public void Read(IndexInput input, FieldInfos fieldInfos, IState state)
{
this.term = null; // invalidate cache
int start = input.ReadVInt(state);
int length = input.ReadVInt(state);
int totalLength = start + length;
if (preUTF8Strings)
{
text.SetLength(totalLength);
input.ReadChars(text.result, start, length, state);
}
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, state);
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, state);
UnicodeUtil.UTF8toUTF16(bytes.result, start, length, text);
}
}
this.field = fieldInfos.FieldName(input.ReadVInt(state));
}
public byte[] GetPayload(byte[] data, int offset, IState state)
{
if (!needToLoadPayload)
{
throw new System.IO.IOException("Either no payload exists at this term position or an attempt was made to load it more than once.");
}
// read payloads lazily
byte[] retArray;
int retOffset;
if (data == null || data.Length - offset < payloadLength)
{
// the array is too small to store the payload data,
// so we allocate a new one
retArray = new byte[payloadLength];
retOffset = 0;
}
else
{
retArray = data;
retOffset = offset;
}
proxStream.ReadBytes(retArray, retOffset, payloadLength, state);
needToLoadPayload = false;
return(retArray);
}
/// <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(), GetBinaryValue(), and TokenStreamValue() must be set.
/// </summary>
public override string StringValue(IState state)
{
Enclosing_Instance.EnsureOpen();
if (internalIsBinary)
{
return(null);
}
if (fieldsData == null)
{
IndexInput localFieldsStream = GetFieldStream(state);
try
{
localFieldsStream.Seek(pointer, state);
if (isCompressed)
{
var b = new byte[toRead];
localFieldsStream.ReadBytes(b, 0, b.Length, state);
fieldsData =
System.Text.Encoding.GetEncoding("UTF-8").GetString(Enclosing_Instance.Uncompress(b));
}
else
{
if (Enclosing_Instance.format >= FieldsWriter.FORMAT_VERSION_UTF8_LENGTH_IN_BYTES)
{
var bytes = new byte[toRead];
localFieldsStream.ReadBytes(bytes, 0, toRead, state);
fieldsData = System.Text.Encoding.GetEncoding("UTF-8").GetString(bytes);
}
else
{
//read in chars b/c we already know the length we need to read
var chars = new char[toRead];
localFieldsStream.ReadChars(chars, 0, toRead, state);
fieldsData = new System.String(chars);
}
}
}
catch (System.IO.IOException e)
{
throw new FieldReaderException(e);
}
}
return((System.String)fieldsData);
}
/// <summary>Expert: implements buffer refill. Reads bytes from the current
/// position in the input.
/// </summary>
/// <param name="b">the array to read bytes into
/// </param>
/// <param name="offset">the offset in the array to start storing bytes
/// </param>
/// <param name="len">the number of bytes to read
/// </param>
public override void ReadInternal(byte[] b, int offset, int len, IState state)
{
long start = FilePointer(state);
if (start + len > length)
{
throw new System.IO.IOException("read past EOF");
}
base_Renamed.Seek(fileOffset + start, state);
base_Renamed.ReadBytes(b, offset, len, false, state);
}
private void AddField(Document doc, FieldInfo fi, bool binary, bool compressed, bool tokenize, IState state)
{
//we have a binary stored field, and it may be compressed
if (binary)
{
int toRead = fieldsStream.ReadVInt(state);
var b = new byte[toRead];
fieldsStream.ReadBytes(b, 0, b.Length, state);
doc.Add(compressed ? new Field(fi.name, Uncompress(b), Field.Store.YES) : new Field(fi.name, b, Field.Store.YES));
}
else
{
const Field.Store store = Field.Store.YES;
Field.Index index = FieldExtensions.ToIndex(fi.isIndexed, tokenize);
Field.TermVector termVector = FieldExtensions.ToTermVector(fi.storeTermVector, fi.storeOffsetWithTermVector, fi.storePositionWithTermVector);
AbstractField f;
if (compressed)
{
int toRead = fieldsStream.ReadVInt(state);
var b = new byte[toRead];
fieldsStream.ReadBytes(b, 0, b.Length, state);
f = new Field(fi.name, false, System.Text.Encoding.GetEncoding("UTF-8").GetString(Uncompress(b)), store, index,
termVector)
{
OmitTermFreqAndPositions = fi.omitTermFreqAndPositions, OmitNorms = fi.omitNorms
};
}
else
{
f = new Field(fi.name, false, fieldsStream.ReadString(state), store, index, termVector)
{
OmitTermFreqAndPositions = fi.omitTermFreqAndPositions, OmitNorms = fi.omitNorms
};
}
doc.Add(f);
}
}
/// <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, IState state)
{
IndexInput isRenamed = null;
try
{
long startPtr = os.FilePointer;
isRenamed = directory.OpenInput(source.file, state);
long length = isRenamed.Length(state);
long remainder = length;
int chunk = buffer.Length;
while (remainder > 0)
{
var len = (int)Math.Min(chunk, remainder);
isRenamed.ReadBytes(buffer, 0, len, false, state);
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, state);
}
}
// 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.FilePointer;
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 (isRenamed != null)
{
isRenamed.Close();
}
}
}
/// <summary>Copy numBytes bytes from input to ourself. </summary>
public virtual void CopyBytes(IndexInput input, long numBytes)
{
long left = numBytes;
if (copyBuffer == null)
copyBuffer = new byte[COPY_BUFFER_SIZE];
while (left > 0)
{
int toCopy;
if (left > COPY_BUFFER_SIZE)
toCopy = COPY_BUFFER_SIZE;
else
toCopy = (int) left;
input.ReadBytes(copyBuffer, 0, toCopy);
WriteBytes(copyBuffer, 0, toCopy);
left -= toCopy;
}
}
/// <summary>Copy numBytes bytes from input to ourself. </summary>
public virtual void CopyBytes(IndexInput input, long numBytes)
{
System.Diagnostics.Debug.Assert(numBytes >= 0, "numBytes=" + numBytes);
long left = numBytes;
if (copyBuffer == null)
copyBuffer = new byte[COPY_BUFFER_SIZE];
while (left > 0)
{
int toCopy;
if (left > COPY_BUFFER_SIZE)
toCopy = COPY_BUFFER_SIZE;
else
toCopy = (int) left;
input.ReadBytes(copyBuffer, 0, toCopy);
WriteBytes(copyBuffer, 0, toCopy);
left -= toCopy;
}
}
public override byte[] GetBinaryValue(byte[] result, IState state)
{
Enclosing_Instance.EnsureOpen();
if (internalIsBinary)
{
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(state);
// 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, state);
localFieldsStream.ReadBytes(b, 0, toRead, state);
fieldsData = isCompressed ? Enclosing_Instance.Uncompress(b) : b;
}
catch (IOException e)
{
throw new FieldReaderException(e);
}
internalbinaryOffset = 0;
internalBinaryLength = toRead;
}
return((byte[])fieldsData);
}
return(null);
}
public override void ReadBytes(byte[] b, int offset, int len)
{
main.ReadBytes(b, offset, len);
digest.Update(b, offset, len);
}
/// <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, IState state)
{
// 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, state);
int numTerms = tvf.ReadVInt(state);
//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(state);
storePositions = (bits & STORE_POSITIONS_WITH_TERMVECTOR) != 0;
storeOffsets = (bits & STORE_OFFSET_WITH_TERMVECTOR) != 0;
}
else
{
tvf.ReadVInt(state);
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(state);
deltaLength = tvf.ReadVInt(state);
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, state);
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, state);
term = System.Text.Encoding.UTF8.GetString(byteBuffer, 0, totalLength);
}
int freq = tvf.ReadVInt(state);
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(state);
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(state);
}
}
}
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(state);
int endOffset = startOffset + tvf.ReadVInt(state);
offsets[j] = new TermVectorOffsetInfo(startOffset, endOffset);
prevOffset = endOffset;
}
}
else
{
for (int j = 0; j < freq; j++)
{
tvf.ReadVInt(state);
tvf.ReadVInt(state);
}
}
}
mapper.Map(term, freq, offsets, positions);
}
}
private void CheckReadBytes(IndexInput input, int size, int pos)
{
// Just to see that "offset" is treated properly in readBytes(), we
// add an arbitrary offset at the beginning of the array
int offset = size % 10; // arbitrary
buffer = ArrayUtil.Grow(buffer, offset + size);
Assert.AreEqual(pos, input.FilePointer);
long left = TEST_FILE_LENGTH - input.FilePointer;
if (left <= 0)
{
return ;
}
else if (left < size)
{
size = (int) left;
}
input.ReadBytes(buffer, offset, size);
Assert.AreEqual(pos + size, input.FilePointer);
for (int i = 0; i < size; i++)
{
Assert.AreEqual(Byten(pos + i), buffer[offset + i], "pos=" + i + " filepos=" + (pos + i));
}
}
public virtual void TestCopyBytesMem()
{
int num = AtLeast(10);
for (int iter = 0; iter < num; iter++)
{
Directory dir = NewDirectory();
if (VERBOSE)
{
Console.WriteLine("TEST: iter=" + iter + " dir=" + dir);
}
// make random file
IndexOutput @out = dir.CreateOutput("test", NewIOContext(Random()));
var bytes = new byte[TestUtil.NextInt(Random(), 1, 77777)];
int size = TestUtil.NextInt(Random(), 1, 1777777);
int upto = 0;
int byteUpto = 0;
while (upto < size)
{
bytes[byteUpto++] = Value(upto);
upto++;
if (byteUpto == bytes.Length)
{
@out.WriteBytes(bytes, 0, bytes.Length);
byteUpto = 0;
}
}
@out.WriteBytes(bytes, 0, byteUpto);
Assert.AreEqual(size, @out.GetFilePointer());
@out.Dispose();
Assert.AreEqual(size, dir.FileLength("test"));
// copy from test -> test2
IndexInput @in = dir.OpenInput("test", NewIOContext(Random()));
@out = dir.CreateOutput("test2", NewIOContext(Random()));
upto = 0;
while (upto < size)
{
if (Random().NextBoolean())
{
@out.WriteByte(@in.ReadByte());
upto++;
}
else
{
int chunk = Math.Min(TestUtil.NextInt(Random(), 1, bytes.Length), size - upto);
@out.CopyBytes(@in, chunk);
upto += chunk;
}
}
Assert.AreEqual(size, upto);
@out.Dispose();
@in.Dispose();
// verify
IndexInput in2 = dir.OpenInput("test2", NewIOContext(Random()));
upto = 0;
while (upto < size)
{
if (Random().NextBoolean())
{
var v = in2.ReadByte();
Assert.AreEqual(Value(upto), v);
upto++;
}
else
{
int limit = Math.Min(TestUtil.NextInt(Random(), 1, bytes.Length), size - upto);
in2.ReadBytes(bytes, 0, limit);
for (int byteIdx = 0; byteIdx < limit; byteIdx++)
{
Assert.AreEqual(Value(upto), bytes[byteIdx]);
upto++;
}
}
}
in2.Dispose();
dir.DeleteFile("test");
dir.DeleteFile("test2");
dir.Dispose();
}
}
internal SkipBuffer(IndexInput input, int length, IState state)
{
data = new byte[length];
pointer = input.FilePointer(state);
input.ReadBytes(data, 0, length, state);
}
public override void ReadInternal(byte[] b, int offset, int length, IState state)
{
SimOutage();
delegate_Renamed.ReadBytes(b, offset, length, null);
}
/// <summary>Read as a bit set </summary>
private void ReadBits(IndexInput input, IState state)
{
count = input.ReadInt(state); // read count
bits = new byte[(size >> 3) + 1]; // allocate bits
input.ReadBytes(bits, 0, bits.Length, state);
}
public override void ReadBytes(byte[] b, int offset, int len)
{
EnsureOpen();
@delegate.ReadBytes(b, offset, len);
}
/// <summary>
/// Read the next block of data (<code>For</code> format).
/// </summary>
/// <param name="in"> the input to use to read data </param>
/// <param name="encoded"> a buffer that can be used to store encoded data </param>
/// <param name="decoded"> where to write decoded data </param>
/// <exception cref="IOException"> If there is a low-level I/O error </exception>
public void ReadBlock(IndexInput @in, sbyte[] encoded, int[] decoded)
{
int numBits = @in.ReadByte();
Debug.Assert(numBits <= 32, numBits.ToString());
if (numBits == ALL_VALUES_EQUAL)
{
int value = @in.ReadVInt();
CollectionsHelper.Fill(decoded, 0, Lucene41PostingsFormat.BLOCK_SIZE, value);
return;
}
int encodedSize = EncodedSizes[numBits];
@in.ReadBytes(encoded, 0, encodedSize);
PackedInts.Decoder decoder = Decoders[numBits];
int iters = Iterations[numBits];
Debug.Assert(iters * decoder.ByteValueCount() >= Lucene41PostingsFormat.BLOCK_SIZE);
decoder.Decode(encoded, 0, decoded, 0, iters);
}
private void AssertSameStreams(string msg, IndexInput expected, IndexInput test)
{
Assert.IsNotNull(expected, msg + " null expected");
Assert.IsNotNull(test, msg + " null test");
Assert.AreEqual(expected.Length(), test.Length(), msg + " length");
Assert.AreEqual(expected.FilePointer, test.FilePointer, msg + " position");
var expectedBuffer = new byte[512];
var testBuffer = new byte[expectedBuffer.Length];
long remainder = expected.Length() - expected.FilePointer;
while (remainder > 0)
{
int readLen = (int)Math.Min(remainder, expectedBuffer.Length);
expected.ReadBytes(expectedBuffer, 0, readLen);
test.ReadBytes(testBuffer, 0, readLen);
AssertEqualArrays(msg + ", remainder " + remainder, expectedBuffer, testBuffer, 0, readLen);
remainder -= readLen;
}
}
public virtual void TestDirectInstantiation()
{
DirectoryInfo path = CreateTempDir("testDirectInstantiation");
byte[] largeBuffer = new byte[Random.Next(256 * 1024)], largeReadBuffer = new byte[largeBuffer.Length];
for (int i = 0; i < largeBuffer.Length; i++)
{
largeBuffer[i] = (byte)i; // automatically loops with modulo
}
var dirs = new FSDirectory[] { new SimpleFSDirectory(path, null), new NIOFSDirectory(path, null), new MMapDirectory(path, null) };
for (int i = 0; i < dirs.Length; i++)
{
FSDirectory dir = dirs[i];
dir.EnsureOpen();
string fname = "foo." + i;
string lockname = "foo" + i + ".lck";
IndexOutput @out = dir.CreateOutput(fname, NewIOContext(Random));
@out.WriteByte((byte)(sbyte)i);
@out.WriteBytes(largeBuffer, largeBuffer.Length);
@out.Dispose();
for (int j = 0; j < dirs.Length; j++)
{
FSDirectory d2 = dirs[j];
d2.EnsureOpen();
Assert.IsTrue(SlowFileExists(d2, fname));
Assert.AreEqual(1 + largeBuffer.Length, d2.FileLength(fname));
// LUCENENET specific - unmap hack not needed
//// don't do read tests if unmapping is not supported!
//if (d2 is MMapDirectory && !((MMapDirectory)d2).UseUnmap)
//{
// continue;
//}
IndexInput input = d2.OpenInput(fname, NewIOContext(Random));
Assert.AreEqual((byte)i, input.ReadByte());
// read array with buffering enabled
Arrays.Fill(largeReadBuffer, (byte)0);
input.ReadBytes(largeReadBuffer, 0, largeReadBuffer.Length, true);
Assert.AreEqual(largeBuffer, largeReadBuffer);
// read again without using buffer
input.Seek(1L);
Arrays.Fill(largeReadBuffer, (byte)0);
input.ReadBytes(largeReadBuffer, 0, largeReadBuffer.Length, false);
Assert.AreEqual(largeBuffer, largeReadBuffer);
input.Dispose();
}
// delete with a different dir
dirs[(i + 1) % dirs.Length].DeleteFile(fname);
for (int j = 0; j < dirs.Length; j++)
{
FSDirectory d2 = dirs[j];
Assert.IsFalse(SlowFileExists(d2, fname));
}
Lock @lock = dir.MakeLock(lockname);
Assert.IsTrue(@lock.Obtain());
for (int j = 0; j < dirs.Length; j++)
{
FSDirectory d2 = dirs[j];
Lock lock2 = d2.MakeLock(lockname);
try
{
Assert.IsFalse(lock2.Obtain(1));
}
#pragma warning disable 168
catch (LockObtainFailedException e)
#pragma warning restore 168
{
// OK
}
}
@lock.Dispose();
// now lock with different dir
@lock = dirs[(i + 1) % dirs.Length].MakeLock(lockname);
Assert.IsTrue(@lock.Obtain());
@lock.Dispose();
}
for (int i = 0; i < dirs.Length; i++)
{
FSDirectory dir = dirs[i];
dir.EnsureOpen();
dir.Dispose();
Assert.IsFalse(dir.IsOpen);
}
}
