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());
}
}
/// <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];
}
}
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());
}
}
/// <summary> </summary>
/// <param name="field">The field to read in
/// </param>
/// <param name="tvfPointer">The pointer within the tvf file where we should start reading
/// </param>
/// <param name="mapper">The mapper used to map the TermVector
/// </param>
/// <throws> IOException </throws>
private void ReadTermVector(System.String field, long tvfPointer, TermVectorMapper mapper)
{
// Now read the data from specified position
//We don't need to offset by the FORMAT here since the pointer already includes the offset
tvf.Seek(tvfPointer);
int numTerms = tvf.ReadVInt();
//System.out.println("Num Terms: " + numTerms);
// If no terms - return a constant empty termvector. However, this should never occur!
if (numTerms == 0)
{
return;
}
bool storePositions;
bool storeOffsets;
if (format >= FORMAT_VERSION)
{
byte bits = tvf.ReadByte();
storePositions = (bits & STORE_POSITIONS_WITH_TERMVECTOR) != 0;
storeOffsets = (bits & STORE_OFFSET_WITH_TERMVECTOR) != 0;
}
else
{
tvf.ReadVInt();
storePositions = false;
storeOffsets = false;
}
mapper.SetExpectations(field, numTerms, storeOffsets, storePositions);
int start = 0;
int deltaLength = 0;
int totalLength = 0;
byte[] byteBuffer;
char[] charBuffer;
bool preUTF8 = format < FORMAT_UTF8_LENGTH_IN_BYTES;
// init the buffers
if (preUTF8)
{
charBuffer = new char[10];
byteBuffer = null;
}
else
{
charBuffer = null;
byteBuffer = new byte[20];
}
for (int i = 0; i < numTerms; i++)
{
start = tvf.ReadVInt();
deltaLength = tvf.ReadVInt();
totalLength = start + deltaLength;
System.String term;
if (preUTF8)
{
// Term stored as java chars
if (charBuffer.Length < totalLength)
{
char[] newCharBuffer = new char[(int)(1.5 * totalLength)];
Array.Copy(charBuffer, 0, newCharBuffer, 0, start);
charBuffer = newCharBuffer;
}
tvf.ReadChars(charBuffer, start, deltaLength);
term = new System.String(charBuffer, 0, totalLength);
}
else
{
// Term stored as utf8 bytes
if (byteBuffer.Length < totalLength)
{
byte[] newByteBuffer = new byte[(int)(1.5 * totalLength)];
Array.Copy(byteBuffer, 0, newByteBuffer, 0, start);
byteBuffer = newByteBuffer;
}
tvf.ReadBytes(byteBuffer, start, deltaLength);
term = System.Text.Encoding.UTF8.GetString(byteBuffer, 0, totalLength);
}
int freq = tvf.ReadVInt();
int[] positions = null;
if (storePositions)
{
//read in the positions
//does the mapper even care about positions?
if (mapper.IsIgnoringPositions() == false)
{
positions = new int[freq];
int prevPosition = 0;
for (int j = 0; j < freq; j++)
{
positions[j] = prevPosition + tvf.ReadVInt();
prevPosition = positions[j];
}
}
else
{
//we need to skip over the positions. Since these are VInts, I don't believe there is anyway to know for sure how far to skip
//
for (int j = 0; j < freq; j++)
{
tvf.ReadVInt();
}
}
}
TermVectorOffsetInfo[] offsets = null;
if (storeOffsets)
{
//does the mapper even care about offsets?
if (mapper.IsIgnoringOffsets() == false)
{
offsets = new TermVectorOffsetInfo[freq];
int prevOffset = 0;
for (int j = 0; j < freq; j++)
{
int startOffset = prevOffset + tvf.ReadVInt();
int endOffset = startOffset + tvf.ReadVInt();
offsets[j] = new TermVectorOffsetInfo(startOffset, endOffset);
prevOffset = endOffset;
}
}
else
{
for (int j = 0; j < freq; j++)
{
tvf.ReadVInt();
tvf.ReadVInt();
}
}
}
mapper.Map(term, freq, offsets, positions);
}
}
/// <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>();
}
}
/// <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> 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>();
}
}