public void Read(IndexInput input, FieldInfos fieldInfos)
{
this.term = null; // invalidate cache
int start = input.ReadVInt();
int length = input.ReadVInt();
int totalLength = start + length;
if (preUTF8Strings)
{
text.SetLength(totalLength);
input.ReadChars(text.result, start, length);
}
else
{
if (dirty)
{
// Fully convert all bytes since bytes is dirty
UnicodeUtil.UTF16toUTF8(text.result, 0, text.length, bytes);
bytes.SetLength(totalLength);
input.ReadBytes(bytes.result, start, length);
UnicodeUtil.UTF8toUTF16(bytes.result, 0, totalLength, text);
dirty = false;
}
else
{
// Incrementally convert only the UTF8 bytes that are new:
bytes.SetLength(totalLength);
input.ReadBytes(bytes.result, start, length);
UnicodeUtil.UTF8toUTF16(bytes.result, start, length, text);
}
}
this.field = fieldInfos.FieldName(input.ReadVInt());
}
private int CopyFieldsNoDeletions(FieldSelector fieldSelectorMerge, FieldsWriter fieldsWriter, IndexReader reader, FieldsReader matchingFieldsReader)
{
int maxDoc = reader.MaxDoc();
int docCount = 0;
if (matchingFieldsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
while (docCount < maxDoc)
{
int len = System.Math.Min(MAX_RAW_MERGE_DOCS, maxDoc - docCount);
IndexInput stream = matchingFieldsReader.RawDocs(rawDocLengths, docCount, len);
fieldsWriter.AddRawDocuments(stream, rawDocLengths, len);
docCount += len;
checkAbort.Work(300 * len);
}
}
else
{
for (; docCount < maxDoc; docCount++)
{
// NOTE: it's very important to first assign to doc then pass it to
// termVectorsWriter.addAllDocVectors; see LUCENE-1282
Document doc = reader.Document(docCount, fieldSelectorMerge);
fieldsWriter.AddDocument(doc);
checkAbort.Work(300);
}
}
return(docCount);
}
protected internal override int ReadSkipData(int level, IndexInput skipStream)
{
int delta;
if (currentFieldStoresPayloads)
{
// the current field stores payloads.
// if the doc delta is odd then we have
// to read the current payload length
// because it differs from the length of the
// previous payload
delta = skipStream.ReadVInt();
if ((delta & 1) != 0)
{
payloadLength[level] = skipStream.ReadVInt();
}
delta = SupportClass.Number.URShift(delta, 1);
}
else
{
delta = skipStream.ReadVInt();
}
freqPointer[level] += skipStream.ReadVInt();
proxPointer[level] += skipStream.ReadVInt();
return(delta);
}
public void Read(IndexInput input, FieldInfos fieldInfos)
{
this.term = null; // invalidate cache
int start = input.ReadVInt();
int length = input.ReadVInt();
int totalLength = start + length;
if (preUTF8Strings)
{
text.SetLength(totalLength);
input.ReadChars(text.result, start, length);
}
else
{
if (dirty)
{
// Fully convert all bytes since bytes is dirty
UnicodeUtil.UTF16toUTF8(text.result, 0, text.length, bytes);
bytes.SetLength(totalLength);
input.ReadBytes(bytes.result, start, length);
UnicodeUtil.UTF8toUTF16(bytes.result, 0, totalLength, text);
dirty = false;
}
else
{
// Incrementally convert only the UTF8 bytes that are new:
bytes.SetLength(totalLength);
input.ReadBytes(bytes.result, start, length);
UnicodeUtil.UTF8toUTF16(bytes.result, start, length, text);
}
}
this.field = fieldInfos.FieldName(input.ReadVInt());
}
private int CopyFieldsWithDeletions(FieldSelector fieldSelectorMerge, FieldsWriter fieldsWriter, IndexReader reader, FieldsReader matchingFieldsReader)
{
int docCount = 0;
int maxDoc = reader.MaxDoc();
if (matchingFieldsReader != null)
{
// We can bulk-copy because the fieldInfos are "congruent"
for (int j = 0; j < maxDoc;)
{
if (reader.IsDeleted(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 (reader.IsDeleted(j))
{
j++;
break;
}
}while (numDocs < MAX_RAW_MERGE_DOCS);
IndexInput stream = matchingFieldsReader.RawDocs(rawDocLengths, start, numDocs);
fieldsWriter.AddRawDocuments(stream, rawDocLengths, numDocs);
docCount += numDocs;
checkAbort.Work(300 * numDocs);
}
}
else
{
for (int j = 0; j < maxDoc; j++)
{
if (reader.IsDeleted(j))
{
// skip deleted docs
continue;
}
// NOTE: it's very important to first assign to doc then pass it to
// termVectorsWriter.addAllDocVectors; see LUCENE-1282
Document doc = reader.Document(j, fieldSelectorMerge);
fieldsWriter.AddDocument(doc);
docCount++;
checkAbort.Work(300);
}
}
return(docCount);
}
private int CheckValidFormat(IndexInput in_Renamed)
{
int format = in_Renamed.ReadInt();
if (format > FORMAT_CURRENT)
{
throw new CorruptIndexException("Incompatible format version: " + format + " expected " + FORMAT_CURRENT + " or less");
}
return(format);
}
public /*protected internal*/ SegmentTermDocs(SegmentReader parent)
{
this.parent = parent;
this.freqStream = (IndexInput)parent.core.freqStream.Clone();
lock (parent)
{
this.deletedDocs = parent.deletedDocs;
}
this.skipInterval = parent.core.GetTermsReader().GetSkipInterval();
this.maxSkipLevels = parent.core.GetTermsReader().GetMaxSkipLevels();
}
private IndexInput GetFieldStream()
{
IndexInput localFieldsStream = (IndexInput)Enclosing_Instance.fieldsStreamTL.Get();
if (localFieldsStream == null)
{
localFieldsStream = (IndexInput)Enclosing_Instance.cloneableFieldsStream.Clone();
Enclosing_Instance.fieldsStreamTL.Set(localFieldsStream);
}
return(localFieldsStream);
}
public /*protected internal*/ SegmentTermDocs(SegmentReader parent)
{
this.parent = parent;
this.freqStream = (IndexInput) parent.core.freqStream.Clone();
lock (parent)
{
this.deletedDocs = parent.deletedDocs;
}
this.skipInterval = parent.core.GetTermsReader().GetSkipInterval();
this.maxSkipLevels = parent.core.GetTermsReader().GetMaxSkipLevels();
}
public CompoundFileReader(Directory dir, System.String name, int readBufferSize)
{
directory = dir;
fileName = name;
this.readBufferSize = readBufferSize;
bool success = false;
try
{
stream = dir.OpenInput(name, readBufferSize);
// read the directory and init files
int count = stream.ReadVInt();
FileEntry entry = null;
for (int i = 0; i < count; i++)
{
long offset = stream.ReadLong();
System.String id = stream.ReadString();
if (entry != null)
{
// set length of the previous entry
entry.length = offset - entry.offset;
}
entry = new FileEntry();
entry.offset = offset;
entries[id] = entry;
}
// set the length of the final entry
if (entry != null)
{
entry.length = stream.Length() - entry.offset;
}
success = true;
}
finally
{
if (!success && (stream != null))
{
try
{
stream.Close();
}
catch (System.IO.IOException e)
{
}
}
}
}
public CompoundFileReader(Directory dir, System.String name, int readBufferSize)
{
directory = dir;
fileName = name;
this.readBufferSize = readBufferSize;
bool success = false;
try
{
stream = dir.OpenInput(name, readBufferSize);
// read the directory and init files
int count = stream.ReadVInt();
FileEntry entry = null;
for (int i = 0; i < count; i++)
{
long offset = stream.ReadLong();
System.String id = stream.ReadString();
if (entry != null)
{
// set length of the previous entry
entry.length = offset - entry.offset;
}
entry = new FileEntry();
entry.offset = offset;
entries[id] = entry;
}
// set the length of the final entry
if (entry != null)
{
entry.length = stream.Length() - entry.offset;
}
success = true;
}
finally
{
if (!success && (stream != null))
{
try
{
stream.Close();
}
catch (System.IO.IOException e)
{
}
}
}
}
public override byte[] GetBinaryValue(byte[] result)
{
Enclosing_Instance.EnsureOpen();
if (isBinary)
{
if (fieldsData == null)
{
// Allocate new buffer if result is null or too small
byte[] b;
if (result == null || result.Length < toRead)
{
b = new byte[toRead];
}
else
{
b = result;
}
IndexInput localFieldsStream = GetFieldStream();
// Throw this IOException since IndexReader.document does so anyway, so probably not that big of a change for people
// since they are already handling this exception when getting the document
try
{
localFieldsStream.Seek(pointer);
localFieldsStream.ReadBytes(b, 0, toRead);
if (isCompressed == true)
{
fieldsData = Enclosing_Instance.Uncompress(b);
}
else
{
fieldsData = b;
}
}
catch (System.IO.IOException e)
{
throw new FieldReaderException(e);
}
binaryOffset = 0;
binaryLength = toRead;
}
return((byte[])fieldsData);
}
else
{
return(null);
}
}
/// <summary>Bulk write a contiguous series of documents. The
/// lengths array is the length (in bytes) of each raw
/// document. The stream IndexInput is the
/// fieldsStream from which we should bulk-copy all
/// bytes.
/// </summary>
internal void AddRawDocuments(IndexInput stream, int[] lengths, int numDocs)
{
long position = fieldsStream.GetFilePointer();
long start = position;
for (int i = 0; i < numDocs; i++)
{
indexStream.WriteLong(position);
position += lengths[i];
}
fieldsStream.CopyBytes(stream, position - start);
System.Diagnostics.Debug.Assert(fieldsStream.GetFilePointer() == position);
}
// Used only by clone
private FieldsReader(FieldInfos fieldInfos, int numTotalDocs, int size, int format, int formatSize, int docStoreOffset, IndexInput cloneableFieldsStream, IndexInput cloneableIndexStream)
{
this.fieldInfos = fieldInfos;
this.numTotalDocs = numTotalDocs;
this.size = size;
this.format = format;
this.formatSize = formatSize;
this.docStoreOffset = docStoreOffset;
this.cloneableFieldsStream = cloneableFieldsStream;
this.cloneableIndexStream = cloneableIndexStream;
fieldsStream = (IndexInput) cloneableFieldsStream.Clone();
indexStream = (IndexInput) cloneableIndexStream.Clone();
}
// Used only by clone
private FieldsReader(FieldInfos fieldInfos, int numTotalDocs, int size, int format, int formatSize, int docStoreOffset, IndexInput cloneableFieldsStream, IndexInput cloneableIndexStream)
{
this.fieldInfos = fieldInfos;
this.numTotalDocs = numTotalDocs;
this.size = size;
this.format = format;
this.formatSize = formatSize;
this.docStoreOffset = docStoreOffset;
this.cloneableFieldsStream = cloneableFieldsStream;
this.cloneableIndexStream = cloneableIndexStream;
fieldsStream = (IndexInput)cloneableFieldsStream.Clone();
indexStream = (IndexInput)cloneableIndexStream.Clone();
}
private void Read(IndexInput input, System.String fileName)
{
int firstInt = input.ReadVInt();
if (firstInt < 0)
{
// This is a real format
format = firstInt;
}
else
{
format = FORMAT_PRE;
}
if (format != FORMAT_PRE & format != FORMAT_START)
{
throw new CorruptIndexException("unrecognized format " + format + " in file \"" + fileName + "\"");
}
int size;
if (format == FORMAT_PRE)
{
size = firstInt;
}
else
{
size = input.ReadVInt(); //read in the size
}
for (int i = 0; i < size; i++)
{
System.String name = StringHelper.Intern(input.ReadString());
byte bits = input.ReadByte();
bool isIndexed = (bits & IS_INDEXED) != 0;
bool storeTermVector = (bits & STORE_TERMVECTOR) != 0;
bool storePositionsWithTermVector = (bits & STORE_POSITIONS_WITH_TERMVECTOR) != 0;
bool storeOffsetWithTermVector = (bits & STORE_OFFSET_WITH_TERMVECTOR) != 0;
bool omitNorms = (bits & OMIT_NORMS) != 0;
bool storePayloads = (bits & STORE_PAYLOADS) != 0;
bool omitTermFreqAndPositions = (bits & OMIT_TERM_FREQ_AND_POSITIONS) != 0;
AddInternal(name, isIndexed, storeTermVector, storePositionsWithTermVector, storeOffsetWithTermVector, omitNorms, storePayloads, omitTermFreqAndPositions);
}
if (input.GetFilePointer() != input.Length())
{
throw new CorruptIndexException("did not read all bytes from file \"" + fileName + "\": read " + input.GetFilePointer() + " vs size " + input.Length());
}
}
public override void Close()
{
lock (this)
{
if (stream == null)
{
throw new System.IO.IOException("Already closed");
}
entries.Clear();
stream.Close();
stream = null;
}
}
/// <summary>Copy the contents of the file with specified extension into the
/// provided output stream. Use the provided buffer for moving data
/// to reduce memory allocation.
/// </summary>
private void CopyFile(FileEntry source, IndexOutput os, byte[] buffer)
{
IndexInput is_Renamed = null;
try
{
long startPtr = os.GetFilePointer();
is_Renamed = directory.OpenInput(source.file);
long length = is_Renamed.Length();
long remainder = length;
int chunk = buffer.Length;
while (remainder > 0)
{
int len = (int)System.Math.Min(chunk, remainder);
is_Renamed.ReadBytes(buffer, 0, len, false);
os.WriteBytes(buffer, len);
remainder -= len;
if (checkAbort != null)
{
// Roughly every 2 MB we will check if
// it's time to abort
checkAbort.Work(80);
}
}
// Verify that remainder is 0
if (remainder != 0)
{
throw new System.IO.IOException("Non-zero remainder length after copying: " + remainder + " (id: " + source.file + ", length: " + length + ", buffer size: " + chunk + ")");
}
// Verify that the output length diff is equal to original file
long endPtr = os.GetFilePointer();
long diff = endPtr - startPtr;
if (diff != length)
{
throw new System.IO.IOException("Difference in the output file offsets " + diff + " does not match the original file length " + length);
}
}
finally
{
if (is_Renamed != null)
{
is_Renamed.Close();
}
}
}
/// <summary>read as a d-gaps list </summary>
private void ReadDgaps(IndexInput input)
{
size = input.ReadInt(); // (re)read size
count = input.ReadInt(); // read count
bits = new byte[(size >> 3) + 1]; // allocate bits
int last = 0;
int n = Count();
while (n > 0)
{
last += input.ReadVInt();
bits[last] = input.ReadByte();
n -= BYTE_COUNTS[bits[last] & 0xFF];
}
}
/// <summary>The value of the field as a String, or null. If null, the Reader value,
/// binary value, or TokenStream value is used. Exactly one of stringValue(),
/// readerValue(), binaryValue(), and tokenStreamValue() must be set.
/// </summary>
public override System.String StringValue()
{
Enclosing_Instance.EnsureOpen();
if (isBinary)
{
return(null);
}
else
{
if (fieldsData == null)
{
IndexInput localFieldsStream = GetFieldStream();
try
{
localFieldsStream.Seek(pointer);
if (isCompressed)
{
byte[] b = new byte[toRead];
localFieldsStream.ReadBytes(b, 0, b.Length);
fieldsData = System.Text.Encoding.GetEncoding("UTF-8").GetString(Enclosing_Instance.Uncompress(b));
}
else
{
if (Enclosing_Instance.format >= FieldsWriter.FORMAT_VERSION_UTF8_LENGTH_IN_BYTES)
{
byte[] bytes = new byte[toRead];
localFieldsStream.ReadBytes(bytes, 0, toRead);
fieldsData = System.Text.Encoding.GetEncoding("UTF-8").GetString(bytes);
}
else
{
//read in chars b/c we already know the length we need to read
char[] chars = new char[toRead];
localFieldsStream.ReadChars(chars, 0, toRead);
fieldsData = new System.String(chars);
}
}
}
catch (System.IO.IOException e)
{
throw new FieldReaderException(e);
}
}
return((System.String)fieldsData);
}
}
public MultiLevelSkipListReader(IndexInput skipStream, int maxSkipLevels, int skipInterval)
{
this.skipStream = new IndexInput[maxSkipLevels];
this.skipPointer = new long[maxSkipLevels];
this.childPointer = new long[maxSkipLevels];
this.numSkipped = new int[maxSkipLevels];
this.maxNumberOfSkipLevels = maxSkipLevels;
this.skipInterval = new int[maxSkipLevels];
this.skipStream[0] = skipStream;
this.inputIsBuffered = (skipStream is BufferedIndexInput);
this.skipInterval[0] = skipInterval;
for (int i = 1; i < maxSkipLevels; i++)
{
// cache skip intervals
this.skipInterval[i] = this.skipInterval[i - 1] * skipInterval;
}
skipDoc = new int[maxSkipLevels];
}
public MultiLevelSkipListReader(IndexInput skipStream, int maxSkipLevels, int skipInterval)
{
this.skipStream = new IndexInput[maxSkipLevels];
this.skipPointer = new long[maxSkipLevels];
this.childPointer = new long[maxSkipLevels];
this.numSkipped = new int[maxSkipLevels];
this.maxNumberOfSkipLevels = maxSkipLevels;
this.skipInterval = new int[maxSkipLevels];
this.skipStream[0] = skipStream;
this.inputIsBuffered = (skipStream is BufferedIndexInput);
this.skipInterval[0] = skipInterval;
for (int i = 1; i < maxSkipLevels; i++)
{
// cache skip intervals
this.skipInterval[i] = this.skipInterval[i - 1] * skipInterval;
}
skipDoc = new int[maxSkipLevels];
}
/// <summary> Construct a FieldInfos object using the directory and the name of the file
/// IndexInput
/// </summary>
/// <param name="d">The directory to open the IndexInput from
/// </param>
/// <param name="name">The name of the file to open the IndexInput from in the Directory
/// </param>
/// <throws> IOException </throws>
public /*internal*/ FieldInfos(Directory d, System.String name)
{
IndexInput input = d.OpenInput(name);
try
{
try
{
Read(input, name);
}
catch (System.IO.IOException ioe)
{
if (format == FORMAT_PRE)
{
// LUCENE-1623: FORMAT_PRE (before there was a
// format) may be 2.3.2 (pre-utf8) or 2.4.x (utf8)
// encoding; retry with input set to pre-utf8
input.Seek(0);
input.SetModifiedUTF8StringsMode();
byNumber.Clear();
byName.Clear();
try
{
Read(input, name);
}
catch (System.Exception t)
{
// Ignore any new exception & throw original IOE
throw ioe;
}
}
else
{
// The IOException cannot be caused by
// LUCENE-1623, so re-throw it
throw ioe;
}
}
}
finally
{
input.Close();
}
}
/// <summary>Constructs a bit vector from the file <code>name</code> in Directory
/// <code>d</code>, as written by the {@link #write} method.
/// </summary>
public BitVector(Directory d, System.String name)
{
IndexInput input = d.OpenInput(name);
try
{
size = input.ReadInt(); // read size
if (size == -1)
{
ReadDgaps(input);
}
else
{
ReadBits(input);
}
}
finally
{
input.Close();
}
}
// It is not always neccessary to move the prox pointer
// to a new document after the freq pointer has been moved.
// Consider for example a phrase query with two terms:
// the freq pointer for term 1 has to move to document x
// to answer the question if the term occurs in that document. But
// only if term 2 also matches document x, the positions have to be
// read to figure out if term 1 and term 2 appear next
// to each other in document x and thus satisfy the query.
// So we move the prox pointer lazily to the document
// as soon as positions are requested.
private void LazySkip()
{
if (proxStream == null)
{
// clone lazily
proxStream = (IndexInput)parent.core.proxStream.Clone();
}
// we might have to skip the current payload
// if it was not read yet
SkipPayload();
if (lazySkipPointer != -1)
{
proxStream.Seek(lazySkipPointer);
lazySkipPointer = -1;
}
if (lazySkipProxCount != 0)
{
SkipPositions(lazySkipProxCount);
lazySkipProxCount = 0;
}
}
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();
if (firstInt == 0)
format = 0;
else
format = firstInt;
if (format > FieldsWriter.FORMAT_CURRENT
/* extra support for Lucene 3.0 indexes: */ && format != FieldsWriter.FORMAT_LUCENE_3_0_NO_COMPRESSED_FIELDS
)
throw new CorruptIndexException("Incompatible format version: " + format + " expected " + FieldsWriter.FORMAT_CURRENT + " or lower");
if (format > FieldsWriter.FORMAT)
formatSize = 4;
else
formatSize = 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)
{
Close();
}
}
}
internal SkipBuffer(IndexInput input, int length)
{
data = new byte[length];
pointer = input.GetFilePointer();
input.ReadBytes(data, 0, length);
}
internal 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 = System.Int32.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 < TermInfosWriter.FORMAT_CURRENT)
{
throw new CorruptIndexException("Unknown format version:" + format + " expected " + TermInfosWriter.FORMAT_CURRENT + " or higher");
}
size = input.ReadLong(); // read the size
if (format == -1)
{
if (!isIndex)
{
indexInterval = input.ReadInt();
formatM1SkipInterval = input.ReadInt();
}
// switch off skipTo optimization for file format prior to 1.4rc2 in order to avoid a bug in
// skipTo implementation of these versions
skipInterval = System.Int32.MaxValue;
}
else
{
indexInterval = input.ReadInt();
skipInterval = input.ReadInt();
if (format <= TermInfosWriter.FORMAT)
{
// this new format introduces multi-level skipping
maxSkipLevels = input.ReadInt();
}
}
System.Diagnostics.Debug.Assert(indexInterval > 0, "indexInterval=" + indexInterval + " is negative; must be > 0");
System.Diagnostics.Debug.Assert(skipInterval > 0, "skipInterval=" + skipInterval + " is negative; must be > 0");
}
if (format > TermInfosWriter.FORMAT_VERSION_UTF8_LENGTH_IN_BYTES)
{
termBuffer.SetPreUTF8Strings();
scanBuffer.SetPreUTF8Strings();
prevBuffer.SetPreUTF8Strings();
}
}
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();
if (firstInt == 0)
{
format = 0;
}
else
{
format = firstInt;
}
if (format > FieldsWriter.FORMAT_CURRENT &&
/* extra support for Lucene 3.0 indexes: */ format != FieldsWriter.FORMAT_LUCENE_3_0_NO_COMPRESSED_FIELDS
)
{
throw new CorruptIndexException("Incompatible format version: " + format + " expected " + FieldsWriter.FORMAT_CURRENT + " or lower");
}
if (format > FieldsWriter.FORMAT)
{
formatSize = 4;
}
else
{
formatSize = 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)
{
Close();
}
}
}
internal CSIndexInput(IndexInput base_Renamed, long fileOffset, long length, int readBufferSize) : base(readBufferSize)
{
this.base_Renamed = (IndexInput)base_Renamed.Clone();
this.fileOffset = fileOffset;
this.length = length;
}
/// <summary> Construct a new SegmentInfo instance by reading a
/// previously saved SegmentInfo from input.
///
/// </summary>
/// <param name="dir">directory to load from
/// </param>
/// <param name="format">format of the segments info file
/// </param>
/// <param name="input">input handle to read segment info from
/// </param>
internal SegmentInfo(Directory dir, int format, IndexInput input)
{
this.dir = dir;
name = input.ReadString();
docCount = input.ReadInt();
if (format <= SegmentInfos.FORMAT_LOCKLESS)
{
delGen = input.ReadLong();
if (format <= SegmentInfos.FORMAT_SHARED_DOC_STORE)
{
docStoreOffset = input.ReadInt();
if (docStoreOffset != - 1)
{
docStoreSegment = input.ReadString();
docStoreIsCompoundFile = (1 == input.ReadByte());
}
else
{
docStoreSegment = name;
docStoreIsCompoundFile = false;
}
}
else
{
docStoreOffset = - 1;
docStoreSegment = name;
docStoreIsCompoundFile = false;
}
if (format <= SegmentInfos.FORMAT_SINGLE_NORM_FILE)
{
hasSingleNormFile = (1 == input.ReadByte());
}
else
{
hasSingleNormFile = false;
}
int numNormGen = input.ReadInt();
if (numNormGen == NO)
{
normGen = null;
}
else
{
normGen = new long[numNormGen];
for (int j = 0; j < numNormGen; j++)
{
normGen[j] = input.ReadLong();
}
}
isCompoundFile = (sbyte) input.ReadByte();
preLockless = (isCompoundFile == CHECK_DIR);
if (format <= SegmentInfos.FORMAT_DEL_COUNT)
{
delCount = input.ReadInt();
System.Diagnostics.Debug.Assert(delCount <= docCount);
}
else
delCount = - 1;
if (format <= SegmentInfos.FORMAT_HAS_PROX)
hasProx = input.ReadByte() == 1;
else
hasProx = true;
if (format <= SegmentInfos.FORMAT_DIAGNOSTICS)
{
diagnostics = input.ReadStringStringMap();
}
else
{
diagnostics = new System.Collections.Generic.Dictionary<string,string>();
}
}
else
{
delGen = CHECK_DIR;
normGen = null;
isCompoundFile = (sbyte) (CHECK_DIR);
preLockless = true;
hasSingleNormFile = false;
docStoreOffset = - 1;
docStoreIsCompoundFile = false;
docStoreSegment = null;
delCount = - 1;
hasProx = true;
diagnostics = new System.Collections.Generic.Dictionary<string,string>();
}
}
internal CoreReaders(SegmentReader origInstance, Directory dir, SegmentInfo si, int readBufferSize, int termsIndexDivisor)
{
segment = si.name;
this.readBufferSize = readBufferSize;
this.dir = dir;
bool success = false;
try
{
Directory dir0 = dir;
if (si.GetUseCompoundFile())
{
cfsReader = new CompoundFileReader(dir, segment + "." + IndexFileNames.COMPOUND_FILE_EXTENSION, readBufferSize);
dir0 = cfsReader;
}
cfsDir = dir0;
fieldInfos = new FieldInfos(cfsDir, segment + "." + IndexFileNames.FIELD_INFOS_EXTENSION);
this.termsIndexDivisor = termsIndexDivisor;
TermInfosReader reader = new TermInfosReader(cfsDir, segment, fieldInfos, readBufferSize, termsIndexDivisor);
if (termsIndexDivisor == - 1)
{
tisNoIndex = reader;
}
else
{
tis = reader;
tisNoIndex = null;
}
// make sure that all index files have been read or are kept open
// so that if an index update removes them we'll still have them
freqStream = cfsDir.OpenInput(segment + "." + IndexFileNames.FREQ_EXTENSION, readBufferSize);
if (fieldInfos.HasProx())
{
proxStream = cfsDir.OpenInput(segment + "." + IndexFileNames.PROX_EXTENSION, readBufferSize);
}
else
{
proxStream = null;
}
success = true;
}
finally
{
if (!success)
{
DecRef();
}
}
// Must assign this at the end -- if we hit an
// exception above core, we don't want to attempt to
// purge the FieldCache (will hit NPE because core is
// not assigned yet).
this.origInstance = origInstance;
}
internal CSIndexInput(IndexInput base_Renamed, long fileOffset, long length, int readBufferSize):base(readBufferSize)
{
this.base_Renamed = (IndexInput) base_Renamed.Clone();
this.fileOffset = fileOffset;
this.length = length;
}
public override void Close()
{
lock (this)
{
if (stream == null)
throw new System.IO.IOException("Already closed");
entries.Clear();
stream.Close();
stream = null;
}
}
/// <summary>read as a d-gaps list </summary>
private void ReadDgaps(IndexInput input)
{
size = input.ReadInt(); // (re)read size
count = input.ReadInt(); // read count
bits = new byte[(size >> 3) + 1]; // allocate bits
int last = 0;
int n = Count();
while (n > 0)
{
last += input.ReadVInt();
bits[last] = input.ReadByte();
n -= BYTE_COUNTS[bits[last] & 0xFF];
}
}
internal TermVectorsReader(Directory d, System.String segment, FieldInfos fieldInfos, int readBufferSize, int docStoreOffset, int size)
{
bool success = false;
try
{
if (d.FileExists(segment + "." + IndexFileNames.VECTORS_INDEX_EXTENSION))
{
tvx = d.OpenInput(segment + "." + IndexFileNames.VECTORS_INDEX_EXTENSION, readBufferSize);
format = CheckValidFormat(tvx);
tvd = d.OpenInput(segment + "." + IndexFileNames.VECTORS_DOCUMENTS_EXTENSION, readBufferSize);
int tvdFormat = CheckValidFormat(tvd);
tvf = d.OpenInput(segment + "." + IndexFileNames.VECTORS_FIELDS_EXTENSION, readBufferSize);
int tvfFormat = CheckValidFormat(tvf);
System.Diagnostics.Debug.Assert(format == tvdFormat);
System.Diagnostics.Debug.Assert(format == tvfFormat);
if (format >= FORMAT_VERSION2)
{
System.Diagnostics.Debug.Assert((tvx.Length() - FORMAT_SIZE) % 16 == 0);
numTotalDocs = (int)(tvx.Length() >> 4);
}
else
{
System.Diagnostics.Debug.Assert((tvx.Length() - FORMAT_SIZE) % 8 == 0);
numTotalDocs = (int)(tvx.Length() >> 3);
}
if (-1 == docStoreOffset)
{
this.docStoreOffset = 0;
this.size = numTotalDocs;
System.Diagnostics.Debug.Assert(size == 0 || numTotalDocs == size);
}
else
{
this.docStoreOffset = docStoreOffset;
this.size = size;
// Verify the file is long enough to hold all of our
// docs
System.Diagnostics.Debug.Assert(numTotalDocs >= size + docStoreOffset, "numTotalDocs=" + numTotalDocs + " size=" + size + " docStoreOffset=" + docStoreOffset);
}
}
else
{
// If all documents flushed in a segment had hit
// non-aborting exceptions, it's possible that
// FieldInfos.hasVectors returns true yet the term
// vector files don't exist.
format = 0;
}
this.fieldInfos = fieldInfos;
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)
{
Close();
}
}
}
internal SegmentTermPositions(SegmentReader p) : base(p)
{
this.proxStream = null; // the proxStream will be cloned lazily when nextPosition() is called for the first time
}
private int CheckValidFormat(IndexInput in_Renamed)
{
int format = in_Renamed.ReadInt();
if (format > FORMAT_CURRENT)
{
throw new CorruptIndexException("Incompatible format version: " + format + " expected " + FORMAT_CURRENT + " or less");
}
return format;
}
/// <summary>Read as a bit set </summary>
private void ReadBits(IndexInput input)
{
count = input.ReadInt(); // read count
bits = new byte[(size >> 3) + 1]; // allocate bits
input.ReadBytes(bits, 0, bits.Length);
}
private void OpenNorms(Directory cfsDir, int readBufferSize)
{
long nextNormSeek = SegmentMerger.NORMS_HEADER.Length; //skip header (header unused for now)
int maxDoc = MaxDoc();
for (int i = 0; i < core.fieldInfos.Size(); i++)
{
FieldInfo fi = core.fieldInfos.FieldInfo(i);
if (norms.Contains(fi.name))
{
// in case this SegmentReader is being re-opened, we might be able to
// reuse some norm instances and skip loading them here
continue;
}
if (fi.isIndexed && !fi.omitNorms)
{
Directory d = Directory();
System.String fileName = si.GetNormFileName(fi.number);
if (!si.HasSeparateNorms(fi.number))
{
d = cfsDir;
}
// singleNormFile means multiple norms share this file
bool singleNormFile = fileName.EndsWith("." + IndexFileNames.NORMS_EXTENSION);
IndexInput normInput = null;
long normSeek;
if (singleNormFile)
{
normSeek = nextNormSeek;
if (singleNormStream == null)
{
singleNormStream = d.OpenInput(fileName, readBufferSize);
singleNormRef = new Ref();
}
else
{
singleNormRef.IncRef();
}
// All norms in the .nrm file can share a single IndexInput since
// they are only used in a synchronized context.
// If this were to change in the future, a clone could be done here.
normInput = singleNormStream;
}
else
{
normSeek = 0;
normInput = d.OpenInput(fileName);
}
norms[fi.name] = new Norm(this, normInput, fi.number, normSeek);
nextNormSeek += maxDoc; // increment also if some norms are separate
}
}
}
internal SkipBuffer(IndexInput input, int length)
{
data = new byte[length];
pointer = input.GetFilePointer();
input.ReadBytes(data, 0, length);
}
public Norm(SegmentReader enclosingInstance, IndexInput in_Renamed, int number, long normSeek)
{
InitBlock(enclosingInstance);
this.in_Renamed = in_Renamed;
this.number = number;
this.normSeek = normSeek;
}
internal CSIndexInput(IndexInput base_Renamed, long fileOffset, long length) : this(base_Renamed, fileOffset, length, BufferedIndexInput.BUFFER_SIZE)
{
}
private void CloseInput()
{
if (in_Renamed != null)
{
if (in_Renamed != Enclosing_Instance.singleNormStream)
{
// It's private to us -- just close it
in_Renamed.Close();
}
else
{
// We are sharing this with others -- decRef and
// maybe close the shared norm stream
if (Enclosing_Instance.singleNormRef.DecRef() == 0)
{
Enclosing_Instance.singleNormStream.Close();
Enclosing_Instance.singleNormStream = null;
}
}
in_Renamed = null;
}
}
internal CSIndexInput(IndexInput base_Renamed, long fileOffset, long length):this(base_Renamed, fileOffset, length, BufferedIndexInput.BUFFER_SIZE)
{
}
/// <summary>Bulk write a contiguous series of documents. The
/// lengths array is the length (in bytes) of each raw
/// document. The stream IndexInput is the
/// fieldsStream from which we should bulk-copy all
/// bytes.
/// </summary>
internal void AddRawDocuments(IndexInput stream, int[] lengths, int numDocs)
{
long position = fieldsStream.GetFilePointer();
long start = position;
for (int i = 0; i < numDocs; i++)
{
indexStream.WriteLong(position);
position += lengths[i];
}
fieldsStream.CopyBytes(stream, position - start);
System.Diagnostics.Debug.Assert(fieldsStream.GetFilePointer() == position);
}
internal DefaultSkipListReader(IndexInput skipStream, int maxSkipLevels, int skipInterval):base(skipStream, maxSkipLevels, skipInterval)
{
freqPointer = new long[maxSkipLevels];
proxPointer = new long[maxSkipLevels];
payloadLength = new int[maxSkipLevels];
}
/// <summary>Returns a {@link 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(System.Collections.IList 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;
System.String 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.GetUserData();
System.String userDataString;
if (sis.GetUserData().Count > 0)
{
userDataString = " userData=" + SupportClass.CollectionsHelper.CollectionToString(sis.GetUserData());
}
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:");
}
System.Collections.IEnumerator it = onlySegments.GetEnumerator();
while (it.MoveNext())
{
if (infoStream != null)
{
infoStream.Write(" " + it.Current);
}
}
System.Collections.IEnumerator e = onlySegments.GetEnumerator();
while (e.MoveNext() == true)
{
result.segmentsChecked.Add(e.Current);
}
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;
}
Status.SegmentInfoStatus 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.GetHasProx());
segInfoStat.hasProx = info.GetHasProx();
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);
System.Collections.Generic.IDictionary <string, string> diagnostics = info.GetDiagnostics();
segInfoStat.diagnostics = diagnostics;
if (diagnostics.Count > 0)
{
Msg(" diagnostics = " + SupportClass.CollectionsHelper.CollectionToString(diagnostics));
}
int docStoreOffset = info.GetDocStoreOffset();
if (docStoreOffset != -1)
{
Msg(" docStoreOffset=" + docStoreOffset);
segInfoStat.docStoreOffset = docStoreOffset;
Msg(" docStoreSegment=" + info.GetDocStoreSegment());
segInfoStat.docStoreSegment = info.GetDocStoreSegment();
Msg(" docStoreIsCompoundFile=" + info.GetDocStoreIsCompoundFile());
segInfoStat.docStoreCompoundFile = info.GetDocStoreIsCompoundFile();
}
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(info);
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..............");
}
System.Collections.Generic.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 System.SystemException("Field Norm test failed");
}
else if (segInfoStat.termIndexStatus.error != null)
{
throw new System.SystemException("Term Index test failed");
}
else if (segInfoStat.storedFieldStatus.error != null)
{
throw new System.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");
System.String comment;
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(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);
}
/// <summary> Subclasses must implement the actual skip data encoding in this method. /// /// </summary> /// <param name="level">the level skip data shall be read from /// </param> /// <param name="skipStream">the skip stream to read from /// </param> protected internal abstract int ReadSkipData(int level, IndexInput skipStream);
internal TermVectorsReader(Directory d, System.String segment, FieldInfos fieldInfos, int readBufferSize, int docStoreOffset, int size)
{
bool success = false;
try
{
if (d.FileExists(segment + "." + IndexFileNames.VECTORS_INDEX_EXTENSION))
{
tvx = d.OpenInput(segment + "." + IndexFileNames.VECTORS_INDEX_EXTENSION, readBufferSize);
format = CheckValidFormat(tvx);
tvd = d.OpenInput(segment + "." + IndexFileNames.VECTORS_DOCUMENTS_EXTENSION, readBufferSize);
int tvdFormat = CheckValidFormat(tvd);
tvf = d.OpenInput(segment + "." + IndexFileNames.VECTORS_FIELDS_EXTENSION, readBufferSize);
int tvfFormat = CheckValidFormat(tvf);
System.Diagnostics.Debug.Assert(format == tvdFormat);
System.Diagnostics.Debug.Assert(format == tvfFormat);
if (format >= FORMAT_VERSION2)
{
System.Diagnostics.Debug.Assert((tvx.Length() - FORMAT_SIZE) % 16 == 0);
numTotalDocs = (int)(tvx.Length() >> 4);
}
else
{
System.Diagnostics.Debug.Assert((tvx.Length() - FORMAT_SIZE) % 8 == 0);
numTotalDocs = (int)(tvx.Length() >> 3);
}
if (-1 == docStoreOffset)
{
this.docStoreOffset = 0;
this.size = numTotalDocs;
System.Diagnostics.Debug.Assert(size == 0 || numTotalDocs == size);
}
else
{
this.docStoreOffset = docStoreOffset;
this.size = size;
// Verify the file is long enough to hold all of our
// docs
System.Diagnostics.Debug.Assert(numTotalDocs >= size + docStoreOffset, "numTotalDocs=" + numTotalDocs + " size=" + size + " docStoreOffset=" + docStoreOffset);
}
}
else
{
// If all documents flushed in a segment had hit
// non-aborting exceptions, it's possible that
// FieldInfos.hasVectors returns true yet the term
// vector files don't exist.
format = 0;
}
this.fieldInfos = fieldInfos;
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)
{
Close();
}
}
}
protected internal override int ReadSkipData(int level, IndexInput skipStream)
{
int delta;
if (currentFieldStoresPayloads)
{
// the current field stores payloads.
// if the doc delta is odd then we have
// to read the current payload length
// because it differs from the length of the
// previous payload
delta = skipStream.ReadVInt();
if ((delta & 1) != 0)
{
payloadLength[level] = skipStream.ReadVInt();
}
delta = SupportClass.Number.URShift(delta, 1);
}
else
{
delta = skipStream.ReadVInt();
}
freqPointer[level] += skipStream.ReadVInt();
proxPointer[level] += skipStream.ReadVInt();
return delta;
}
/// <summary> Construct a new SegmentInfo instance by reading a
/// previously saved SegmentInfo from input.
///
/// </summary>
/// <param name="dir">directory to load from
/// </param>
/// <param name="format">format of the segments info file
/// </param>
/// <param name="input">input handle to read segment info from
/// </param>
internal SegmentInfo(Directory dir, int format, IndexInput input)
{
this.dir = dir;
name = input.ReadString();
docCount = input.ReadInt();
if (format <= SegmentInfos.FORMAT_LOCKLESS)
{
delGen = input.ReadLong();
if (format <= SegmentInfos.FORMAT_SHARED_DOC_STORE)
{
docStoreOffset = input.ReadInt();
if (docStoreOffset != -1)
{
docStoreSegment = input.ReadString();
docStoreIsCompoundFile = (1 == input.ReadByte());
}
else
{
docStoreSegment = name;
docStoreIsCompoundFile = false;
}
}
else
{
docStoreOffset = -1;
docStoreSegment = name;
docStoreIsCompoundFile = false;
}
if (format <= SegmentInfos.FORMAT_SINGLE_NORM_FILE)
{
hasSingleNormFile = (1 == input.ReadByte());
}
else
{
hasSingleNormFile = false;
}
int numNormGen = input.ReadInt();
if (numNormGen == NO)
{
normGen = null;
}
else
{
normGen = new long[numNormGen];
for (int j = 0; j < numNormGen; j++)
{
normGen[j] = input.ReadLong();
}
}
isCompoundFile = (sbyte)input.ReadByte();
preLockless = (isCompoundFile == CHECK_DIR);
if (format <= SegmentInfos.FORMAT_DEL_COUNT)
{
delCount = input.ReadInt();
System.Diagnostics.Debug.Assert(delCount <= docCount);
}
else
{
delCount = -1;
}
if (format <= SegmentInfos.FORMAT_HAS_PROX)
{
hasProx = input.ReadByte() == 1;
}
else
{
hasProx = true;
}
if (format <= SegmentInfos.FORMAT_DIAGNOSTICS)
{
diagnostics = input.ReadStringStringMap();
}
else
{
diagnostics = new System.Collections.Generic.Dictionary <string, string>();
}
}
else
{
delGen = CHECK_DIR;
normGen = null;
isCompoundFile = (sbyte)(CHECK_DIR);
preLockless = true;
hasSingleNormFile = false;
docStoreOffset = -1;
docStoreIsCompoundFile = false;
docStoreSegment = null;
delCount = -1;
hasProx = true;
diagnostics = new System.Collections.Generic.Dictionary <string, string>();
}
}
// It is not always neccessary to move the prox pointer
// to a new document after the freq pointer has been moved.
// Consider for example a phrase query with two terms:
// the freq pointer for term 1 has to move to document x
// to answer the question if the term occurs in that document. But
// only if term 2 also matches document x, the positions have to be
// read to figure out if term 1 and term 2 appear next
// to each other in document x and thus satisfy the query.
// So we move the prox pointer lazily to the document
// as soon as positions are requested.
private void LazySkip()
{
if (proxStream == null)
{
// clone lazily
proxStream = (IndexInput) parent.core.proxStream.Clone();
}
// we might have to skip the current payload
// if it was not read yet
SkipPayload();
if (lazySkipPointer != - 1)
{
proxStream.Seek(lazySkipPointer);
lazySkipPointer = - 1;
}
if (lazySkipProxCount != 0)
{
SkipPositions(lazySkipProxCount);
lazySkipProxCount = 0;
}
}
/// <summary>Read as a bit set </summary>
private void ReadBits(IndexInput input)
{
count = input.ReadInt(); // read count
bits = new byte[(size >> 3) + 1]; // allocate bits
input.ReadBytes(bits, 0, bits.Length);
}
private void Read(IndexInput input, System.String fileName)
{
int firstInt = input.ReadVInt();
if (firstInt < 0)
{
// This is a real format
format = firstInt;
}
else
{
format = FORMAT_PRE;
}
if (format != FORMAT_PRE & format != FORMAT_START)
{
throw new CorruptIndexException("unrecognized format " + format + " in file \"" + fileName + "\"");
}
int size;
if (format == FORMAT_PRE)
{
size = firstInt;
}
else
{
size = input.ReadVInt(); //read in the size
}
for (int i = 0; i < size; i++)
{
System.String name = StringHelper.Intern(input.ReadString());
byte bits = input.ReadByte();
bool isIndexed = (bits & IS_INDEXED) != 0;
bool storeTermVector = (bits & STORE_TERMVECTOR) != 0;
bool storePositionsWithTermVector = (bits & STORE_POSITIONS_WITH_TERMVECTOR) != 0;
bool storeOffsetWithTermVector = (bits & STORE_OFFSET_WITH_TERMVECTOR) != 0;
bool omitNorms = (bits & OMIT_NORMS) != 0;
bool storePayloads = (bits & STORE_PAYLOADS) != 0;
bool omitTermFreqAndPositions = (bits & OMIT_TERM_FREQ_AND_POSITIONS) != 0;
AddInternal(name, isIndexed, storeTermVector, storePositionsWithTermVector, storeOffsetWithTermVector, omitNorms, storePayloads, omitTermFreqAndPositions);
}
if (input.GetFilePointer() != input.Length())
{
throw new CorruptIndexException("did not read all bytes from file \"" + fileName + "\": read " + input.GetFilePointer() + " vs size " + input.Length());
}
}
internal SegmentTermPositions(SegmentReader p):base(p)
{
this.proxStream = null; // the proxStream will be cloned lazily when nextPosition() is called for the first time
}
