public void Read(IndexInput input, FieldInfos fieldInfos)
{
this.Term = null; // invalidate cache
NewSuffixStart = input.ReadVInt();
int length = input.ReadVInt();
int totalLength = NewSuffixStart + length;
Debug.Assert(totalLength <= ByteBlockPool.BYTE_BLOCK_SIZE - 2, "termLength=" + totalLength + ",resource=" + input);
if (Bytes.Bytes.Length < totalLength)
{
Bytes.Grow(totalLength);
}
Bytes.Length = totalLength;
input.ReadBytes(Bytes.Bytes, NewSuffixStart, length);
int fieldNumber = input.ReadVInt();
if (fieldNumber != CurrentFieldNumber)
{
CurrentFieldNumber = fieldNumber;
// NOTE: too much sneakiness here, seriously this is a negative vint?!
if (CurrentFieldNumber == -1)
{
Field = "";
}
else
{
Debug.Assert(fieldInfos.FieldInfo(CurrentFieldNumber) != null, CurrentFieldNumber.ToString());
Field = String.Intern(fieldInfos.FieldInfo(CurrentFieldNumber).Name);
}
}
else
{
Debug.Assert(Field.Equals(fieldInfos.FieldInfo(fieldNumber).Name), "currentFieldNumber=" + CurrentFieldNumber + " field=" + Field + " vs " + fieldInfos.FieldInfo(fieldNumber) == null ? "null" : fieldInfos.FieldInfo(fieldNumber).Name);
}
}
/// <summary>
/// Save a single segment's info. </summary>
public override void Write(Directory dir, SegmentInfo si, FieldInfos fis, IOContext ioContext)
{
string fileName = IndexFileNames.SegmentFileName(si.Name, "", Lucene46SegmentInfoFormat.SI_EXTENSION);
si.AddFile(fileName);
IndexOutput output = dir.CreateOutput(fileName, ioContext);
bool success = false;
try
{
CodecUtil.WriteHeader(output, Lucene46SegmentInfoFormat.CODEC_NAME, Lucene46SegmentInfoFormat.VERSION_CURRENT);
// Write the Lucene version that created this segment, since 3.1
output.WriteString(si.Version);
output.WriteInt(si.DocCount);
output.WriteByte((byte)(sbyte)(si.UseCompoundFile ? SegmentInfo.YES : SegmentInfo.NO));
output.WriteStringStringMap(si.Diagnostics);
output.WriteStringSet(si.Files);
CodecUtil.WriteFooter(output);
success = true;
}
finally
{
if (!success)
{
IOUtils.CloseWhileHandlingException(output);
si.Dir.DeleteFile(fileName);
}
else
{
output.Dispose();
}
}
}
public SegmentTermDocs(IndexInput freqStream, TermInfosReader tis, FieldInfos fieldInfos)
{
this.FreqStream = (IndexInput)freqStream.Clone();
this.Tis = tis;
this.FieldInfos = fieldInfos;
SkipInterval = tis.SkipInterval;
MaxSkipLevels = tis.MaxSkipLevels;
}
internal TermInfosReader(Directory dir, string seg, FieldInfos fis, IOContext context, int indexDivisor)
{
bool success = false;
if (indexDivisor < 1 && indexDivisor != -1)
{
throw new System.ArgumentException("indexDivisor must be -1 (don't load terms index) or greater than 0: got " + indexDivisor);
}
try
{
Directory = dir;
Segment = seg;
FieldInfos = fis;
OrigEnum = new SegmentTermEnum(Directory.OpenInput(IndexFileNames.SegmentFileName(Segment, "", Lucene3xPostingsFormat.TERMS_EXTENSION), context), FieldInfos, false);
Size_Renamed = OrigEnum.Size;
if (indexDivisor != -1)
{
// Load terms index
TotalIndexInterval = OrigEnum.IndexInterval * indexDivisor;
string indexFileName = IndexFileNames.SegmentFileName(Segment, "", Lucene3xPostingsFormat.TERMS_INDEX_EXTENSION);
SegmentTermEnum indexEnum = new SegmentTermEnum(Directory.OpenInput(indexFileName, context), FieldInfos, true);
try
{
Index = new TermInfosReaderIndex(indexEnum, indexDivisor, dir.FileLength(indexFileName), TotalIndexInterval);
IndexLength = Index.Length();
}
finally
{
indexEnum.Dispose();
}
}
else
{
// Do not load terms index:
TotalIndexInterval = -1;
Index = null;
IndexLength = -1;
}
success = true;
}
finally
{
// With lock-less commits, it's entirely possible (and
// fine) to hit a FileNotFound exception above. In
// this case, we want to explicitly close any subset
// of things that were opened so that we don't have to
// wait for a GC to do so.
if (!success)
{
Dispose();
}
}
}
public void Read(IndexInput input, FieldInfos fieldInfos)
{
this.term = null; // invalidate cache
newSuffixStart = input.ReadVInt32();
int length = input.ReadVInt32();
int totalLength = newSuffixStart + length;
if (Debugging.AssertsEnabled)
{
Debugging.Assert(totalLength <= ByteBlockPool.BYTE_BLOCK_SIZE - 2, () => "termLength=" + totalLength + ",resource=" + input);
}
if (bytes.Bytes.Length < totalLength)
{
bytes.Grow(totalLength);
}
bytes.Length = totalLength;
input.ReadBytes(bytes.Bytes, newSuffixStart, length);
int fieldNumber = input.ReadVInt32();
if (fieldNumber != currentFieldNumber)
{
currentFieldNumber = fieldNumber;
// NOTE: too much sneakiness here, seriously this is a negative vint?!
if (currentFieldNumber == -1)
{
field = "";
}
else
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(fieldInfos.FieldInfo(currentFieldNumber) != null, currentFieldNumber.ToString);
}
field = fieldInfos.FieldInfo(currentFieldNumber).Name.Intern();
}
}
else
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(field.Equals(fieldInfos.FieldInfo(fieldNumber).Name, StringComparison.Ordinal),
() => "currentFieldNumber=" + currentFieldNumber +
" field=" + field +
" vs " + fieldInfos.FieldInfo(fieldNumber) == null ? "null" : fieldInfos.FieldInfo(fieldNumber).Name);
}
}
}
public SegmentTermEnum(IndexInput i, FieldInfos fis, bool isi)
{
input = i;
fieldInfos = fis;
isIndex = isi;
maxSkipLevels = 1; // use single-level skip lists for formats > -3
int firstInt = input.ReadInt32();
if (firstInt >= 0)
{
// original-format file, without explicit format version number
format = 0;
size = firstInt;
// back-compatible settings
indexInterval = 128;
skipInterval = int.MaxValue; // switch off skipTo optimization
}
else
{
// we have a format version number
format = firstInt;
// check that it is a format we can understand
if (format > FORMAT_MINIMUM)
{
throw new IndexFormatTooOldException(input, format, FORMAT_MINIMUM, FORMAT_CURRENT);
}
if (format < FORMAT_CURRENT)
{
throw new IndexFormatTooNewException(input, format, FORMAT_MINIMUM, FORMAT_CURRENT);
}
size = input.ReadInt64(); // read the size
indexInterval = input.ReadInt32();
skipInterval = input.ReadInt32();
maxSkipLevels = input.ReadInt32();
if (Debugging.AssertsEnabled)
{
Debugging.Assert(indexInterval > 0, () => "indexInterval=" + indexInterval + " is negative; must be > 0");
Debugging.Assert(skipInterval > 0, () => "skipInterval=" + skipInterval + " is negative; must be > 0");
}
}
}
public override TermVectorsReader VectorsReader(Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context)
{
string fileName = IndexFileNames.SegmentFileName(Lucene3xSegmentInfoFormat.GetDocStoreSegment(segmentInfo), "", Lucene3xTermVectorsReader.VECTORS_FIELDS_EXTENSION);
// Unfortunately, for 3.x indices, each segment's
// FieldInfos can lie about hasVectors (claim it's true
// when really it's false).... so we have to carefully
// check if the files really exist before trying to open
// them (4.x has fixed this):
bool exists;
if (Lucene3xSegmentInfoFormat.GetDocStoreOffset(segmentInfo) != -1 && Lucene3xSegmentInfoFormat.GetDocStoreIsCompoundFile(segmentInfo))
{
string cfxFileName = IndexFileNames.SegmentFileName(Lucene3xSegmentInfoFormat.GetDocStoreSegment(segmentInfo), "", Lucene3xCodec.COMPOUND_FILE_STORE_EXTENSION);
if (segmentInfo.Dir.FileExists(cfxFileName))
{
Directory cfsDir = new CompoundFileDirectory(segmentInfo.Dir, cfxFileName, context, false);
try
{
exists = cfsDir.FileExists(fileName);
}
finally
{
cfsDir.Dispose();
}
}
else
{
exists = false;
}
}
else
{
exists = directory.FileExists(fileName);
}
if (!exists)
{
// 3x's FieldInfos sometimes lies and claims a segment
// has vectors when it doesn't:
return null;
}
else
{
return new Lucene3xTermVectorsReader(directory, segmentInfo, fieldInfos, context);
}
}
/// <summary>
/// Sole constructor. </summary>
public Lucene40StoredFieldsReader(Directory d, SegmentInfo si, FieldInfos fn, IOContext context)
{
string segment = si.Name;
bool success = false;
FieldInfos = fn;
try
{
FieldsStream = d.OpenInput(IndexFileNames.SegmentFileName(segment, "", Lucene40StoredFieldsWriter.FIELDS_EXTENSION), context);
string indexStreamFN = IndexFileNames.SegmentFileName(segment, "", Lucene40StoredFieldsWriter.FIELDS_INDEX_EXTENSION);
IndexStream = d.OpenInput(indexStreamFN, context);
CodecUtil.CheckHeader(IndexStream, Lucene40StoredFieldsWriter.CODEC_NAME_IDX, Lucene40StoredFieldsWriter.VERSION_START, Lucene40StoredFieldsWriter.VERSION_CURRENT);
CodecUtil.CheckHeader(FieldsStream, Lucene40StoredFieldsWriter.CODEC_NAME_DAT, Lucene40StoredFieldsWriter.VERSION_START, Lucene40StoredFieldsWriter.VERSION_CURRENT);
Debug.Assert(Lucene40StoredFieldsWriter.HEADER_LENGTH_DAT == FieldsStream.FilePointer);
Debug.Assert(Lucene40StoredFieldsWriter.HEADER_LENGTH_IDX == IndexStream.FilePointer);
long indexSize = IndexStream.Length() - Lucene40StoredFieldsWriter.HEADER_LENGTH_IDX;
this.Size_Renamed = (int)(indexSize >> 3);
// Verify two sources of "maxDoc" agree:
if (this.Size_Renamed != si.DocCount)
{
throw new CorruptIndexException("doc counts differ for segment " + segment + ": fieldsReader shows " + this.Size_Renamed + " but segmentInfo shows " + si.DocCount);
}
NumTotalDocs = (int)(indexSize >> 3);
success = true;
}
finally
{
// With lock-less commits, it's entirely possible (and
// fine) to hit a FileNotFound exception above. In
// this case, we want to explicitly close any subset
// of things that were opened so that we don't have to
// wait for a GC to do so.
if (!success)
{
try
{
Dispose();
} // ensure we throw our original exception
catch (Exception)
{
}
}
}
}
/// <summary>
/// Save a single segment's info. </summary>
public override void Write(Directory dir, SegmentInfo si, FieldInfos fis, IOContext ioContext)
{
string fileName = IndexFileNames.SegmentFileName(si.Name, "", Lucene46SegmentInfoFormat.SI_EXTENSION);
si.AddFile(fileName);
IndexOutput output = dir.CreateOutput(fileName, ioContext);
bool success = false;
try
{
CodecUtil.WriteHeader(output, Lucene46SegmentInfoFormat.CODEC_NAME, Lucene46SegmentInfoFormat.VERSION_CURRENT);
// Write the Lucene version that created this segment, since 3.1
output.WriteString(si.Version);
output.WriteInt32(si.DocCount);
output.WriteByte((byte)(si.UseCompoundFile ? SegmentInfo.YES : SegmentInfo.NO));
output.WriteStringStringMap(si.Diagnostics);
output.WriteStringSet(si.GetFiles());
CodecUtil.WriteFooter(output);
success = true;
}
finally
{
if (!success)
{
IOUtils.DisposeWhileHandlingException(output);
si.Dir.DeleteFile(fileName);
}
else
{
output.Dispose();
}
}
}
/// <summary>
/// sugar method for startDocument() + writeField() for every stored field in the document </summary>
protected internal void AddDocument <T1>(IEnumerable <T1> doc, FieldInfos fieldInfos) where T1 : Lucene.Net.Index.IndexableField
{
int storedCount = 0;
foreach (IndexableField field in doc)
{
if (field.FieldType().Stored)
{
storedCount++;
}
}
StartDocument(storedCount);
foreach (IndexableField field in doc)
{
if (field.FieldType().Stored)
{
WriteField(fieldInfos.FieldInfo(field.Name()), field);
}
}
FinishDocument();
}
public override void Finish(FieldInfos fis, int numDocs)
{
if (4 + ((long)numDocs) * 16 != Tvx.FilePointer)
// this is most likely a bug in Sun JRE 1.6.0_04/_05;
// we detect that the bug has struck, here, and
// throw an exception to prevent the corruption from
// entering the index. See LUCENE-1282 for
// details.
{
throw new Exception("tvx size mismatch: mergedDocs is " + numDocs + " but tvx size is " + Tvx.FilePointer + " file=" + Tvx.ToString() + "; now aborting this merge to prevent index corruption");
}
}
/// <summary> /// Returns a <seealso cref="StoredFieldsReader"/> to load stored /// fields. /// </summary> public abstract StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context);
public override void Write(Directory directory, string segmentName, string segmentSuffix, FieldInfos infos, IOContext context)
{
string fileName = IndexFileNames.SegmentFileName(segmentName, "", FIELD_INFOS_EXTENSION);
IndexOutput output = directory.CreateOutput(fileName, context);
bool success = false;
try
{
output.WriteVInt(FORMAT_PREFLEX_RW);
output.WriteVInt(infos.Size());
foreach (FieldInfo fi in infos)
{
sbyte bits = 0x0;
if (fi.HasVectors())
{
bits |= STORE_TERMVECTOR;
}
if (fi.OmitsNorms())
{
bits |= OMIT_NORMS;
}
if (fi.HasPayloads())
{
bits |= STORE_PAYLOADS;
}
if (fi.Indexed)
{
bits |= IS_INDEXED;
Debug.Assert(fi.FieldIndexOptions == FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS || !fi.HasPayloads());
if (fi.FieldIndexOptions == FieldInfo.IndexOptions.DOCS_ONLY)
{
bits |= OMIT_TERM_FREQ_AND_POSITIONS;
}
else if (fi.FieldIndexOptions == FieldInfo.IndexOptions.DOCS_AND_FREQS)
{
bits |= OMIT_POSITIONS;
}
}
output.WriteString(fi.Name);
/*
* we need to write the field number since IW tries
* to stabelize the field numbers across segments so the
* FI ordinal is not necessarily equivalent to the field number
*/
output.WriteInt(fi.Number);
output.WriteByte((byte)bits);
if (fi.Indexed && !fi.OmitsNorms())
{
// to allow null norm types we need to indicate if norms are written
// only in RW case
output.WriteByte((byte)(sbyte)(fi.NormType == null ? 0 : 1));
}
Debug.Assert(fi.Attributes() == null); // not used or supported
}
success = true;
}
finally
{
if (success)
{
output.Dispose();
}
else
{
IOUtils.CloseWhileHandlingException(output);
}
}
}
public override void Write(Directory directory, string segmentName, string segmentSuffix, FieldInfos infos, IOContext context)
{
string fileName = IndexFileNames.SegmentFileName(segmentName, "", Lucene42FieldInfosFormat.EXTENSION);
IndexOutput output = directory.CreateOutput(fileName, context);
bool success = false;
try
{
CodecUtil.WriteHeader(output, Lucene42FieldInfosFormat.CODEC_NAME, Lucene42FieldInfosFormat.FORMAT_CURRENT);
output.WriteVInt(infos.Size());
foreach (FieldInfo fi in infos)
{
FieldInfo.IndexOptions? indexOptions = fi.FieldIndexOptions;
sbyte bits = 0x0;
if (fi.HasVectors())
{
bits |= Lucene42FieldInfosFormat.STORE_TERMVECTOR;
}
if (fi.OmitsNorms())
{
bits |= Lucene42FieldInfosFormat.OMIT_NORMS;
}
if (fi.HasPayloads())
{
bits |= Lucene42FieldInfosFormat.STORE_PAYLOADS;
}
if (fi.Indexed)
{
bits |= Lucene42FieldInfosFormat.IS_INDEXED;
Debug.Assert(indexOptions >= FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS || !fi.HasPayloads());
if (indexOptions == FieldInfo.IndexOptions.DOCS_ONLY)
{
bits |= Lucene42FieldInfosFormat.OMIT_TERM_FREQ_AND_POSITIONS;
}
else if (indexOptions == FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS)
{
bits |= Lucene42FieldInfosFormat.STORE_OFFSETS_IN_POSTINGS;
}
else if (indexOptions == FieldInfo.IndexOptions.DOCS_AND_FREQS)
{
bits |= Lucene42FieldInfosFormat.OMIT_POSITIONS;
}
}
output.WriteString(fi.Name);
output.WriteVInt(fi.Number);
output.WriteByte((byte)bits);
// pack the DV types in one byte
var dv = DocValuesByte(fi.DocValuesType);
var nrm = DocValuesByte(fi.NormType);
Debug.Assert((dv & (~0xF)) == 0 && (nrm & (~0x0F)) == 0);
var val = unchecked((sbyte)(0xff & ((nrm << 4) | dv)));
output.WriteByte((byte)val);
output.WriteStringStringMap(fi.Attributes());
}
success = true;
}
finally
{
if (success)
{
output.Dispose();
}
else
{
IOUtils.CloseWhileHandlingException(output);
}
}
}
public override TermVectorsReader VectorsReader(Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context)
{
return new Lucene3xTermVectorsReaderAnonymousInnerClassHelper(this, directory, segmentInfo, fieldInfos, context);
}
public override void Finish(FieldInfos fis, int numDocs)
{
Debug.Assert(DocCount == numDocs);
Debug.Assert(DocStatus == (numDocs > 0 ? Status.FINISHED : Status.UNDEFINED));
Debug.Assert(FieldStatus != Status.STARTED);
Debug.Assert(TermStatus != Status.STARTED);
@in.Finish(fis, numDocs);
}
private TermInfosWriter(Directory directory, string segment, FieldInfos fis, int interval, bool isIndex)
{
Initialize(directory, segment, fis, interval, isIndex);
}
/// <summary> /// Write <seealso cref="SegmentInfo"/> data. </summary> /// <exception cref="IOException"> If an I/O error occurs </exception> public abstract void Write(Directory dir, SegmentInfo info, FieldInfos fis, IOContext ioContext);
public override TermVectorsReader VectorsReader(Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context)
{
return(new Lucene40TermVectorsReader(directory, segmentInfo, fieldInfos, context));
}
public override StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context)
{
return new Lucene40StoredFieldsReader(directory, si, fn, context);
}
private void Initialize(Directory directory, string segment, FieldInfos fis, int interval, bool isi)
{
IndexInterval = interval;
FieldInfos = fis;
IsIndex = isi;
Output = directory.CreateOutput(IndexFileNames.SegmentFileName(segment, "", (IsIndex ? Lucene3xPostingsFormat.TERMS_INDEX_EXTENSION : Lucene3xPostingsFormat.TERMS_EXTENSION)), IOContext.DEFAULT);
bool success = false;
try
{
Output.WriteInt(FORMAT_CURRENT); // write format
Output.WriteLong(0); // leave space for size
Output.WriteInt(IndexInterval); // write indexInterval
Output.WriteInt(SkipInterval); // write skipInterval
Output.WriteInt(MaxSkipLevels); // write maxSkipLevels
Debug.Assert(InitUTF16Results());
success = true;
}
finally
{
if (!success)
{
IOUtils.CloseWhileHandlingException(Output);
try
{
directory.DeleteFile(IndexFileNames.SegmentFileName(segment, "", (IsIndex ? Lucene3xPostingsFormat.TERMS_INDEX_EXTENSION : Lucene3xPostingsFormat.TERMS_EXTENSION)));
}
catch (IOException ignored)
{
}
}
}
}
/// <summary>
/// note: -1 is the empty field: "" !!!! </summary>
internal static string FieldName(FieldInfos infos, int fieldNumber)
{
if (fieldNumber == -1)
{
return "";
}
else
{
return infos.FieldInfo(fieldNumber).Name;
}
}
public override void Finish(FieldInfos fis, int numDocs)
{
Debug.Assert(DocStatus == (numDocs > 0 ? Status.FINISHED : Status.UNDEFINED));
@in.Finish(fis, numDocs);
Debug.Assert(FieldCount == 0);
Debug.Assert(numDocs == NumWritten);
}
public override StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context)
{
return new CompressingStoredFieldsReader(directory, si, SegmentSuffix, fn, context, FormatName, CompressionMode);
}
public override sealed TermVectorsReader VectorsReader(Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context)
{
return(new CompressingTermVectorsReader(directory, segmentInfo, SegmentSuffix, fieldInfos, context, FormatName, CompressionMode));
}
internal TermInfosWriter(Directory directory, string segment, FieldInfos fis, int interval)
{
Initialize(directory, segment, fis, interval, false);
bool success = false;
try
{
Other = new TermInfosWriter(directory, segment, fis, interval, true);
Other.Other = this;
success = true;
}
finally
{
if (!success)
{
IOUtils.CloseWhileHandlingException(Output);
try
{
directory.DeleteFile(IndexFileNames.SegmentFileName(segment, "", (IsIndex ? Lucene3xPostingsFormat.TERMS_INDEX_EXTENSION : Lucene3xPostingsFormat.TERMS_EXTENSION)));
}
catch (IOException ignored)
{
}
}
}
}
public override TermVectorsReader VectorsReader(Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context)
{
return new AssertingTermVectorsReader(@in.VectorsReader(directory, segmentInfo, fieldInfos, context));
}
private void ReadFields(IndexInput meta, FieldInfos infos)
{
int fieldNumber = meta.ReadVInt();
while (fieldNumber != -1)
{
// check should be: infos.fieldInfo(fieldNumber) != null, which incorporates negative check
// but docvalues updates are currently buggy here (loading extra stuff, etc): LUCENE-5616
if (fieldNumber < 0)
{
// trickier to validate more: because we re-use for norms, because we use multiple entries
// for "composite" types like sortedset, etc.
throw new Exception("Invalid field number: " + fieldNumber + " (resource=" + meta + ")");
}
byte type = meta.ReadByte();
if (type == Lucene45DocValuesFormat.NUMERIC)
{
Numerics[fieldNumber] = ReadNumericEntry(meta);
}
else if (type == Lucene45DocValuesFormat.BINARY)
{
BinaryEntry b = ReadBinaryEntry(meta);
Binaries[fieldNumber] = b;
}
else if (type == Lucene45DocValuesFormat.SORTED)
{
ReadSortedField(fieldNumber, meta, infos);
}
else if (type == Lucene45DocValuesFormat.SORTED_SET)
{
SortedSetEntry ss = ReadSortedSetEntry(meta);
SortedSets[fieldNumber] = ss;
if (ss.Format == Lucene45DocValuesConsumer.SORTED_SET_WITH_ADDRESSES)
{
ReadSortedSetFieldWithAddresses(fieldNumber, meta, infos);
}
else if (ss.Format == Lucene45DocValuesConsumer.SORTED_SET_SINGLE_VALUED_SORTED)
{
if (meta.ReadVInt() != fieldNumber)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
if (meta.ReadByte() != Lucene45DocValuesFormat.SORTED)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
ReadSortedField(fieldNumber, meta, infos);
}
else
{
throw new Exception();
}
}
else
{
throw new Exception("invalid type: " + type + ", resource=" + meta);
}
fieldNumber = meta.ReadVInt();
}
}
public override FieldInfos Read(Directory directory, string segmentName, string segmentSuffix, IOContext iocontext)
{
string fileName = IndexFileNames.SegmentFileName(segmentName, "", Lucene40FieldInfosFormat.FIELD_INFOS_EXTENSION);
IndexInput input = directory.OpenInput(fileName, iocontext);
bool success = false;
try
{
CodecUtil.CheckHeader(input, Lucene40FieldInfosFormat.CODEC_NAME, Lucene40FieldInfosFormat.FORMAT_START, Lucene40FieldInfosFormat.FORMAT_CURRENT);
int size = input.ReadVInt(); //read in the size
FieldInfo[] infos = new FieldInfo[size];
for (int i = 0; i < size; i++)
{
string name = input.ReadString();
int fieldNumber = input.ReadVInt();
byte bits = input.ReadByte();
bool isIndexed = (bits & Lucene40FieldInfosFormat.IS_INDEXED) != 0;
bool storeTermVector = (bits & Lucene40FieldInfosFormat.STORE_TERMVECTOR) != 0;
bool omitNorms = (bits & Lucene40FieldInfosFormat.OMIT_NORMS) != 0;
bool storePayloads = (bits & Lucene40FieldInfosFormat.STORE_PAYLOADS) != 0;
FieldInfo.IndexOptions indexOptions;
if (!isIndexed)
{
indexOptions = default(FieldInfo.IndexOptions);
}
else if ((bits & Lucene40FieldInfosFormat.OMIT_TERM_FREQ_AND_POSITIONS) != 0)
{
indexOptions = FieldInfo.IndexOptions.DOCS_ONLY;
}
else if ((bits & Lucene40FieldInfosFormat.OMIT_POSITIONS) != 0)
{
indexOptions = FieldInfo.IndexOptions.DOCS_AND_FREQS;
}
else if ((bits & Lucene40FieldInfosFormat.STORE_OFFSETS_IN_POSTINGS) != 0)
{
indexOptions = FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
}
else
{
indexOptions = FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS;
}
// LUCENE-3027: past indices were able to write
// storePayloads=true when omitTFAP is also true,
// which is invalid. We correct that, here:
if (isIndexed && indexOptions.CompareTo(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS) < 0)
{
storePayloads = false;
}
// DV Types are packed in one byte
byte val = input.ReadByte();
LegacyDocValuesType oldValuesType = GetDocValuesType((sbyte)(val & 0x0F));
LegacyDocValuesType oldNormsType = GetDocValuesType((sbyte)(((int)((uint)val >> 4)) & 0x0F));
IDictionary<string, string> attributes = input.ReadStringStringMap();
if (oldValuesType.Mapping != null)
{
attributes[LEGACY_DV_TYPE_KEY] = oldValuesType.Name;
}
if (oldNormsType.Mapping != null)
{
if (oldNormsType.Mapping != FieldInfo.DocValuesType_e.NUMERIC)
{
throw new CorruptIndexException("invalid norm type: " + oldNormsType + " (resource=" + input + ")");
}
attributes[LEGACY_NORM_TYPE_KEY] = oldNormsType.Name;
}
infos[i] = new FieldInfo(name, isIndexed, fieldNumber, storeTermVector, omitNorms, storePayloads, indexOptions, oldValuesType.Mapping, oldNormsType.Mapping, attributes);
}
CodecUtil.CheckEOF(input);
FieldInfos fieldInfos = new FieldInfos(infos);
success = true;
return fieldInfos;
}
finally
{
if (success)
{
input.Dispose();
}
else
{
IOUtils.CloseWhileHandlingException(input);
}
}
}
private void ReadSortedField(int fieldNumber, IndexInput meta, FieldInfos infos)
{
// sorted = binary + numeric
if (meta.ReadVInt() != fieldNumber)
{
throw new Exception("sorted entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
if (meta.ReadByte() != Lucene45DocValuesFormat.BINARY)
{
throw new Exception("sorted entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
BinaryEntry b = ReadBinaryEntry(meta);
Binaries[fieldNumber] = b;
if (meta.ReadVInt() != fieldNumber)
{
throw new Exception("sorted entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
if (meta.ReadByte() != Lucene45DocValuesFormat.NUMERIC)
{
throw new Exception("sorted entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
NumericEntry n = ReadNumericEntry(meta);
Ords[fieldNumber] = n;
}
/// <summary> /// Writes the provided <see cref="FieldInfos"/> to the /// directory. /// </summary> public abstract void Write(Directory directory, string segmentName, string segmentSuffix, FieldInfos infos, IOContext context);
private void ReadSortedSetFieldWithAddresses(int fieldNumber, IndexInput meta, FieldInfos infos)
{
// sortedset = binary + numeric (addresses) + ordIndex
if (meta.ReadVInt() != fieldNumber)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
if (meta.ReadByte() != Lucene45DocValuesFormat.BINARY)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
BinaryEntry b = ReadBinaryEntry(meta);
Binaries[fieldNumber] = b;
if (meta.ReadVInt() != fieldNumber)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
if (meta.ReadByte() != Lucene45DocValuesFormat.NUMERIC)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
NumericEntry n1 = ReadNumericEntry(meta);
Ords[fieldNumber] = n1;
if (meta.ReadVInt() != fieldNumber)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
if (meta.ReadByte() != Lucene45DocValuesFormat.NUMERIC)
{
throw new Exception("sortedset entry for field: " + fieldNumber + " is corrupt (resource=" + meta + ")");
}
NumericEntry n2 = ReadNumericEntry(meta);
OrdIndexes[fieldNumber] = n2;
}
public Lucene3xTermVectorsReaderAnonymousInnerClassHelper(PreFlexRWTermVectorsFormat outerInstance, Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context)
: base(directory, segmentInfo, fieldInfos, context)
{
this.OuterInstance = outerInstance;
}
/// <summary> /// Write <see cref="SegmentInfo"/> data. </summary> /// <exception cref="System.IO.IOException"> If an I/O error occurs. </exception> public abstract void Write(Directory dir, SegmentInfo info, FieldInfos fis, IOContext ioContext);
/// <summary> /// Called before <seealso cref="#close()"/>, passing in the number /// of documents that were written. Note that this is /// intentionally redundant (equivalent to the number of /// calls to <seealso cref="#startDocument(int)"/>, but a Codec should /// check that this is the case to detect the JRE bug described /// in LUCENE-1282. /// </summary> public abstract void Finish(FieldInfos fis, int numDocs);
public override void Finish(FieldInfos fis, int numDocs)
{
if (HEADER_LENGTH_IDX + ((long)numDocs) * 8 != IndexStream.FilePointer)
// this is most likely a bug in Sun JRE 1.6.0_04/_05;
// we detect that the bug has struck, here, and
// throw an exception to prevent the corruption from
// entering the index. See LUCENE-1282 for
// details.
{
throw new Exception("fdx size mismatch: docCount is " + numDocs + " but fdx file size is " + IndexStream.FilePointer + " file=" + IndexStream.ToString() + "; now aborting this merge to prevent index corruption");
}
}
// NOTE: this is not "really" 3.x format, because we are
// writing each SI to its own file, vs 3.x where the list
// of segments and SI for each segment is written into a
// single segments_N file
/// <summary>
/// Save a single segment's info. </summary>
public override void Write(Directory dir, SegmentInfo si, FieldInfos fis, IOContext ioContext)
{
SegmentInfos.Write3xInfo(dir, si, ioContext);
}
// NOTE: this is not "really" 3.x format, because we are
// writing each SI to its own file, vs 3.x where the list
// of segments and SI for each segment is written into a
// single segments_N file
/// <summary>
/// Save a single segment's info. </summary>
public override void Write(Directory dir, SegmentInfo si, FieldInfos fis, IOContext ioContext)
{
SegmentInfos.Write3xInfo(dir, si, ioContext);
}
public override StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context)
{
return(new CompressingStoredFieldsReader(directory, si, segmentSuffix, fn, context, formatName, compressionMode));
}
public override StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context)
{
return(new Lucene40StoredFieldsReader(directory, si, fn, context));
}
/// <summary> /// Returns a <seealso cref="StoredFieldsReader"/> to load stored /// fields. /// </summary> public abstract StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context);
/// <summary> /// Returns a <see cref="TermVectorsReader"/> to read term /// vectors. /// </summary> public abstract TermVectorsReader VectorsReader(Directory directory, SegmentInfo segmentInfo, FieldInfos fieldInfos, IOContext context);
public override StoredFieldsReader FieldsReader(Directory directory, SegmentInfo si, FieldInfos fn, IOContext context)
{
return new AssertingStoredFieldsReader(@in.FieldsReader(directory, si, fn, context), si.DocCount);
}
/// <summary>
/// Used only by clone. </summary>
private Lucene40StoredFieldsReader(FieldInfos fieldInfos, int numTotalDocs, int size, IndexInput fieldsStream, IndexInput indexStream)
{
this.FieldInfos = fieldInfos;
this.NumTotalDocs = numTotalDocs;
this.Size_Renamed = size;
this.FieldsStream = fieldsStream;
this.IndexStream = indexStream;
}
