private FacetResult GetDim(string dim, OrdRange ordRange, int topN)
{
TopOrdAndInt32Queue q = null;
int bottomCount = 0;
int dimCount = 0;
int childCount = 0;
TopOrdAndInt32Queue.OrdAndValue reuse = null;
//System.out.println("getDim : " + ordRange.start + " - " + ordRange.end);
for (int ord = ordRange.Start; ord <= ordRange.End; ord++)
{
//System.out.println(" ord=" + ord + " count=" + counts[ord]);
if (counts[ord] > 0)
{
dimCount += counts[ord];
childCount++;
if (counts[ord] > bottomCount)
{
if (reuse == null)
{
reuse = new TopOrdAndInt32Queue.OrdAndValue();
}
reuse.Ord = ord;
reuse.Value = counts[ord];
if (q == null)
{
// Lazy init, so we don't create this for the
// sparse case unnecessarily
q = new TopOrdAndInt32Queue(topN);
}
reuse = q.InsertWithOverflow(reuse);
if (q.Count == topN)
{
bottomCount = q.Top.Value;
}
}
}
}
if (q == null)
{
return(null);
}
LabelAndValue[] labelValues = new LabelAndValue[q.Count];
for (int i = labelValues.Length - 1; i >= 0; i--)
{
TopOrdAndInt32Queue.OrdAndValue ordAndValue = q.Pop();
var term = new BytesRef();
dv.LookupOrd(ordAndValue.Ord, term);
string[] parts = FacetsConfig.StringToPath(term.Utf8ToString());
labelValues[i] = new LabelAndValue(parts[1], ordAndValue.Value);
}
return(new FacetResult(dim, Arrays.Empty <string>(), dimCount, labelValues, childCount));
}
public virtual void TestSortedSetDocValuesField()
{
AssumeTrue("default codec does not support SORTED_SET", DefaultCodecSupportsSortedSet);
SortedSetDocValues dv = reader.GetSortedSetDocValues(SORTED_SET_DV_FIELD);
int maxDoc = reader.MaxDoc;
BytesRef bytes = new BytesRef();
for (int i = 0; i < maxDoc; i++)
{
dv.SetDocument(i);
dv.LookupOrd(dv.NextOrd(), bytes);
int value = sortedValues[i];
assertEquals("incorrect sorted-set DocValues for doc " + i, value.toString(), bytes.Utf8ToString());
dv.LookupOrd(dv.NextOrd(), bytes);
assertEquals("incorrect sorted-set DocValues for doc " + i, (value + 1).ToString(), bytes.Utf8ToString());
assertEquals(SortedSetDocValues.NO_MORE_ORDS, dv.NextOrd());
}
}
/// <summary>
/// Creates this, pulling doc values from the specified
/// field.
/// </summary>
public DefaultSortedSetDocValuesReaderState(IndexReader reader, string field = FacetsConfig.DEFAULT_INDEX_FIELD_NAME)
{
this.field = field;
this.origReader = reader;
// We need this to create thread-safe MultiSortedSetDV
// per collector:
topReader = SlowCompositeReaderWrapper.Wrap(reader);
SortedSetDocValues dv = topReader.GetSortedSetDocValues(field);
if (dv is null)
{
throw new ArgumentException("field \"" + field + "\" was not indexed with SortedSetDocValues");
}
if (dv.ValueCount > int.MaxValue)
{
throw new ArgumentException("can only handle valueCount < System.Int32.MaxValue; got " + dv.ValueCount);
}
valueCount = (int)dv.ValueCount;
// TODO: we can make this more efficient if eg we can be
// "involved" when IOrdinalMap is being created? Ie see
// each term/ord it's assigning as it goes...
string lastDim = null;
int startOrd = -1;
BytesRef spare = new BytesRef();
// TODO: this approach can work for full hierarchy?;
// TaxoReader can't do this since ords are not in
// "sorted order" ... but we should generalize this to
// support arbitrary hierarchy:
for (int ord = 0; ord < valueCount; ord++)
{
dv.LookupOrd(ord, spare);
string[] components = FacetsConfig.StringToPath(spare.Utf8ToString());
if (components.Length != 2)
{
throw new ArgumentException("this class can only handle 2 level hierarchy (dim/value); got: " + Arrays.ToString(components) + " " + spare.Utf8ToString());
}
if (!components[0].Equals(lastDim, StringComparison.Ordinal))
{
if (lastDim != null)
{
prefixToOrdRange[lastDim] = new OrdRange(startOrd, ord - 1);
}
startOrd = ord;
lastDim = components[0];
}
}
if (lastDim != null)
{
prefixToOrdRange[lastDim] = new OrdRange(startOrd, valueCount - 1);
}
}
public override void Collect(int doc)
{
_docTermOrds.SetDocument(doc);
long ord;
while ((ord = _docTermOrds.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
_docTermOrds.LookupOrd(ord, _scratch);
_collectorTerms.Add(_scratch);
}
}
private FacetResult GetDim(string dim, OrdRange ordRange, int topN)
{
TopOrdAndInt32Queue q = null;
int bottomCount = 0;
int dimCount = 0;
int childCount = 0;
//System.out.println("getDim : " + ordRange.start + " - " + ordRange.end);
for (int ord = ordRange.Start; ord <= ordRange.End; ord++)
{
//System.out.println(" ord=" + ord + " count=" + counts[ord]);
if (counts[ord] > 0)
{
dimCount += counts[ord];
childCount++;
if (counts[ord] > bottomCount)
{
if (q == null)
{
// Lazy init, so we don't create this for the
// sparse case unnecessarily
q = new TopOrdAndInt32Queue(topN);
}
// LUCENENET specific - use struct instead of reusing class instance for better performance
q.Insert(new OrdAndValue <int>(ord, counts[ord]));
if (q.Count == topN)
{
bottomCount = q.Top.Value;
}
}
}
}
if (q == null)
{
return(null);
}
var scratch = new BytesRef();
LabelAndValue[] labelValues = new LabelAndValue[q.Count];
for (int i = labelValues.Length - 1; i >= 0; i--)
{
var ordAndValue = q.Pop();
dv.LookupOrd(ordAndValue.Ord, scratch);
string[] parts = FacetsConfig.StringToPath(scratch.Utf8ToString());
labelValues[i] = new LabelAndValue(parts[1], ordAndValue.Value);
}
return(new FacetResult(dim, Arrays.Empty <string>(), dimCount, labelValues, childCount));
}
public virtual void Collect(int doc)
{
docTermOrds.SetDocument(doc);
long ord;
while ((ord = docTermOrds.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
docTermOrds.LookupOrd(ord, joinValue);
if (!joinValueToJoinScores.TryGetValue(joinValue, out JoinScore joinScore) || joinScore == null)
{
joinValueToJoinScores[BytesRef.DeepCopyOf(joinValue)] = joinScore = new JoinScore();
}
joinScore.AddScore(scorer.GetScore());
}
}
public virtual void Collect(int doc)
{
docTermOrds.SetDocument(doc);
long ord;
while ((ord = docTermOrds.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
docTermOrds.LookupOrd(ord, joinValue);
var joinScore = JoinValueToJoinScores.ContainsKey(joinValue) ? JoinValueToJoinScores[joinValue] : null;
if (joinScore == null)
{
JoinValueToJoinScores[BytesRef.DeepCopyOf(joinValue)] = joinScore = new JoinScore();
}
joinScore.AddScore(scorer.GetScore());
}
}
public override void Collect(int doc)
{
docTermOrds.Document = doc;
long ord;
while ((ord = docTermOrds.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
docTermOrds.LookupOrd(ord, joinValue);
var joinScore = JoinValueToJoinScores[joinValue];
if (joinScore == null)
{
JoinValueToJoinScores[BytesRef.DeepCopyOf(joinValue)] = joinScore = new JoinScore();
}
joinScore.AddScore(scorer.Score());
}
}
public virtual void Collect(int doc)
{
docTermOrds.SetDocument(doc);
long ord;
while ((ord = docTermOrds.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
docTermOrds.LookupOrd(ord, scratch);
if (!joinValueToJoinScores.TryGetValue(scratch, out JoinScore joinScore) || joinScore == null)
{
continue;
}
int basedDoc = docBase + doc;
// First encountered join value determines the score.
// Something to keep in mind for many-to-many relations.
if (!docToJoinScore.ContainsKey(basedDoc))
{
docToJoinScore[basedDoc] = joinScore;
}
}
}
public override void Collect(int doc)
{
docTermOrds.Document = doc;
long ord;
while ((ord = docTermOrds.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
docTermOrds.LookupOrd(ord, scratch);
JoinScore joinScore = _joinValueToJoinScores.ContainsKey(scratch) ? _joinValueToJoinScores[scratch] : null;
if (joinScore == null)
{
continue;
}
int basedDoc = docBase + doc;
// First encountered join value determines the score.
// Something to keep in mind for many-to-many relations.
if (!_docToJoinScore.ContainsKey(basedDoc))
{
_docToJoinScore[basedDoc] = joinScore;
}
}
}
public override void LookupOrd(long ord, BytesRef result)
{
@in.LookupOrd(ord, result);
}
/// <summary>
/// Create the results based on the search hits.
/// Can be overridden by subclass to add particular behavior (e.g. weight transformation) </summary>
/// <exception cref="System.IO.IOException"> If there are problems reading fields from the underlying Lucene index. </exception>
protected internal virtual IList <LookupResult> CreateResults(IndexSearcher searcher, TopFieldDocs hits, int num, string charSequence, bool doHighlight, IEnumerable <string> matchedTokens, string prefixToken)
{
BinaryDocValues textDV = MultiDocValues.GetBinaryValues(searcher.IndexReader, TEXT_FIELD_NAME);
// This will just be null if app didn't pass payloads to build():
// TODO: maybe just stored fields? they compress...
BinaryDocValues payloadsDV = MultiDocValues.GetBinaryValues(searcher.IndexReader, "payloads");
IList <AtomicReaderContext> leaves = searcher.IndexReader.Leaves;
List <LookupResult> results = new List <LookupResult>();
BytesRef scratch = new BytesRef();
for (int i = 0; i < hits.ScoreDocs.Length; i++)
{
FieldDoc fd = (FieldDoc)hits.ScoreDocs[i];
textDV.Get(fd.Doc, scratch);
string text = scratch.Utf8ToString();
long score = (long)fd.Fields[0];
BytesRef payload;
if (payloadsDV != null)
{
payload = new BytesRef();
payloadsDV.Get(fd.Doc, payload);
}
else
{
payload = null;
}
// Must look up sorted-set by segment:
int segment = ReaderUtil.SubIndex(fd.Doc, leaves);
SortedSetDocValues contextsDV = leaves[segment].AtomicReader.GetSortedSetDocValues(CONTEXTS_FIELD_NAME);
HashSet <BytesRef> contexts;
if (contextsDV != null)
{
contexts = new HashSet <BytesRef>();
contextsDV.SetDocument(fd.Doc - leaves[segment].DocBase);
long ord;
while ((ord = contextsDV.NextOrd()) != SortedSetDocValues.NO_MORE_ORDS)
{
BytesRef context = new BytesRef();
contextsDV.LookupOrd(ord, context);
contexts.Add(context);
}
}
else
{
contexts = null;
}
LookupResult result;
if (doHighlight)
{
object highlightKey = Highlight(text, matchedTokens, prefixToken);
result = new LookupResult(highlightKey.ToString(), highlightKey, score, payload, contexts);
}
else
{
result = new LookupResult(text, score, payload, contexts);
}
results.Add(result);
}
return(results);
}
public virtual void Test()
{
#pragma warning disable 612, 618
IFieldCache cache = FieldCache.DEFAULT;
FieldCache.Doubles doubles = cache.GetDoubles(Reader, "theDouble", Random.NextBoolean());
Assert.AreSame(doubles, cache.GetDoubles(Reader, "theDouble", Random.NextBoolean()), "Second request to cache return same array");
Assert.AreSame(doubles, cache.GetDoubles(Reader, "theDouble", FieldCache.DEFAULT_DOUBLE_PARSER, Random.NextBoolean()), "Second request with explicit parser return same array");
for (int i = 0; i < NUM_DOCS; i++)
{
Assert.IsTrue(doubles.Get(i) == (double.MaxValue - i), doubles.Get(i) + " does not equal: " + (double.MaxValue - i));
}
FieldCache.Int64s longs = cache.GetInt64s(Reader, "theLong", Random.NextBoolean());
Assert.AreSame(longs, cache.GetInt64s(Reader, "theLong", Random.NextBoolean()), "Second request to cache return same array");
Assert.AreSame(longs, cache.GetInt64s(Reader, "theLong", FieldCache.DEFAULT_INT64_PARSER, Random.NextBoolean()), "Second request with explicit parser return same array");
for (int i = 0; i < NUM_DOCS; i++)
{
Assert.IsTrue(longs.Get(i) == (long.MaxValue - i), longs.Get(i) + " does not equal: " + (long.MaxValue - i) + " i=" + i);
}
FieldCache.Bytes bytes = cache.GetBytes(Reader, "theByte", Random.NextBoolean());
Assert.AreSame(bytes, cache.GetBytes(Reader, "theByte", Random.NextBoolean()), "Second request to cache return same array");
Assert.AreSame(bytes, cache.GetBytes(Reader, "theByte", FieldCache.DEFAULT_BYTE_PARSER, Random.NextBoolean()), "Second request with explicit parser return same array");
for (int i = 0; i < NUM_DOCS; i++)
{
Assert.IsTrue((sbyte)bytes.Get(i) == (sbyte)(sbyte.MaxValue - i), (sbyte)bytes.Get(i) + " does not equal: " + (sbyte.MaxValue - i));
}
FieldCache.Int16s shorts = cache.GetInt16s(Reader, "theShort", Random.NextBoolean());
Assert.AreSame(shorts, cache.GetInt16s(Reader, "theShort", Random.NextBoolean()), "Second request to cache return same array");
Assert.AreSame(shorts, cache.GetInt16s(Reader, "theShort", FieldCache.DEFAULT_INT16_PARSER, Random.NextBoolean()), "Second request with explicit parser return same array");
for (int i = 0; i < NUM_DOCS; i++)
{
Assert.IsTrue(shorts.Get(i) == (short)(short.MaxValue - i), shorts.Get(i) + " does not equal: " + (short.MaxValue - i));
}
FieldCache.Int32s ints = cache.GetInt32s(Reader, "theInt", Random.NextBoolean());
Assert.AreSame(ints, cache.GetInt32s(Reader, "theInt", Random.NextBoolean()), "Second request to cache return same array");
Assert.AreSame(ints, cache.GetInt32s(Reader, "theInt", FieldCache.DEFAULT_INT32_PARSER, Random.NextBoolean()), "Second request with explicit parser return same array");
for (int i = 0; i < NUM_DOCS; i++)
{
Assert.IsTrue(ints.Get(i) == (int.MaxValue - i), ints.Get(i) + " does not equal: " + (int.MaxValue - i));
}
FieldCache.Singles floats = cache.GetSingles(Reader, "theFloat", Random.NextBoolean());
Assert.AreSame(floats, cache.GetSingles(Reader, "theFloat", Random.NextBoolean()), "Second request to cache return same array");
Assert.AreSame(floats, cache.GetSingles(Reader, "theFloat", FieldCache.DEFAULT_SINGLE_PARSER, Random.NextBoolean()), "Second request with explicit parser return same array");
for (int i = 0; i < NUM_DOCS; i++)
{
Assert.IsTrue(floats.Get(i) == (float.MaxValue - i), floats.Get(i) + " does not equal: " + (float.MaxValue - i));
}
#pragma warning restore 612, 618
IBits docsWithField = cache.GetDocsWithField(Reader, "theLong");
Assert.AreSame(docsWithField, cache.GetDocsWithField(Reader, "theLong"), "Second request to cache return same array");
Assert.IsTrue(docsWithField is Bits.MatchAllBits, "docsWithField(theLong) must be class Bits.MatchAllBits");
Assert.IsTrue(docsWithField.Length == NUM_DOCS, "docsWithField(theLong) Size: " + docsWithField.Length + " is not: " + NUM_DOCS);
for (int i = 0; i < docsWithField.Length; i++)
{
Assert.IsTrue(docsWithField.Get(i));
}
docsWithField = cache.GetDocsWithField(Reader, "sparse");
Assert.AreSame(docsWithField, cache.GetDocsWithField(Reader, "sparse"), "Second request to cache return same array");
Assert.IsFalse(docsWithField is Bits.MatchAllBits, "docsWithField(sparse) must not be class Bits.MatchAllBits");
Assert.IsTrue(docsWithField.Length == NUM_DOCS, "docsWithField(sparse) Size: " + docsWithField.Length + " is not: " + NUM_DOCS);
for (int i = 0; i < docsWithField.Length; i++)
{
Assert.AreEqual(i % 2 == 0, docsWithField.Get(i));
}
// getTermsIndex
SortedDocValues termsIndex = cache.GetTermsIndex(Reader, "theRandomUnicodeString");
Assert.AreSame(termsIndex, cache.GetTermsIndex(Reader, "theRandomUnicodeString"), "Second request to cache return same array");
BytesRef br = new BytesRef();
for (int i = 0; i < NUM_DOCS; i++)
{
BytesRef term;
int ord = termsIndex.GetOrd(i);
if (ord == -1)
{
term = null;
}
else
{
termsIndex.LookupOrd(ord, br);
term = br;
}
string s = term == null ? null : term.Utf8ToString();
Assert.IsTrue(UnicodeStrings[i] == null || UnicodeStrings[i].Equals(s, StringComparison.Ordinal), "for doc " + i + ": " + s + " does not equal: " + UnicodeStrings[i]);
}
int nTerms = termsIndex.ValueCount;
TermsEnum tenum = termsIndex.GetTermsEnum();
BytesRef val = new BytesRef();
for (int i = 0; i < nTerms; i++)
{
BytesRef val1 = tenum.Next();
termsIndex.LookupOrd(i, val);
// System.out.println("i="+i);
Assert.AreEqual(val, val1);
}
// seek the enum around (note this isn't a great test here)
int num = AtLeast(100);
for (int i = 0; i < num; i++)
{
int k = Random.Next(nTerms);
termsIndex.LookupOrd(k, val);
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, tenum.SeekCeil(val));
Assert.AreEqual(val, tenum.Term);
}
for (int i = 0; i < nTerms; i++)
{
termsIndex.LookupOrd(i, val);
Assert.AreEqual(TermsEnum.SeekStatus.FOUND, tenum.SeekCeil(val));
Assert.AreEqual(val, tenum.Term);
}
// test bad field
termsIndex = cache.GetTermsIndex(Reader, "bogusfield");
// getTerms
BinaryDocValues terms = cache.GetTerms(Reader, "theRandomUnicodeString", true);
Assert.AreSame(terms, cache.GetTerms(Reader, "theRandomUnicodeString", true), "Second request to cache return same array");
IBits bits = cache.GetDocsWithField(Reader, "theRandomUnicodeString");
for (int i = 0; i < NUM_DOCS; i++)
{
terms.Get(i, br);
BytesRef term;
if (!bits.Get(i))
{
term = null;
}
else
{
term = br;
}
string s = term == null ? null : term.Utf8ToString();
Assert.IsTrue(UnicodeStrings[i] == null || UnicodeStrings[i].Equals(s, StringComparison.Ordinal), "for doc " + i + ": " + s + " does not equal: " + UnicodeStrings[i]);
}
// test bad field
terms = cache.GetTerms(Reader, "bogusfield", false);
// getDocTermOrds
SortedSetDocValues termOrds = cache.GetDocTermOrds(Reader, "theRandomUnicodeMultiValuedField");
int numEntries = cache.GetCacheEntries().Length;
// ask for it again, and check that we didnt create any additional entries:
termOrds = cache.GetDocTermOrds(Reader, "theRandomUnicodeMultiValuedField");
Assert.AreEqual(numEntries, cache.GetCacheEntries().Length);
for (int i = 0; i < NUM_DOCS; i++)
{
termOrds.SetDocument(i);
// this will remove identical terms. A DocTermOrds doesn't return duplicate ords for a docId
IList <BytesRef> values = MultiValued[i].Distinct().ToList();
foreach (BytesRef v in values)
{
if (v == null)
{
// why does this test use null values... instead of an empty list: confusing
break;
}
long ord = termOrds.NextOrd();
Debug.Assert(ord != SortedSetDocValues.NO_MORE_ORDS);
BytesRef scratch = new BytesRef();
termOrds.LookupOrd(ord, scratch);
Assert.AreEqual(v, scratch);
}
Assert.AreEqual(SortedSetDocValues.NO_MORE_ORDS, termOrds.NextOrd());
}
// test bad field
termOrds = cache.GetDocTermOrds(Reader, "bogusfield");
Assert.IsTrue(termOrds.ValueCount == 0);
FieldCache.DEFAULT.PurgeByCacheKey(Reader.CoreCacheKey);
}
