; ie we just have to renumber the field number
// on the fly?
// NOTE: it's very important to first assign to doc then pass it to
// fieldsWriter.addDocument; see LUCENE-1282
Document doc = reader.Document(j);
AddDocument(doc, mergeState.FieldInfos);
docCount++;
mergeState.CheckAbort.Work(300);
}
}
return(docCount);
}
public override DocComparer GetComparer(AtomicReader reader, int docbase)
{
FieldCache.Int32s values = FieldCache.DEFAULT.GetInt32s(reader, m_field, true);
return(new Int32DocComparer(values));
}
/// <summary>
/// alternate scorer skipTo(),skipTo(),next(),next(),skipTo(),skipTo(), etc
/// and ensure a hitcollector receives same docs and scores
/// </summary>
/// <param name = "similarity" >
/// LUCENENET specific
/// Removes dependency on <see cref="LuceneTestCase.ClassEnv.Similarity"/>
/// </param>
public static void CheckSkipTo(Query q, IndexSearcher s, Similarity similarity)
{
//System.out.println("Checking "+q);
IList<AtomicReaderContext> readerContextArray = s.TopReaderContext.Leaves;
if (s.CreateNormalizedWeight(q).ScoresDocsOutOfOrder()) // in this case order of skipTo() might differ from that of next().
{
return;
}
const int skip_op = 0;
const int next_op = 1;
int[][] orders = new int[][] { new int[] { next_op }, new int[] { skip_op }, new int[] { skip_op, next_op }, new int[] { next_op, skip_op }, new int[] { skip_op, skip_op, next_op, next_op }, new int[] { next_op, next_op, skip_op, skip_op }, new int[] { skip_op, skip_op, skip_op, next_op, next_op } };
for (int k = 0; k < orders.Length; k++)
{
int[] order = orders[k];
// System.out.print("Order:");for (int i = 0; i < order.Length; i++)
// System.out.print(order[i]==skip_op ? " skip()":" next()");
// System.out.println();
int[] opidx = new int[] { 0 };
int[] lastDoc = new int[] { -1 };
// FUTURE: ensure scorer.Doc()==-1
const float maxDiff = 1e-5f;
AtomicReader[] lastReader = new AtomicReader[] { null };
s.Search(q, new CollectorAnonymousInnerClassHelper(q, s, readerContextArray, skip_op, order, opidx, lastDoc, maxDiff, lastReader, similarity));
if (lastReader[0] != null)
{
// confirm that skipping beyond the last doc, on the
// previous reader, hits NO_MORE_DOCS
AtomicReader previousReader = lastReader[0];
IndexSearcher indexSearcher = LuceneTestCase.NewSearcher(previousReader, false, similarity);
indexSearcher.Similarity = s.Similarity;
Weight w = indexSearcher.CreateNormalizedWeight(q);
AtomicReaderContext ctx = (AtomicReaderContext)previousReader.Context;
Scorer scorer = w.Scorer(ctx, ((AtomicReader)ctx.Reader).LiveDocs);
if (scorer != null)
{
bool more = scorer.Advance(lastDoc[0] + 1) != DocIdSetIterator.NO_MORE_DOCS;
Assert.IsFalse(more, "query's last doc was " + lastDoc[0] + " but skipTo(" + (lastDoc[0] + 1) + ") got to " + scorer.DocID());
}
}
}
}
private static void AssertSorted(AtomicReader reader)
{
NumericDocValues ndv = reader.GetNumericDocValues("ndv");
for (int i = 1; i < reader.MaxDoc; ++i)
{
assertTrue("ndv(" + (i - 1) + ")=" + ndv.Get(i - 1) + ",ndv(" + i + ")=" + ndv.Get(i), ndv.Get(i - 1) <= ndv.Get(i));
}
}
/// <summary>
/// check that first skip on just created scorers always goes to the right doc</summary>
/// <param name = "similarity" >
/// LUCENENET specific
/// Removes dependency on <see cref="LuceneTestCase.ClassEnv.Similarity"/>
/// </param>
public static void CheckFirstSkipTo(Query q, IndexSearcher s, Similarity similarity)
{
//System.out.println("checkFirstSkipTo: "+q);
const float maxDiff = 1e-3f;
int[] lastDoc = new int[] { -1 };
AtomicReader[] lastReader = new AtomicReader[] { null };
IList<AtomicReaderContext> context = s.TopReaderContext.Leaves;
s.Search(q, new CollectorAnonymousInnerClassHelper2(q, s, maxDiff, lastDoc, lastReader, context, similarity));
if (lastReader[0] != null)
{
// confirm that skipping beyond the last doc, on the
// previous reader, hits NO_MORE_DOCS
AtomicReader previousReader = lastReader[0];
IndexSearcher indexSearcher = LuceneTestCase.NewSearcher(previousReader, similarity);
indexSearcher.Similarity = s.Similarity;
Weight w = indexSearcher.CreateNormalizedWeight(q);
Scorer scorer = w.Scorer((AtomicReaderContext)indexSearcher.TopReaderContext, previousReader.LiveDocs);
if (scorer != null)
{
bool more = scorer.Advance(lastDoc[0] + 1) != DocIdSetIterator.NO_MORE_DOCS;
Assert.IsFalse(more, "query's last doc was " + lastDoc[0] + " but skipTo(" + (lastDoc[0] + 1) + ") got to " + scorer.DocID());
}
}
}
public virtual void TestPayloadsPos0()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random(), dir, new MockPayloadAnalyzer(), Similarity, TimeZone);
Document doc = new Document();
doc.Add(new TextField("content", new StringReader("a a b c d e a f g h i j a b k k")));
writer.AddDocument(doc);
IndexReader readerFromWriter = writer.Reader;
AtomicReader r = SlowCompositeReaderWrapper.Wrap(readerFromWriter);
DocsAndPositionsEnum tp = r.GetTermPositionsEnum(new Term("content", "a"));
int count = 0;
Assert.IsTrue(tp.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
// "a" occurs 4 times
Assert.AreEqual(4, tp.Freq);
Assert.AreEqual(0, tp.NextPosition());
Assert.AreEqual(1, tp.NextPosition());
Assert.AreEqual(3, tp.NextPosition());
Assert.AreEqual(6, tp.NextPosition());
// only one doc has "a"
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, tp.NextDoc());
IndexSearcher @is = NewSearcher(readerFromWriter);
SpanTermQuery stq1 = new SpanTermQuery(new Term("content", "a"));
SpanTermQuery stq2 = new SpanTermQuery(new Term("content", "k"));
SpanQuery[] sqs = new SpanQuery[] { stq1, stq2 };
SpanNearQuery snq = new SpanNearQuery(sqs, 30, false);
count = 0;
bool sawZero = false;
if (VERBOSE)
{
Console.WriteLine("\ngetPayloadSpans test");
}
Search.Spans.Spans pspans = MultiSpansWrapper.Wrap(@is.TopReaderContext, snq);
while (pspans.Next())
{
if (VERBOSE)
{
Console.WriteLine("doc " + pspans.Doc + ": span " + pspans.Start + " to " + pspans.End);
}
var payloads = pspans.GetPayload();
sawZero |= pspans.Start == 0;
foreach (var bytes in payloads)
{
count++;
if (VERBOSE)
{
Console.WriteLine(" payload: " + Encoding.UTF8.GetString(bytes));
}
}
}
Assert.IsTrue(sawZero);
Assert.AreEqual(5, count);
// System.out.println("\ngetSpans test");
Search.Spans.Spans spans = MultiSpansWrapper.Wrap(@is.TopReaderContext, snq);
count = 0;
sawZero = false;
while (spans.Next())
{
count++;
sawZero |= spans.Start == 0;
// System.out.println(spans.Doc() + " - " + spans.Start() + " - " +
// spans.End());
}
Assert.AreEqual(4, count);
Assert.IsTrue(sawZero);
// System.out.println("\nPayloadSpanUtil test");
sawZero = false;
PayloadSpanUtil psu = new PayloadSpanUtil(@is.TopReaderContext);
var pls = psu.GetPayloadsForQuery(snq);
count = pls.Count;
foreach (var bytes in pls)
{
string s = Encoding.UTF8.GetString(bytes);
//System.out.println(s);
sawZero |= s.Equals("pos: 0");
}
Assert.AreEqual(5, count);
Assert.IsTrue(sawZero);
writer.Dispose();
@is.IndexReader.Dispose();
dir.Dispose();
}
private IEnumerable<long?> GetMergeSortedFieldDocToOrdEnumerable(AtomicReader[] readers, SortedDocValues[] dvs, OrdinalMap map)
{
int readerUpTo = -1;
int docIDUpTo = 0;
AtomicReader currentReader = null;
Bits currentLiveDocs = null;
while (true)
{
if (readerUpTo == readers.Length)
{
yield break;
}
if (currentReader == null || docIDUpTo == currentReader.MaxDoc)
{
readerUpTo++;
if (readerUpTo < readers.Length)
{
currentReader = readers[readerUpTo];
currentLiveDocs = currentReader.LiveDocs;
}
docIDUpTo = 0;
continue;
}
if (currentLiveDocs == null || currentLiveDocs.Get(docIDUpTo))
{
int segOrd = dvs[readerUpTo].GetOrd(docIDUpTo);
docIDUpTo++;
yield return segOrd == -1 ? -1 : map.GetGlobalOrd(readerUpTo, segOrd);
continue;
}
docIDUpTo++;
}
}
private void DuellReaders(CompositeReader other, AtomicReader memIndexReader)
{
AtomicReader competitor = SlowCompositeReaderWrapper.Wrap(other);
Fields memFields = memIndexReader.Fields;
foreach (string field in competitor.Fields)
{
Terms memTerms = memFields.GetTerms(field);
Terms iwTerms = memIndexReader.GetTerms(field);
if (iwTerms == null)
{
assertNull(memTerms);
}
else
{
NumericDocValues normValues = competitor.GetNormValues(field);
NumericDocValues memNormValues = memIndexReader.GetNormValues(field);
if (normValues != null)
{
// mem idx always computes norms on the fly
assertNotNull(memNormValues);
assertEquals(normValues.Get(0), memNormValues.Get(0));
}
assertNotNull(memTerms);
assertEquals(iwTerms.DocCount, memTerms.DocCount);
assertEquals(iwTerms.SumDocFreq, memTerms.SumDocFreq);
assertEquals(iwTerms.SumTotalTermFreq, memTerms.SumTotalTermFreq);
TermsEnum iwTermsIter = iwTerms.GetIterator(null);
TermsEnum memTermsIter = memTerms.GetIterator(null);
if (iwTerms.HasPositions)
{
bool offsets = iwTerms.HasOffsets && memTerms.HasOffsets;
while (iwTermsIter.Next() != null)
{
assertNotNull(memTermsIter.Next());
assertEquals(iwTermsIter.Term, memTermsIter.Term);
DocsAndPositionsEnum iwDocsAndPos = iwTermsIter.DocsAndPositions(null, null);
DocsAndPositionsEnum memDocsAndPos = memTermsIter.DocsAndPositions(null, null);
while (iwDocsAndPos.NextDoc() != DocsAndPositionsEnum.NO_MORE_DOCS)
{
assertEquals(iwDocsAndPos.DocID, memDocsAndPos.NextDoc());
assertEquals(iwDocsAndPos.Freq, memDocsAndPos.Freq);
for (int i = 0; i < iwDocsAndPos.Freq; i++)
{
assertEquals("term: " + iwTermsIter.Term.Utf8ToString(), iwDocsAndPos.NextPosition(), memDocsAndPos.NextPosition());
if (offsets)
{
assertEquals(iwDocsAndPos.StartOffset, memDocsAndPos.StartOffset);
assertEquals(iwDocsAndPos.EndOffset, memDocsAndPos.EndOffset);
}
}
}
}
}
else
{
while (iwTermsIter.Next() != null)
{
assertEquals(iwTermsIter.Term, memTermsIter.Term);
DocsEnum iwDocsAndPos = iwTermsIter.Docs(null, null);
DocsEnum memDocsAndPos = memTermsIter.Docs(null, null);
while (iwDocsAndPos.NextDoc() != DocsAndPositionsEnum.NO_MORE_DOCS)
{
assertEquals(iwDocsAndPos.DocID, memDocsAndPos.NextDoc());
assertEquals(iwDocsAndPos.Freq, memDocsAndPos.Freq);
}
}
}
}
}
}
private IEnumerable<long?> GetMergeSortedSetOrdsEnumerable(AtomicReader[] readers, SortedSetDocValues[] dvs, OrdinalMap map)
{
int readerUpto = -1;
int docIDUpto = 0;
AtomicReader currentReader = null;
Bits currentLiveDocs = null;
var ords = new long[8];
int ordUpto = 0;
int ordLength = 0;
while (true)
{
if (readerUpto == readers.Length)
{
yield break;
}
if (ordUpto < ordLength)
{
var value = ords[ordUpto];
ordUpto++;
yield return value;
continue;
}
if (currentReader == null || docIDUpto == currentReader.MaxDoc)
{
readerUpto++;
if (readerUpto < readers.Length)
{
currentReader = readers[readerUpto];
currentLiveDocs = currentReader.LiveDocs;
}
docIDUpto = 0;
continue;
}
if (currentLiveDocs == null || currentLiveDocs.Get(docIDUpto))
{
Debug.Assert(docIDUpto < currentReader.MaxDoc);
SortedSetDocValues dv = dvs[readerUpto];
dv.Document = docIDUpto;
ordUpto = ordLength = 0;
long ord;
while ((ord = dv.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
if (ordLength == ords.Length)
{
ords = ArrayUtil.Grow(ords, ordLength + 1);
}
ords[ordLength] = map.GetGlobalOrd(readerUpto, ord);
ordLength++;
}
docIDUpto++;
continue;
}
docIDUpto++;
}
}
public virtual void TestNonIndexedFields()
{
Directory dir = NewDirectory();
RandomIndexWriter iw = new RandomIndexWriter(Random(), dir, Similarity, TimeZone);
Document doc = new Document();
doc.Add(new StoredField("bogusbytes", "bogus"));
doc.Add(new StoredField("bogusshorts", "bogus"));
doc.Add(new StoredField("bogusints", "bogus"));
doc.Add(new StoredField("boguslongs", "bogus"));
doc.Add(new StoredField("bogusfloats", "bogus"));
doc.Add(new StoredField("bogusdoubles", "bogus"));
doc.Add(new StoredField("bogusterms", "bogus"));
doc.Add(new StoredField("bogustermsindex", "bogus"));
doc.Add(new StoredField("bogusmultivalued", "bogus"));
doc.Add(new StoredField("bogusbits", "bogus"));
iw.AddDocument(doc);
DirectoryReader ir = iw.Reader;
iw.Dispose();
AtomicReader ar = GetOnlySegmentReader(ir);
IFieldCache cache = FieldCache.DEFAULT;
cache.PurgeAllCaches();
Assert.AreEqual(0, cache.CacheEntries.Length);
Bytes bytes = cache.GetBytes(ar, "bogusbytes", true);
Assert.AreEqual(0, bytes.Get(0));
Shorts shorts = cache.GetShorts(ar, "bogusshorts", true);
Assert.AreEqual(0, shorts.Get(0));
Ints ints = cache.GetInts(ar, "bogusints", true);
Assert.AreEqual(0, ints.Get(0));
Longs longs = cache.GetLongs(ar, "boguslongs", true);
Assert.AreEqual(0, longs.Get(0));
Floats floats = cache.GetFloats(ar, "bogusfloats", true);
Assert.AreEqual(0, floats.Get(0), 0.0f);
Doubles doubles = cache.GetDoubles(ar, "bogusdoubles", true);
Assert.AreEqual(0, doubles.Get(0), 0.0D);
BytesRef scratch = new BytesRef();
BinaryDocValues binaries = cache.GetTerms(ar, "bogusterms", true);
binaries.Get(0, scratch);
Assert.AreEqual(0, scratch.Length);
SortedDocValues sorted = cache.GetTermsIndex(ar, "bogustermsindex");
Assert.AreEqual(-1, sorted.GetOrd(0));
sorted.Get(0, scratch);
Assert.AreEqual(0, scratch.Length);
SortedSetDocValues sortedSet = cache.GetDocTermOrds(ar, "bogusmultivalued");
sortedSet.Document = 0;
Assert.AreEqual(SortedSetDocValues.NO_MORE_ORDS, sortedSet.NextOrd());
Bits bits = cache.GetDocsWithField(ar, "bogusbits");
Assert.IsFalse(bits.Get(0));
// check that we cached nothing
Assert.AreEqual(0, cache.CacheEntries.Length);
ir.Dispose();
dir.Dispose();
}
/// <summary>
/// Returns {@code true} if the given {@code reader} is sorted by the specified {@code sort}. </summary>
public static bool isSorted(AtomicReader reader, Sort sort)
{
if (reader is SegmentReader)
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.index.SegmentReader segReader = (org.apache.lucene.index.SegmentReader) reader;
SegmentReader segReader = (SegmentReader) reader;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final java.util.Map<String, String> diagnostics = segReader.getSegmentInfo().info.getDiagnostics();
IDictionary<string, string> diagnostics = segReader.SegmentInfo.info.Diagnostics;
if (diagnostics != null && sort.ToString().Equals(diagnostics[SORTER_ID_PROP]))
{
return true;
}
}
return false;
}
public virtual void TestDocValuesIntegration()
{
AssumeTrue("3.x does not support docvalues", DefaultCodecSupportsDocValues);
Directory dir = NewDirectory();
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, null);
RandomIndexWriter iw = new RandomIndexWriter(Random, dir, iwc);
Document doc = new Document();
doc.Add(new BinaryDocValuesField("binary", new BytesRef("binary value")));
doc.Add(new SortedDocValuesField("sorted", new BytesRef("sorted value")));
doc.Add(new NumericDocValuesField("numeric", 42));
if (DefaultCodecSupportsSortedSet)
{
doc.Add(new SortedSetDocValuesField("sortedset", new BytesRef("sortedset value1")));
doc.Add(new SortedSetDocValuesField("sortedset", new BytesRef("sortedset value2")));
}
iw.AddDocument(doc);
DirectoryReader ir = iw.GetReader();
iw.Dispose();
AtomicReader ar = GetOnlySegmentReader(ir);
BytesRef scratch = new BytesRef();
// Binary type: can be retrieved via getTerms()
try
{
FieldCache.DEFAULT.GetInt32s(ar, "binary", false);
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
BinaryDocValues binary = FieldCache.DEFAULT.GetTerms(ar, "binary", true);
binary.Get(0, scratch);
Assert.AreEqual("binary value", scratch.Utf8ToString());
try
{
FieldCache.DEFAULT.GetTermsIndex(ar, "binary");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
FieldCache.DEFAULT.GetDocTermOrds(ar, "binary");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
new DocTermOrds(ar, null, "binary");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
IBits bits = FieldCache.DEFAULT.GetDocsWithField(ar, "binary");
Assert.IsTrue(bits.Get(0));
// Sorted type: can be retrieved via getTerms(), getTermsIndex(), getDocTermOrds()
try
{
FieldCache.DEFAULT.GetInt32s(ar, "sorted", false);
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
new DocTermOrds(ar, null, "sorted");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
binary = FieldCache.DEFAULT.GetTerms(ar, "sorted", true);
binary.Get(0, scratch);
Assert.AreEqual("sorted value", scratch.Utf8ToString());
SortedDocValues sorted = FieldCache.DEFAULT.GetTermsIndex(ar, "sorted");
Assert.AreEqual(0, sorted.GetOrd(0));
Assert.AreEqual(1, sorted.ValueCount);
sorted.Get(0, scratch);
Assert.AreEqual("sorted value", scratch.Utf8ToString());
SortedSetDocValues sortedSet = FieldCache.DEFAULT.GetDocTermOrds(ar, "sorted");
sortedSet.SetDocument(0);
Assert.AreEqual(0, sortedSet.NextOrd());
Assert.AreEqual(SortedSetDocValues.NO_MORE_ORDS, sortedSet.NextOrd());
Assert.AreEqual(1, sortedSet.ValueCount);
bits = FieldCache.DEFAULT.GetDocsWithField(ar, "sorted");
Assert.IsTrue(bits.Get(0));
// Numeric type: can be retrieved via getInts() and so on
Int32s numeric = FieldCache.DEFAULT.GetInt32s(ar, "numeric", false);
Assert.AreEqual(42, numeric.Get(0));
try
{
FieldCache.DEFAULT.GetTerms(ar, "numeric", true);
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
FieldCache.DEFAULT.GetTermsIndex(ar, "numeric");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
FieldCache.DEFAULT.GetDocTermOrds(ar, "numeric");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
new DocTermOrds(ar, null, "numeric");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
bits = FieldCache.DEFAULT.GetDocsWithField(ar, "numeric");
Assert.IsTrue(bits.Get(0));
// SortedSet type: can be retrieved via getDocTermOrds()
if (DefaultCodecSupportsSortedSet)
{
try
{
FieldCache.DEFAULT.GetInt32s(ar, "sortedset", false);
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
FieldCache.DEFAULT.GetTerms(ar, "sortedset", true);
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
FieldCache.DEFAULT.GetTermsIndex(ar, "sortedset");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
try
{
new DocTermOrds(ar, null, "sortedset");
Assert.Fail();
}
#pragma warning disable 168
catch (InvalidOperationException expected)
#pragma warning restore 168
{
}
sortedSet = FieldCache.DEFAULT.GetDocTermOrds(ar, "sortedset");
sortedSet.SetDocument(0);
Assert.AreEqual(0, sortedSet.NextOrd());
Assert.AreEqual(1, sortedSet.NextOrd());
Assert.AreEqual(SortedSetDocValues.NO_MORE_ORDS, sortedSet.NextOrd());
Assert.AreEqual(2, sortedSet.ValueCount);
bits = FieldCache.DEFAULT.GetDocsWithField(ar, "sortedset");
Assert.IsTrue(bits.Get(0));
}
ir.Dispose();
dir.Dispose();
}
internal bool TermNotInReader(AtomicReader reader, Term term)
{
// only called from assert
//System.out.println("TQ.termNotInReader reader=" + reader + " term=" + field + ":" + bytes.utf8ToString());
return reader.DocFreq(term) == 0;
}
public virtual void TestNonIndexedFields()
{
Directory dir = NewDirectory();
RandomIndexWriter iw = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
Document doc = new Document();
doc.Add(new StoredField("bogusbytes", "bogus"));
doc.Add(new StoredField("bogusshorts", "bogus"));
doc.Add(new StoredField("bogusints", "bogus"));
doc.Add(new StoredField("boguslongs", "bogus"));
doc.Add(new StoredField("bogusfloats", "bogus"));
doc.Add(new StoredField("bogusdoubles", "bogus"));
doc.Add(new StoredField("bogusterms", "bogus"));
doc.Add(new StoredField("bogustermsindex", "bogus"));
doc.Add(new StoredField("bogusmultivalued", "bogus"));
doc.Add(new StoredField("bogusbits", "bogus"));
iw.AddDocument(doc);
DirectoryReader ir = iw.GetReader();
iw.Dispose();
AtomicReader ar = GetOnlySegmentReader(ir);
IFieldCache cache = FieldCache.DEFAULT;
cache.PurgeAllCaches();
Assert.AreEqual(0, cache.GetCacheEntries().Length);
#pragma warning disable 612, 618
Bytes bytes = cache.GetBytes(ar, "bogusbytes", true);
Assert.AreEqual((byte)0, bytes.Get(0));
Int16s shorts = cache.GetInt16s(ar, "bogusshorts", true);
Assert.AreEqual(0, shorts.Get(0));
#pragma warning restore 612, 618
Int32s ints = cache.GetInt32s(ar, "bogusints", true);
Assert.AreEqual(0, ints.Get(0));
Int64s longs = cache.GetInt64s(ar, "boguslongs", true);
Assert.AreEqual(0, longs.Get(0));
Singles floats = cache.GetSingles(ar, "bogusfloats", true);
Assert.AreEqual(0, floats.Get(0), 0.0f);
Doubles doubles = cache.GetDoubles(ar, "bogusdoubles", true);
Assert.AreEqual(0, doubles.Get(0), 0.0D);
BytesRef scratch = new BytesRef();
BinaryDocValues binaries = cache.GetTerms(ar, "bogusterms", true);
binaries.Get(0, scratch);
Assert.AreEqual(0, scratch.Length);
SortedDocValues sorted = cache.GetTermsIndex(ar, "bogustermsindex");
Assert.AreEqual(-1, sorted.GetOrd(0));
sorted.Get(0, scratch);
Assert.AreEqual(0, scratch.Length);
SortedSetDocValues sortedSet = cache.GetDocTermOrds(ar, "bogusmultivalued");
sortedSet.SetDocument(0);
Assert.AreEqual(SortedSetDocValues.NO_MORE_ORDS, sortedSet.NextOrd());
IBits bits = cache.GetDocsWithField(ar, "bogusbits");
Assert.IsFalse(bits.Get(0));
// check that we cached nothing
Assert.AreEqual(0, cache.GetCacheEntries().Length);
ir.Dispose();
dir.Dispose();
}
/// <summary>
/// Expert: same as <seealso cref="#wrap(AtomicReader, Sort)"/> but operates directly on a <seealso cref="Sorter.DocMap"/>. </summary>
internal static AtomicReader wrap(AtomicReader reader, Sorter.DocMap docMap)
{
if (docMap == null)
{
// the reader is already sorter
return reader;
}
if (reader.maxDoc() != docMap.size())
{
throw new System.ArgumentException("reader.maxDoc() should be equal to docMap.size(), got" + reader.maxDoc() + " != " + docMap.size());
}
Debug.Assert(Sorter.isConsistent(docMap));
return new SortingAtomicReader(reader, docMap);
}
public override AtomicReader Wrap(AtomicReader reader)
{
return(new AssertingAtomicReader(reader));
}
protected override DocIdSet CacheImpl(DocIdSetIterator iterator, AtomicReader reader)
{
FixedBitSet cached = new FixedBitSet(reader.MaxDoc);
cached.Or(iterator);
return cached;
}
private readonly Sorter.DocMap docMap; // pkg-protected to avoid synthetic accessor methods
private SortingAtomicReader(AtomicReader @in, Sorter.DocMap docMap) : base(@in)
{
this.docMap = docMap;
}