public static void ReadLegacyInfos(SegmentInfos infos, Directory directory, IndexInput input, int format)
{
infos.Version = input.ReadLong(); // read version
infos.Counter = input.ReadInt(); // read counter
Lucene3xSegmentInfoReader reader = new Lucene3xSegmentInfoReader();
for (int i = input.ReadInt(); i > 0; i--) // read segmentInfos
{
SegmentCommitInfo siPerCommit = reader.ReadLegacySegmentInfo(directory, format, input);
SegmentInfo si = siPerCommit.Info;
if (si.Version == null)
{
// Could be a 3.0 - try to open the doc stores - if it fails, it's a
// 2.x segment, and an IndexFormatTooOldException will be thrown,
// which is what we want.
Directory dir = directory;
if (Lucene3xSegmentInfoFormat.GetDocStoreOffset(si) != -1)
{
if (Lucene3xSegmentInfoFormat.GetDocStoreIsCompoundFile(si))
{
dir = new CompoundFileDirectory(dir, IndexFileNames.SegmentFileName(Lucene3xSegmentInfoFormat.GetDocStoreSegment(si), "", Lucene3xCodec.COMPOUND_FILE_STORE_EXTENSION), IOContext.READONCE, false);
}
}
else if (si.UseCompoundFile)
{
dir = new CompoundFileDirectory(dir, IndexFileNames.SegmentFileName(si.Name, "", IndexFileNames.COMPOUND_FILE_EXTENSION), IOContext.READONCE, false);
}
try
{
Lucene3xStoredFieldsReader.CheckCodeVersion(dir, Lucene3xSegmentInfoFormat.GetDocStoreSegment(si));
}
finally
{
// If we opened the directory, close it
if (dir != directory)
{
dir.Dispose();
}
}
// Above call succeeded, so it's a 3.0 segment. Upgrade it so the next
// time the segment is read, its version won't be null and we won't
// need to open FieldsReader every time for each such segment.
si.Version = "3.0";
}
else if (si.Version.Equals("2.x"))
{
// If it's a 3x index touched by 3.1+ code, then segments record their
// version, whether they are 2.x ones or not. We detect that and throw
// appropriate exception.
throw new IndexFormatTooOldException("segment " + si.Name + " in resource " + input, si.Version);
}
infos.Add(siPerCommit);
}
infos.UserData = input.ReadStringStringMap();
}
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.ReadInt();
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.ReadLong(); // read the size
IndexInterval = Input.ReadInt();
SkipInterval = Input.ReadInt();
MaxSkipLevels = Input.ReadInt();
Debug.Assert(IndexInterval > 0, "indexInterval=" + IndexInterval + " is negative; must be > 0");
Debug.Assert(SkipInterval > 0, "skipInterval=" + SkipInterval + " is negative; must be > 0");
}
}
public FixedGapTermsIndexReader(Directory dir, FieldInfos fieldInfos, String segment, int indexDivisor,
IComparer<BytesRef> termComp, String segmentSuffix, IOContext context)
{
_termComp = termComp;
Debug.Assert(indexDivisor == -1 || indexDivisor > 0);
_input =
dir.OpenInput(
IndexFileNames.SegmentFileName(segment, segmentSuffix,
FixedGapTermsIndexWriter.TERMS_INDEX_EXTENSION),
context);
var success = false;
try
{
_version = ReadHeader(_input);
if (_version >= FixedGapTermsIndexWriter.VERSION_CHECKSUM)
CodecUtil.ChecksumEntireFile(_input);
indexInterval = _input.ReadInt();
if (indexInterval < 1)
{
throw new CorruptIndexException(String.Format("Invalid indexInterval: {0}, Resource: {1}",
indexInterval, _input));
}
_indexDivisor = indexDivisor;
if (indexDivisor < 0)
{
_totalIndexInterval = indexInterval;
}
else
{
// In case terms index gets loaded, later, on demand
_totalIndexInterval = indexInterval*indexDivisor;
}
Debug.Assert(_totalIndexInterval > 0);
SeekDir(_input, _dirOffset);
// Read directory
int numFields = _input.ReadVInt();
if (numFields < 0)
throw new CorruptIndexException(String.Format("Invalid numFields: {0}, Resource: {1}", numFields,
_input));
for (int i = 0; i < numFields; i++)
{
int field = _input.ReadVInt();
int numIndexTerms = _input.ReadVInt();
if (numIndexTerms < 0)
throw new CorruptIndexException(String.Format("Invalid numIndexTerms: {0}, Resource: {1}",
numIndexTerms,
_input));
long termsStart = _input.ReadVLong();
long indexStart = _input.ReadVLong();
long packedIndexStart = _input.ReadVLong();
long packedOffsetsStart = _input.ReadVLong();
if (packedIndexStart < indexStart)
throw new CorruptIndexException(
String.Format(
"Invalid packedIndexStart: {0}, IndexStart: {1}, NumIndexTerms: {2}, Resource: {3}",
packedIndexStart,
indexStart, numIndexTerms, _input));
FieldInfo fieldInfo = fieldInfos.FieldInfo(field);
try
{
_fields.Add(fieldInfo,
new FieldIndexData(numIndexTerms, indexStart, termsStart, packedIndexStart,
packedOffsetsStart, this));
}
catch (ArgumentException)
{
throw new CorruptIndexException(String.Format("Duplicate field: {0}, Resource {1}",
fieldInfo.Name,
_input));
}
}
success = true;
}
finally
{
if (!success)
{
IOUtils.CloseWhileHandlingException(_input);
}
if (indexDivisor > 0)
{
_input.Dispose();
_input = null;
if (success)
_indexLoaded = true;
_termBytesReader = _termBytes.Freeze(true);
}
}
}
public override void Init(IndexInput termsIn)
{
// Make sure we are talking to the matching past writer
CodecUtil.CheckHeader(termsIn, TERMS_CODEC, VERSION_START, VERSION_CURRENT);
SkipInterval = termsIn.ReadInt();
MaxSkipLevels = termsIn.ReadInt();
SkipMinimum = termsIn.ReadInt();
}
public Lucene3xStoredFieldsReader(Directory d, SegmentInfo si, FieldInfos fn, IOContext context)
{
string segment = Lucene3xSegmentInfoFormat.GetDocStoreSegment(si);
int docStoreOffset = Lucene3xSegmentInfoFormat.GetDocStoreOffset(si);
int size = si.DocCount;
bool success = false;
FieldInfos = fn;
try
{
if (docStoreOffset != -1 && Lucene3xSegmentInfoFormat.GetDocStoreIsCompoundFile(si))
{
d = StoreCFSReader = new CompoundFileDirectory(si.Dir, IndexFileNames.SegmentFileName(segment, "", Lucene3xCodec.COMPOUND_FILE_STORE_EXTENSION), context, false);
}
else
{
StoreCFSReader = null;
}
FieldsStream = d.OpenInput(IndexFileNames.SegmentFileName(segment, "", FIELDS_EXTENSION), context);
string indexStreamFN = IndexFileNames.SegmentFileName(segment, "", FIELDS_INDEX_EXTENSION);
IndexStream = d.OpenInput(indexStreamFN, context);
Format = IndexStream.ReadInt();
if (Format < FORMAT_MINIMUM)
{
throw new IndexFormatTooOldException(IndexStream, Format, FORMAT_MINIMUM, FORMAT_CURRENT);
}
if (Format > FORMAT_CURRENT)
{
throw new IndexFormatTooNewException(IndexStream, Format, FORMAT_MINIMUM, FORMAT_CURRENT);
}
long indexSize = IndexStream.Length() - FORMAT_SIZE;
if (docStoreOffset != -1)
{
// We read only a slice out of this shared fields file
this.DocStoreOffset = docStoreOffset;
this.Size = size;
// Verify the file is long enough to hold all of our
// docs
Debug.Assert(((int)(indexSize / 8)) >= size + this.DocStoreOffset, "indexSize=" + indexSize + " size=" + size + " docStoreOffset=" + docStoreOffset);
}
else
{
this.DocStoreOffset = 0;
this.Size = (int)(indexSize >> 3);
// Verify two sources of "maxDoc" agree:
if (this.Size != si.DocCount)
{
throw new CorruptIndexException("doc counts differ for segment " + segment + ": fieldsReader shows " + this.Size + " 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();
} // keep our original exception
catch (Exception t)
{
}
}
}
}
/// <summary>
/// reads from legacy 3.x segments_N </summary>
private SegmentCommitInfo ReadLegacySegmentInfo(Directory dir, int format, IndexInput input)
{
// check that it is a format we can understand
if (format > Lucene3xSegmentInfoFormat.FORMAT_DIAGNOSTICS)
{
throw new IndexFormatTooOldException(input, format, Lucene3xSegmentInfoFormat.FORMAT_DIAGNOSTICS, Lucene3xSegmentInfoFormat.FORMAT_3_1);
}
if (format < Lucene3xSegmentInfoFormat.FORMAT_3_1)
{
throw new IndexFormatTooNewException(input, format, Lucene3xSegmentInfoFormat.FORMAT_DIAGNOSTICS, Lucene3xSegmentInfoFormat.FORMAT_3_1);
}
string version;
if (format <= Lucene3xSegmentInfoFormat.FORMAT_3_1)
{
version = input.ReadString();
}
else
{
version = null;
}
string name = input.ReadString();
int docCount = input.ReadInt();
long delGen = input.ReadLong();
int docStoreOffset = input.ReadInt();
IDictionary<string, string> attributes = new Dictionary<string, string>();
// parse the docstore stuff and shove it into attributes
string docStoreSegment;
bool docStoreIsCompoundFile;
if (docStoreOffset != -1)
{
docStoreSegment = input.ReadString();
docStoreIsCompoundFile = input.ReadByte() == SegmentInfo.YES;
attributes[Lucene3xSegmentInfoFormat.DS_OFFSET_KEY] = Convert.ToString(docStoreOffset);
attributes[Lucene3xSegmentInfoFormat.DS_NAME_KEY] = docStoreSegment;
attributes[Lucene3xSegmentInfoFormat.DS_COMPOUND_KEY] = Convert.ToString(docStoreIsCompoundFile);
}
else
{
docStoreSegment = name;
docStoreIsCompoundFile = false;
}
// pre-4.0 indexes write a byte if there is a single norms file
byte b = input.ReadByte();
//System.out.println("version=" + version + " name=" + name + " docCount=" + docCount + " delGen=" + delGen + " dso=" + docStoreOffset + " dss=" + docStoreSegment + " dssCFs=" + docStoreIsCompoundFile + " b=" + b + " format=" + format);
Debug.Assert(1 == b, "expected 1 but was: " + b + " format: " + format);
int numNormGen = input.ReadInt();
IDictionary<int, long> normGen;
if (numNormGen == SegmentInfo.NO)
{
normGen = null;
}
else
{
normGen = new Dictionary<int, long>();
for (int j = 0; j < numNormGen; j++)
{
normGen[j] = input.ReadLong();
}
}
bool isCompoundFile = input.ReadByte() == SegmentInfo.YES;
int delCount = input.ReadInt();
Debug.Assert(delCount <= docCount);
bool hasProx = input.ReadByte() == 1;
IDictionary<string, string> diagnostics = input.ReadStringStringMap();
if (format <= Lucene3xSegmentInfoFormat.FORMAT_HAS_VECTORS)
{
// NOTE: unused
int hasVectors = input.ReadByte();
}
// Replicate logic from 3.x's SegmentInfo.files():
ISet<string> files = new HashSet<string>();
if (isCompoundFile)
{
files.Add(IndexFileNames.SegmentFileName(name, "", IndexFileNames.COMPOUND_FILE_EXTENSION));
}
else
{
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xFieldInfosReader.FIELD_INFOS_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xPostingsFormat.FREQ_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xPostingsFormat.PROX_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xPostingsFormat.TERMS_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xPostingsFormat.TERMS_INDEX_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xNormsProducer.NORMS_EXTENSION));
}
if (docStoreOffset != -1)
{
if (docStoreIsCompoundFile)
{
files.Add(IndexFileNames.SegmentFileName(docStoreSegment, "", Lucene3xCodec.COMPOUND_FILE_STORE_EXTENSION));
}
else
{
files.Add(IndexFileNames.SegmentFileName(docStoreSegment, "", Lucene3xStoredFieldsReader.FIELDS_INDEX_EXTENSION));
files.Add(IndexFileNames.SegmentFileName(docStoreSegment, "", Lucene3xStoredFieldsReader.FIELDS_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(docStoreSegment, "", Lucene3xTermVectorsReader.VECTORS_INDEX_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(docStoreSegment, "", Lucene3xTermVectorsReader.VECTORS_FIELDS_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(docStoreSegment, "", Lucene3xTermVectorsReader.VECTORS_DOCUMENTS_EXTENSION));
}
}
else if (!isCompoundFile)
{
files.Add(IndexFileNames.SegmentFileName(name, "", Lucene3xStoredFieldsReader.FIELDS_INDEX_EXTENSION));
files.Add(IndexFileNames.SegmentFileName(name, "", Lucene3xStoredFieldsReader.FIELDS_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xTermVectorsReader.VECTORS_INDEX_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xTermVectorsReader.VECTORS_FIELDS_EXTENSION));
AddIfExists(dir, files, IndexFileNames.SegmentFileName(name, "", Lucene3xTermVectorsReader.VECTORS_DOCUMENTS_EXTENSION));
}
// parse the normgen stuff and shove it into attributes
if (normGen != null)
{
attributes[Lucene3xSegmentInfoFormat.NORMGEN_KEY] = Convert.ToString(numNormGen);
foreach (KeyValuePair<int, long> ent in normGen)
{
long gen = ent.Value;
if (gen >= SegmentInfo.YES)
{
// Definitely a separate norm file, with generation:
files.Add(IndexFileNames.FileNameFromGeneration(name, "s" + ent.Key, gen));
attributes[Lucene3xSegmentInfoFormat.NORMGEN_PREFIX + ent.Key] = Convert.ToString(gen);
}
else if (gen == SegmentInfo.NO)
{
// No separate norm
}
else
{
// We should have already hit indexformat too old exception
Debug.Assert(false);
}
}
}
SegmentInfo info = new SegmentInfo(dir, version, name, docCount, isCompoundFile, null, diagnostics, CollectionsHelper.UnmodifiableMap(attributes));
info.Files = files;
SegmentCommitInfo infoPerCommit = new SegmentCommitInfo(info, delCount, delGen, -1);
return infoPerCommit;
}
private SegmentInfo ReadUpgradedSegmentInfo(string name, Directory dir, IndexInput input)
{
CodecUtil.CheckHeader(input, Lucene3xSegmentInfoFormat.UPGRADED_SI_CODEC_NAME, Lucene3xSegmentInfoFormat.UPGRADED_SI_VERSION_START, Lucene3xSegmentInfoFormat.UPGRADED_SI_VERSION_CURRENT);
string version = input.ReadString();
int docCount = input.ReadInt();
IDictionary<string, string> attributes = input.ReadStringStringMap();
bool isCompoundFile = input.ReadByte() == SegmentInfo.YES;
IDictionary<string, string> diagnostics = input.ReadStringStringMap();
ISet<string> files = input.ReadStringSet();
SegmentInfo info = new SegmentInfo(dir, version, name, docCount, isCompoundFile, null, diagnostics, CollectionsHelper.UnmodifiableMap(attributes));
info.Files = files;
return info;
}
private NumericDocValues LoadDoubleField(FieldInfo field, IndexInput input)
{
CodecUtil.CheckHeader(input, Lucene40DocValuesFormat.FLOATS_CODEC_NAME, Lucene40DocValuesFormat.FLOATS_VERSION_START, Lucene40DocValuesFormat.FLOATS_VERSION_CURRENT);
int valueSize = input.ReadInt();
if (valueSize != 8)
{
throw new CorruptIndexException("invalid valueSize: " + valueSize);
}
int maxDoc = State.SegmentInfo.DocCount;
long[] values = new long[maxDoc];
for (int i = 0; i < values.Length; i++)
{
values[i] = input.ReadLong();
}
RamBytesUsed_Renamed.AddAndGet(RamUsageEstimator.SizeOf(values));
return new NumericDocValuesAnonymousInnerClassHelper8(this, values);
}
private SortedDocValues LoadBytesFixedSorted(FieldInfo field, IndexInput data, IndexInput index)
{
CodecUtil.CheckHeader(data, Lucene40DocValuesFormat.BYTES_FIXED_SORTED_CODEC_NAME_DAT, Lucene40DocValuesFormat.BYTES_FIXED_SORTED_VERSION_START, Lucene40DocValuesFormat.BYTES_FIXED_SORTED_VERSION_CURRENT);
CodecUtil.CheckHeader(index, Lucene40DocValuesFormat.BYTES_FIXED_SORTED_CODEC_NAME_IDX, Lucene40DocValuesFormat.BYTES_FIXED_SORTED_VERSION_START, Lucene40DocValuesFormat.BYTES_FIXED_SORTED_VERSION_CURRENT);
int fixedLength = data.ReadInt();
int valueCount = index.ReadInt();
PagedBytes bytes = new PagedBytes(16);
bytes.Copy(data, fixedLength * (long)valueCount);
PagedBytes.Reader bytesReader = bytes.Freeze(true);
PackedInts.Reader reader = PackedInts.GetReader(index);
RamBytesUsed_Renamed.AddAndGet(bytes.RamBytesUsed() + reader.RamBytesUsed());
return CorrectBuggyOrds(new SortedDocValuesAnonymousInnerClassHelper(this, fixedLength, valueCount, bytesReader, reader));
}
/// <summary>Returns a <see cref="Status" /> instance detailing
/// the state of the index.
///
/// </summary>
/// <param name="onlySegments">list of specific segment names to check
///
/// <p/>As this method checks every byte in the specified
/// segments, on a large index it can take quite a long
/// time to run.
///
/// <p/><b>WARNING</b>: make sure
/// you only call this when the index is not opened by any
/// writer.
/// </param>
public virtual Status CheckIndex_Renamed_Method(List <string> onlySegments)
{
System.Globalization.NumberFormatInfo nf = System.Globalization.CultureInfo.CurrentCulture.NumberFormat;
SegmentInfos sis = new SegmentInfos();
Status result = new Status();
result.dir = dir;
try
{
sis.Read(dir);
}
catch (System.Exception t)
{
Msg("ERROR: could not read any segments file in directory");
result.missingSegments = true;
if (infoStream != null)
{
infoStream.WriteLine(t.StackTrace);
}
return(result);
}
int numSegments = sis.Count;
var segmentsFileName = sis.GetCurrentSegmentFileName();
IndexInput input = null;
try
{
input = dir.OpenInput(segmentsFileName);
}
catch (System.Exception t)
{
Msg("ERROR: could not open segments file in directory");
if (infoStream != null)
{
infoStream.WriteLine(t.StackTrace);
}
result.cantOpenSegments = true;
return(result);
}
int format = 0;
try
{
format = input.ReadInt();
}
catch (System.Exception t)
{
Msg("ERROR: could not read segment file version in directory");
if (infoStream != null)
{
infoStream.WriteLine(t.StackTrace);
}
result.missingSegmentVersion = true;
return(result);
}
finally
{
if (input != null)
{
input.Close();
}
}
System.String sFormat = "";
bool skip = false;
if (format == SegmentInfos.FORMAT)
{
sFormat = "FORMAT [Lucene Pre-2.1]";
}
if (format == SegmentInfos.FORMAT_LOCKLESS)
{
sFormat = "FORMAT_LOCKLESS [Lucene 2.1]";
}
else if (format == SegmentInfos.FORMAT_SINGLE_NORM_FILE)
{
sFormat = "FORMAT_SINGLE_NORM_FILE [Lucene 2.2]";
}
else if (format == SegmentInfos.FORMAT_SHARED_DOC_STORE)
{
sFormat = "FORMAT_SHARED_DOC_STORE [Lucene 2.3]";
}
else
{
if (format == SegmentInfos.FORMAT_CHECKSUM)
{
sFormat = "FORMAT_CHECKSUM [Lucene 2.4]";
}
else if (format == SegmentInfos.FORMAT_DEL_COUNT)
{
sFormat = "FORMAT_DEL_COUNT [Lucene 2.4]";
}
else if (format == SegmentInfos.FORMAT_HAS_PROX)
{
sFormat = "FORMAT_HAS_PROX [Lucene 2.4]";
}
else if (format == SegmentInfos.FORMAT_USER_DATA)
{
sFormat = "FORMAT_USER_DATA [Lucene 2.9]";
}
else if (format == SegmentInfos.FORMAT_DIAGNOSTICS)
{
sFormat = "FORMAT_DIAGNOSTICS [Lucene 2.9]";
}
else if (format < SegmentInfos.CURRENT_FORMAT)
{
sFormat = "int=" + format + " [newer version of Lucene than this tool]";
skip = true;
}
else
{
sFormat = format + " [Lucene 1.3 or prior]";
}
}
result.segmentsFileName = segmentsFileName;
result.numSegments = numSegments;
result.segmentFormat = sFormat;
result.userData = sis.UserData;
System.String userDataString;
if (sis.UserData.Count > 0)
{
userDataString = " userData=" + CollectionsHelper.CollectionToString(sis.UserData);
}
else
{
userDataString = "";
}
Msg("Segments file=" + segmentsFileName + " numSegments=" + numSegments + " version=" + sFormat + userDataString);
if (onlySegments != null)
{
result.partial = true;
if (infoStream != null)
{
infoStream.Write("\nChecking only these segments:");
}
foreach (string s in onlySegments)
{
if (infoStream != null)
{
infoStream.Write(" " + s);
}
}
result.segmentsChecked.AddRange(onlySegments);
Msg(":");
}
if (skip)
{
Msg("\nERROR: this index appears to be created by a newer version of Lucene than this tool was compiled on; please re-compile this tool on the matching version of Lucene; exiting");
result.toolOutOfDate = true;
return(result);
}
result.newSegments = (SegmentInfos)sis.Clone();
result.newSegments.Clear();
for (int i = 0; i < numSegments; i++)
{
SegmentInfo info = sis.Info(i);
if (onlySegments != null && !onlySegments.Contains(info.name))
{
continue;
}
var segInfoStat = new Status.SegmentInfoStatus();
result.segmentInfos.Add(segInfoStat);
Msg(" " + (1 + i) + " of " + numSegments + ": name=" + info.name + " docCount=" + info.docCount);
segInfoStat.name = info.name;
segInfoStat.docCount = info.docCount;
int toLoseDocCount = info.docCount;
SegmentReader reader = null;
try
{
Msg(" compound=" + info.GetUseCompoundFile());
segInfoStat.compound = info.GetUseCompoundFile();
Msg(" hasProx=" + info.HasProx);
segInfoStat.hasProx = info.HasProx;
Msg(" numFiles=" + info.Files().Count);
segInfoStat.numFiles = info.Files().Count;
Msg(System.String.Format(nf, " size (MB)={0:f}", new System.Object[] { (info.SizeInBytes() / (1024.0 * 1024.0)) }));
segInfoStat.sizeMB = info.SizeInBytes() / (1024.0 * 1024.0);
IDictionary <string, string> diagnostics = info.Diagnostics;
segInfoStat.diagnostics = diagnostics;
if (diagnostics.Count > 0)
{
Msg(" diagnostics = " + CollectionsHelper.CollectionToString(diagnostics));
}
int docStoreOffset = info.DocStoreOffset;
if (docStoreOffset != -1)
{
Msg(" docStoreOffset=" + docStoreOffset);
segInfoStat.docStoreOffset = docStoreOffset;
Msg(" docStoreSegment=" + info.DocStoreSegment);
segInfoStat.docStoreSegment = info.DocStoreSegment;
Msg(" docStoreIsCompoundFile=" + info.DocStoreIsCompoundFile);
segInfoStat.docStoreCompoundFile = info.DocStoreIsCompoundFile;
}
System.String delFileName = info.GetDelFileName();
if (delFileName == null)
{
Msg(" no deletions");
segInfoStat.hasDeletions = false;
}
else
{
Msg(" has deletions [delFileName=" + delFileName + "]");
segInfoStat.hasDeletions = true;
segInfoStat.deletionsFileName = delFileName;
}
if (infoStream != null)
{
infoStream.Write(" test: open reader.........");
}
reader = SegmentReader.Get(true, info, IndexReader.DEFAULT_TERMS_INDEX_DIVISOR);
segInfoStat.openReaderPassed = true;
int numDocs = reader.NumDocs();
toLoseDocCount = numDocs;
if (reader.HasDeletions)
{
if (reader.deletedDocs.Count() != info.GetDelCount())
{
throw new System.SystemException("delete count mismatch: info=" + info.GetDelCount() + " vs deletedDocs.count()=" + reader.deletedDocs.Count());
}
if (reader.deletedDocs.Count() > reader.MaxDoc)
{
throw new System.SystemException("too many deleted docs: MaxDoc=" + reader.MaxDoc + " vs deletedDocs.count()=" + reader.deletedDocs.Count());
}
if (info.docCount - numDocs != info.GetDelCount())
{
throw new System.SystemException("delete count mismatch: info=" + info.GetDelCount() + " vs reader=" + (info.docCount - numDocs));
}
segInfoStat.numDeleted = info.docCount - numDocs;
Msg("OK [" + (segInfoStat.numDeleted) + " deleted docs]");
}
else
{
if (info.GetDelCount() != 0)
{
throw new System.SystemException("delete count mismatch: info=" + info.GetDelCount() + " vs reader=" + (info.docCount - numDocs));
}
Msg("OK");
}
if (reader.MaxDoc != info.docCount)
{
throw new System.SystemException("SegmentReader.MaxDoc " + reader.MaxDoc + " != SegmentInfos.docCount " + info.docCount);
}
// Test getFieldNames()
if (infoStream != null)
{
infoStream.Write(" test: fields..............");
}
ICollection <string> fieldNames = reader.GetFieldNames(IndexReader.FieldOption.ALL);
Msg("OK [" + fieldNames.Count + " fields]");
segInfoStat.numFields = fieldNames.Count;
// Test Field Norms
segInfoStat.fieldNormStatus = TestFieldNorms(fieldNames, reader);
// Test the Term Index
segInfoStat.termIndexStatus = TestTermIndex(info, reader);
// Test Stored Fields
segInfoStat.storedFieldStatus = TestStoredFields(info, reader, nf);
// Test Term Vectors
segInfoStat.termVectorStatus = TestTermVectors(info, reader, nf);
// Rethrow the first exception we encountered
// This will cause stats for failed segments to be incremented properly
if (segInfoStat.fieldNormStatus.error != null)
{
throw new SystemException("Field Norm test failed");
}
else if (segInfoStat.termIndexStatus.error != null)
{
throw new SystemException("Term Index test failed");
}
else if (segInfoStat.storedFieldStatus.error != null)
{
throw new SystemException("Stored Field test failed");
}
else if (segInfoStat.termVectorStatus.error != null)
{
throw new System.SystemException("Term Vector test failed");
}
Msg("");
}
catch (System.Exception t)
{
Msg("FAILED");
const string comment = "fixIndex() would remove reference to this segment";
Msg(" WARNING: " + comment + "; full exception:");
if (infoStream != null)
{
infoStream.WriteLine(t.StackTrace);
}
Msg("");
result.totLoseDocCount += toLoseDocCount;
result.numBadSegments++;
continue;
}
finally
{
if (reader != null)
{
reader.Close();
}
}
// Keeper
result.newSegments.Add((SegmentInfo)info.Clone());
}
if (0 == result.numBadSegments)
{
result.clean = true;
Msg("No problems were detected with this index.\n");
}
else
{
Msg("WARNING: " + result.numBadSegments + " broken segments (containing " + result.totLoseDocCount + " documents) detected");
}
return(result);
}
private NumericDocValues LoadByteField(FieldInfo field, IndexInput input)
{
CodecUtil.CheckHeader(input, Lucene40DocValuesFormat.INTS_CODEC_NAME, Lucene40DocValuesFormat.INTS_VERSION_START, Lucene40DocValuesFormat.INTS_VERSION_CURRENT);
int valueSize = input.ReadInt();
if (valueSize != 1)
{
throw new CorruptIndexException("invalid valueSize: " + valueSize);
}
int maxDoc = State.SegmentInfo.DocCount;
var values = new byte[maxDoc];
input.ReadBytes(values, 0, values.Length);
RamBytesUsed_Renamed.AddAndGet(RamUsageEstimator.SizeOf(values));
return new NumericDocValuesAnonymousInnerClassHelper3(this, values);
}
public Lucene3xStoredFieldsReader(Directory d, SegmentInfo si, FieldInfos fn, IOContext context)
{
string segment = Lucene3xSegmentInfoFormat.GetDocStoreSegment(si);
int docStoreOffset = Lucene3xSegmentInfoFormat.GetDocStoreOffset(si);
int size = si.DocCount;
bool success = false;
FieldInfos = fn;
try
{
if (docStoreOffset != -1 && Lucene3xSegmentInfoFormat.GetDocStoreIsCompoundFile(si))
{
d = StoreCFSReader = new CompoundFileDirectory(si.Dir, IndexFileNames.SegmentFileName(segment, "", Lucene3xCodec.COMPOUND_FILE_STORE_EXTENSION), context, false);
}
else
{
StoreCFSReader = null;
}
FieldsStream = d.OpenInput(IndexFileNames.SegmentFileName(segment, "", FIELDS_EXTENSION), context);
string indexStreamFN = IndexFileNames.SegmentFileName(segment, "", FIELDS_INDEX_EXTENSION);
IndexStream = d.OpenInput(indexStreamFN, context);
Format = IndexStream.ReadInt();
if (Format < FORMAT_MINIMUM)
{
throw new IndexFormatTooOldException(IndexStream, Format, FORMAT_MINIMUM, FORMAT_CURRENT);
}
if (Format > FORMAT_CURRENT)
{
throw new IndexFormatTooNewException(IndexStream, Format, FORMAT_MINIMUM, FORMAT_CURRENT);
}
long indexSize = IndexStream.Length() - FORMAT_SIZE;
if (docStoreOffset != -1)
{
// We read only a slice out of this shared fields file
this.DocStoreOffset = docStoreOffset;
this.Size = size;
// Verify the file is long enough to hold all of our
// docs
Debug.Assert(((int)(indexSize / 8)) >= size + this.DocStoreOffset, "indexSize=" + indexSize + " size=" + size + " docStoreOffset=" + docStoreOffset);
}
else
{
this.DocStoreOffset = 0;
this.Size = (int)(indexSize >> 3);
// Verify two sources of "maxDoc" agree:
if (this.Size != si.DocCount)
{
throw new CorruptIndexException("doc counts differ for segment " + segment + ": fieldsReader shows " + this.Size + " 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();
} // keep our original exception
catch (Exception t)
{
}
}
}
}
internal FieldsReader(Directory d, System.String segment, FieldInfos fn, int readBufferSize, int docStoreOffset, int size)
{
bool success = false;
isOriginal = true;
try
{
fieldInfos = fn;
cloneableFieldsStream = d.OpenInput(segment + "." + IndexFileNames.FIELDS_EXTENSION, readBufferSize);
cloneableIndexStream = d.OpenInput(segment + "." + IndexFileNames.FIELDS_INDEX_EXTENSION, readBufferSize);
// First version of fdx did not include a format
// header, but, the first int will always be 0 in that
// case
int firstInt = cloneableIndexStream.ReadInt();
format = firstInt == 0 ? 0 : firstInt;
if (format > FieldsWriter.FORMAT_CURRENT)
{
throw new CorruptIndexException("Incompatible format version: " + format + " expected " + FieldsWriter.FORMAT_CURRENT + " or lower");
}
formatSize = format > FieldsWriter.FORMAT ? 4 : 0;
if (format < FieldsWriter.FORMAT_VERSION_UTF8_LENGTH_IN_BYTES)
{
cloneableFieldsStream.SetModifiedUTF8StringsMode();
}
fieldsStream = (IndexInput)cloneableFieldsStream.Clone();
long indexSize = cloneableIndexStream.Length() - formatSize;
if (docStoreOffset != -1)
{
// We read only a slice out of this shared fields file
this.docStoreOffset = docStoreOffset;
this.size = size;
// Verify the file is long enough to hold all of our
// docs
System.Diagnostics.Debug.Assert(((int)(indexSize / 8)) >= size + this.docStoreOffset, "indexSize=" + indexSize + " size=" + size + " docStoreOffset=" + docStoreOffset);
}
else
{
this.docStoreOffset = 0;
this.size = (int)(indexSize >> 3);
}
indexStream = (IndexInput)cloneableIndexStream.Clone();
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)
{
Dispose();
}
}
}
public FixedGapTermsIndexReader(Directory dir, FieldInfos fieldInfos, String segment, int indexDivisor,
IComparer <BytesRef> termComp, String segmentSuffix, IOContext context)
{
_termComp = termComp;
Debug.Assert(indexDivisor == -1 || indexDivisor > 0);
_input =
dir.OpenInput(
IndexFileNames.SegmentFileName(segment, segmentSuffix,
FixedGapTermsIndexWriter.TERMS_INDEX_EXTENSION),
context);
var success = false;
try
{
_version = ReadHeader(_input);
if (_version >= FixedGapTermsIndexWriter.VERSION_CHECKSUM)
{
CodecUtil.ChecksumEntireFile(_input);
}
indexInterval = _input.ReadInt();
if (indexInterval < 1)
{
throw new CorruptIndexException(String.Format("Invalid indexInterval: {0}, Resource: {1}",
indexInterval, _input));
}
_indexDivisor = indexDivisor;
if (indexDivisor < 0)
{
_totalIndexInterval = indexInterval;
}
else
{
// In case terms index gets loaded, later, on demand
_totalIndexInterval = indexInterval * indexDivisor;
}
Debug.Assert(_totalIndexInterval > 0);
SeekDir(_input, _dirOffset);
// Read directory
int numFields = _input.ReadVInt();
if (numFields < 0)
{
throw new CorruptIndexException(String.Format("Invalid numFields: {0}, Resource: {1}", numFields,
_input));
}
for (int i = 0; i < numFields; i++)
{
int field = _input.ReadVInt();
int numIndexTerms = _input.ReadVInt();
if (numIndexTerms < 0)
{
throw new CorruptIndexException(String.Format("Invalid numIndexTerms: {0}, Resource: {1}",
numIndexTerms,
_input));
}
long termsStart = _input.ReadVLong();
long indexStart = _input.ReadVLong();
long packedIndexStart = _input.ReadVLong();
long packedOffsetsStart = _input.ReadVLong();
if (packedIndexStart < indexStart)
{
throw new CorruptIndexException(
String.Format(
"Invalid packedIndexStart: {0}, IndexStart: {1}, NumIndexTerms: {2}, Resource: {3}",
packedIndexStart,
indexStart, numIndexTerms, _input));
}
FieldInfo fieldInfo = fieldInfos.FieldInfo(field);
try
{
_fields.Add(fieldInfo,
new FieldIndexData(numIndexTerms, indexStart, termsStart, packedIndexStart,
packedOffsetsStart, this));
}
catch (ArgumentException)
{
throw new CorruptIndexException(String.Format("Duplicate field: {0}, Resource {1}",
fieldInfo.Name,
_input));
}
}
success = true;
}
finally
{
if (!success)
{
IOUtils.CloseWhileHandlingException(_input);
}
if (indexDivisor > 0)
{
_input.Dispose();
_input = null;
if (success)
{
_indexLoaded = true;
}
_termBytesReader = _termBytes.Freeze(true);
}
}
}
public override void Init(IndexInput termsIn)
{
// Make sure we are talking to the matching past writer
CodecUtil.CheckHeader(termsIn, SepPostingsWriter.CODEC, SepPostingsWriter.VERSION_START,
SepPostingsWriter.VERSION_START);
_skipInterval = termsIn.ReadInt();
_maxSkipLevels = termsIn.ReadInt();
_skipMinimum = termsIn.ReadInt();
}
/// <summary>
/// Read as a bit set </summary>
private void ReadBits(IndexInput input)
{
Count_Renamed = input.ReadInt(); // read count
Bits = new byte[GetNumBytes(Size_Renamed)]; // allocate bits
input.ReadBytes(Bits, 0, Bits.Length);
}