示例#1
0
        public /*internal*/ Document Doc(int n, FieldSelector fieldSelector, IState state)
        {
            SeekIndex(n, state);
            long position = indexStream.ReadLong(state);

            fieldsStream.Seek(position, state);

            var doc       = new Document();
            int numFields = fieldsStream.ReadVInt(state);

            for (int i = 0; i < numFields; i++)
            {
                int                 fieldNumber = fieldsStream.ReadVInt(state);
                FieldInfo           fi          = fieldInfos.FieldInfo(fieldNumber);
                FieldSelectorResult acceptField = fieldSelector == null?FieldSelectorResult.LOAD:fieldSelector.Accept(fi.name);

                byte bits = fieldsStream.ReadByte(state);
                System.Diagnostics.Debug.Assert(bits <= FieldsWriter.FIELD_IS_COMPRESSED + FieldsWriter.FIELD_IS_TOKENIZED + FieldsWriter.FIELD_IS_BINARY);

                bool compressed = (bits & FieldsWriter.FIELD_IS_COMPRESSED) != 0;
                System.Diagnostics.Debug.Assert(
                    (!compressed || (format < FieldsWriter.FORMAT_LUCENE_3_0_NO_COMPRESSED_FIELDS)),
                    "compressed fields are only allowed in indexes of version <= 2.9");
                bool tokenize = (bits & FieldsWriter.FIELD_IS_TOKENIZED) != 0;
                bool binary   = (bits & FieldsWriter.FIELD_IS_BINARY) != 0;
                //TODO: Find an alternative approach here if this list continues to grow beyond the
                //list of 5 or 6 currently here.  See Lucene 762 for discussion
                if (acceptField.Equals(FieldSelectorResult.LOAD))
                {
                    AddField(doc, fi, binary, compressed, tokenize, state);
                }
                else if (acceptField.Equals(FieldSelectorResult.LOAD_AND_BREAK))
                {
                    AddField(doc, fi, binary, compressed, tokenize, state);
                    break;                     //Get out of this loop
                }
                else if (acceptField.Equals(FieldSelectorResult.LAZY_LOAD))
                {
                    AddFieldLazy(doc, fi, binary, compressed, tokenize, state);
                }
                else if (acceptField.Equals(FieldSelectorResult.SIZE))
                {
                    SkipField(binary, compressed, AddFieldSize(doc, fi, binary, compressed, state), state);
                }
                else if (acceptField.Equals(FieldSelectorResult.SIZE_AND_BREAK))
                {
                    AddFieldSize(doc, fi, binary, compressed, state);
                    break;
                }
                else
                {
                    SkipField(binary, compressed, state);
                }
            }

            return(doc);
        }
示例#2
0
        private void  Read(IndexInput input, String fileName, IState state)
        {
            int firstInt = input.ReadVInt(state);

            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(state);                 //read in the size
            }

            for (int i = 0; i < size; i++)
            {
                String name            = StringHelper.Intern(input.ReadString(state));
                byte   bits            = input.ReadByte(state);
                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.FilePointer(state) != input.Length(state))
            {
                throw new CorruptIndexException("did not read all bytes from file \"" + fileName + "\": read " + input.FilePointer(state) + " vs size " + input.Length(state));
            }
        }
示例#3
0
        /// <summary>read as a d-gaps list </summary>
        private void  ReadDgaps(IndexInput input, IState state)
        {
            size  = input.ReadInt(state);            // (re)read size
            count = input.ReadInt(state);            // read count
            bits  = new byte[(size >> 3) + 1];       // allocate bits
            int last = 0;
            int n    = Count();

            while (n > 0)
            {
                last      += input.ReadVInt(state);
                bits[last] = input.ReadByte(state);
                n         -= BYTE_COUNTS[bits[last] & 0xFF];
            }
        }
示例#4
0
 /// <summary>
 /// Sole constructor. </summary>
 public BlockPackedReader(IndexInput @in, int packedIntsVersion, int blockSize, long valueCount, bool direct)
 {
     this.ValueCount = valueCount;
     BlockShift = PackedInts.CheckBlockSize(blockSize, AbstractBlockPackedWriter.MIN_BLOCK_SIZE, AbstractBlockPackedWriter.MAX_BLOCK_SIZE);
     BlockMask = blockSize - 1;
     int numBlocks = PackedInts.NumBlocks(valueCount, blockSize);
     long[] minValues = null;
     SubReaders = new PackedInts.Reader[numBlocks];
     for (int i = 0; i < numBlocks; ++i)
     {
         int token = @in.ReadByte() & 0xFF;
         int bitsPerValue = (int)((uint)token >> AbstractBlockPackedWriter.BPV_SHIFT);
         if (bitsPerValue > 64)
         {
             throw new Exception("Corrupted");
         }
         if ((token & AbstractBlockPackedWriter.MIN_VALUE_EQUALS_0) == 0)
         {
             if (minValues == null)
             {
                 minValues = new long[numBlocks];
             }
             minValues[i] = BlockPackedReaderIterator.ZigZagDecode(1L + BlockPackedReaderIterator.ReadVLong(@in));
         }
         if (bitsPerValue == 0)
         {
             SubReaders[i] = new PackedInts.NullReader(blockSize);
         }
         else
         {
             int size = (int)Math.Min(blockSize, valueCount - (long)i * blockSize);
             if (direct)
             {
                 long pointer = @in.FilePointer;
                 SubReaders[i] = PackedInts.GetDirectReaderNoHeader(@in, PackedInts.Format.PACKED, packedIntsVersion, size, bitsPerValue);
                 @in.Seek(pointer + PackedInts.Format.PACKED.ByteCount(packedIntsVersion, size, bitsPerValue));
             }
             else
             {
                 SubReaders[i] = PackedInts.GetReaderNoHeader(@in, PackedInts.Format.PACKED, packedIntsVersion, size, bitsPerValue);
             }
         }
     }
     this.MinValues = minValues;
 }
示例#5
0
        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));

                    // 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));
                    }
                    catch (LockObtainFailedException e)
                    {
                        // 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);
            }
        }
示例#6
0
 /// <summary>
 /// Skip the next block of data.
 /// </summary>
 /// <param name="in">      the input where to read data </param>
 /// <exception cref="IOException"> If there is a low-level I/O error </exception>
 public void SkipBlock(IndexInput @in)
 {
     int numBits = @in.ReadByte();
     if (numBits == ALL_VALUES_EQUAL)
     {
         @in.ReadVInt();
         return;
     }
     Debug.Assert(numBits > 0 && numBits <= 32, numBits.ToString());
     int encodedSize = EncodedSizes[numBits];
     @in.Seek(@in.FilePointer + encodedSize);
 }
示例#7
0
        /// <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);
        }
示例#8
0
        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();
            }
        }
 private void ReadFields(IndexInput meta)
 {
     int fieldNumber = meta.ReadVInt();
     while (fieldNumber != -1)
     {
         int fieldType = meta.ReadByte();
         if (fieldType == NUMBER)
         {
             numerics[fieldNumber] = ReadNumericEntry(meta);
         }
         else if (fieldType == BYTES)
         {
             binaries[fieldNumber] = ReadBinaryEntry(meta);
         }
         else if (fieldType == SORTED)
         {
             sorteds[fieldNumber] = ReadSortedEntry(meta);
         }
         else if (fieldType == SORTED_SET)
         {
             sortedSets[fieldNumber] = ReadSortedSetEntry(meta);
         }
         else
         {
             throw new CorruptIndexException("invalid entry type: " + fieldType + ", input=" + meta);
         }
         fieldNumber = meta.ReadVInt();
     }
 }
        private static NumericEntry ReadNumericEntry(IndexInput meta)
        {
            var entry = new NumericEntry { offset = meta.ReadLong(), count = meta.ReadInt(), missingOffset = meta.ReadLong() };
            if (entry.missingOffset != -1)
            {
                entry.missingBytes = meta.ReadLong();
            }
            else
            {
                entry.missingBytes = 0;
            }
            entry.byteWidth = meta.ReadByte();

            return entry;
        }
 private void ReadFields(IndexInput meta, FieldInfos infos)
 {
     int fieldNumber = meta.ReadVInt();
     while (fieldNumber != -1)
     {
         int fieldType = meta.ReadByte();
         if (fieldType == NUMBER)
         {
             var entry = new NumericEntry {offset = meta.ReadLong(), missingOffset = meta.ReadLong()};
             if (entry.missingOffset != -1)
             {
                 entry.missingBytes = meta.ReadLong();
             }
             else
             {
                 entry.missingBytes = 0;
             }
             entry.format = meta.ReadByte();
             switch (entry.format)
             {
                 case DELTA_COMPRESSED:
                 case TABLE_COMPRESSED:
                 case GCD_COMPRESSED:
                 case UNCOMPRESSED:
                     break;
                 default:
                     throw new CorruptIndexException("Unknown format: " + entry.format + ", input=" + meta);
             }
             if (entry.format != UNCOMPRESSED)
             {
                 entry.packedIntsVersion = meta.ReadVInt();
             }
             numerics[fieldNumber] = entry;
         }
         else if (fieldType == BYTES)
         {
             var entry = new BinaryEntry
             {
                 offset = meta.ReadLong(),
                 numBytes = meta.ReadLong(),
                 missingOffset = meta.ReadLong()
             };
             if (entry.missingOffset != -1)
             {
                 entry.missingBytes = meta.ReadLong();
             }
             else
             {
                 entry.missingBytes = 0;
             }
             entry.minLength = meta.ReadVInt();
             entry.maxLength = meta.ReadVInt();
             if (entry.minLength != entry.maxLength)
             {
                 entry.packedIntsVersion = meta.ReadVInt();
                 entry.blockSize = meta.ReadVInt();
             }
             binaries[fieldNumber] = entry;
         }
         else if (fieldType == FST)
         {
             var entry = new FSTEntry {offset = meta.ReadLong(), numOrds = meta.ReadVLong()};
             fsts[fieldNumber] = entry;
         }
         else
         {
             throw new CorruptIndexException("invalid entry type: " + fieldType + ", input=" + meta);
         }
         fieldNumber = meta.ReadVInt();
     }
 }
示例#12
0
 public override byte ReadByte()
 {
     EnsureOpen();
     return(@delegate.ReadByte());
 }
        private void ReadFields(IndexInput meta, FieldInfos infos)
        {
            int fieldNumber = meta.ReadVInt();
            while (fieldNumber != -1)
            {
                // check should be: infos.fieldInfo(fieldNumber) != null, which incorporates negative check
                // but docvalues updates are currently buggy here (loading extra stuff, etc): LUCENE-5616
                if (fieldNumber < 0)
                {
                    // trickier to validate more: because we re-use for norms, because we use multiple entries
                    // for "composite" types like sortedset, etc.
                    throw new CorruptIndexException("Invalid field number: " + fieldNumber + ", input=" + meta);
                }
                int fieldType = meta.ReadByte();
                if (fieldType == NUMBER)
                {
                    var entry = new NumericEntry {Offset = meta.ReadLong(), Format = (sbyte)meta.ReadByte()};
                    switch (entry.Format)
                    {
                        case DELTA_COMPRESSED:
                        case TABLE_COMPRESSED:
                        case GCD_COMPRESSED:
                        case UNCOMPRESSED:
                            break;

                        default:
                            throw new CorruptIndexException("Unknown format: " + entry.Format + ", input=" + meta);
                    }
                    if (entry.Format != UNCOMPRESSED)
                    {
                        entry.PackedIntsVersion = meta.ReadVInt();
                    }
                    Numerics[fieldNumber] = entry;
                }
                else if (fieldType == BYTES)
                {
                    BinaryEntry entry = new BinaryEntry();
                    entry.Offset = meta.ReadLong();
                    entry.NumBytes = meta.ReadLong();
                    entry.MinLength = meta.ReadVInt();
                    entry.MaxLength = meta.ReadVInt();
                    if (entry.MinLength != entry.MaxLength)
                    {
                        entry.PackedIntsVersion = meta.ReadVInt();
                        entry.BlockSize = meta.ReadVInt();
                    }
                    Binaries[fieldNumber] = entry;
                }
                else if (fieldType == FST)
                {
                    FSTEntry entry = new FSTEntry();
                    entry.Offset = meta.ReadLong();
                    entry.NumOrds = meta.ReadVLong();
                    Fsts[fieldNumber] = entry;
                }
                else
                {
                    throw new CorruptIndexException("invalid entry type: " + fieldType + ", input=" + meta);
                }
                fieldNumber = meta.ReadVInt();
            }
        }
示例#14
0
        /// <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, IState state)
        {
            this.dir = dir;
            name     = input.ReadString(state);
            docCount = input.ReadInt(state);
            if (format <= SegmentInfos.FORMAT_LOCKLESS)
            {
                delGen = input.ReadLong(state);
                if (format <= SegmentInfos.FORMAT_SHARED_DOC_STORE)
                {
                    docStoreOffset = input.ReadInt(state);
                    if (docStoreOffset != -1)
                    {
                        docStoreSegment        = input.ReadString(state);
                        docStoreIsCompoundFile = (1 == input.ReadByte(state));
                    }
                    else
                    {
                        docStoreSegment        = name;
                        docStoreIsCompoundFile = false;
                    }
                }
                else
                {
                    docStoreOffset         = -1;
                    docStoreSegment        = name;
                    docStoreIsCompoundFile = false;
                }
                if (format <= SegmentInfos.FORMAT_SINGLE_NORM_FILE)
                {
                    hasSingleNormFile = (1 == input.ReadByte(state));
                }
                else
                {
                    hasSingleNormFile = false;
                }
                int numNormGen = input.ReadInt(state);
                if (numNormGen == NO)
                {
                    normGen = null;
                }
                else
                {
                    normGen = new long[numNormGen];
                    for (int j = 0; j < numNormGen; j++)
                    {
                        normGen[j] = input.ReadLong(state);
                    }
                }
                isCompoundFile = (sbyte)input.ReadByte(state);
                preLockless    = (isCompoundFile == CHECK_DIR);
                if (format <= SegmentInfos.FORMAT_DEL_COUNT)
                {
                    delCount = input.ReadInt(state);
                    System.Diagnostics.Debug.Assert(delCount <= docCount);
                }
                else
                {
                    delCount = -1;
                }
                if (format <= SegmentInfos.FORMAT_HAS_PROX)
                {
                    hasProx = input.ReadByte(state) == 1;
                }
                else
                {
                    hasProx = true;
                }

                if (format <= SegmentInfos.FORMAT_DIAGNOSTICS)
                {
                    diagnostics = input.ReadStringStringMap(state);
                }
                else
                {
                    diagnostics = new 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 Dictionary <string, string>();
            }
        }
示例#15
0
        /// <summary>
        /// Helper method that reads CFS entries from an input stream </summary>
        private static IDictionary <string, FileEntry> ReadEntries(IndexInputSlicer handle, Directory dir, string name)
        {
            IOException        priorE        = null;
            IndexInput         stream        = null;
            ChecksumIndexInput entriesStream = null;

            // read the first VInt. If it is negative, it's the version number
            // otherwise it's the count (pre-3.1 indexes)
            try
            {
                IDictionary <string, FileEntry> mapping;
#pragma warning disable 612, 618
                stream = handle.OpenFullSlice();
#pragma warning restore 612, 618
                int firstInt = stream.ReadVInt32();
                // impossible for 3.0 to have 63 files in a .cfs, CFS writer was not visible
                // and separate norms/etc are outside of cfs.
                if (firstInt == CODEC_MAGIC_BYTE1)
                {
                    sbyte secondByte = (sbyte)stream.ReadByte();
                    sbyte thirdByte  = (sbyte)stream.ReadByte();
                    sbyte fourthByte = (sbyte)stream.ReadByte();
                    if (secondByte != CODEC_MAGIC_BYTE2 || thirdByte != CODEC_MAGIC_BYTE3 || fourthByte != CODEC_MAGIC_BYTE4)
                    {
                        throw new CorruptIndexException("Illegal/impossible header for CFS file: " + secondByte + "," + thirdByte + "," + fourthByte);
                    }
                    int    version         = CodecUtil.CheckHeaderNoMagic(stream, CompoundFileWriter.DATA_CODEC, CompoundFileWriter.VERSION_START, CompoundFileWriter.VERSION_CURRENT);
                    string entriesFileName = IndexFileNames.SegmentFileName(
                        IndexFileNames.StripExtension(name), "",
                        IndexFileNames.COMPOUND_FILE_ENTRIES_EXTENSION);
                    entriesStream = dir.OpenChecksumInput(entriesFileName, IOContext.READ_ONCE);
                    CodecUtil.CheckHeader(entriesStream, CompoundFileWriter.ENTRY_CODEC, CompoundFileWriter.VERSION_START, CompoundFileWriter.VERSION_CURRENT);
                    int numEntries = entriesStream.ReadVInt32();
                    mapping = new Dictionary <string, FileEntry>(numEntries);
                    for (int i = 0; i < numEntries; i++)
                    {
                        FileEntry fileEntry = new FileEntry();
                        string    id        = entriesStream.ReadString();
                        FileEntry previous  = mapping.Put(id, fileEntry);
                        if (previous != null)
                        {
                            throw new CorruptIndexException("Duplicate cfs entry id=" + id + " in CFS: " + entriesStream);
                        }
                        fileEntry.Offset = entriesStream.ReadInt64();
                        fileEntry.Length = entriesStream.ReadInt64();
                    }
                    if (version >= CompoundFileWriter.VERSION_CHECKSUM)
                    {
                        CodecUtil.CheckFooter(entriesStream);
                    }
                    else
                    {
#pragma warning disable 612, 618
                        CodecUtil.CheckEOF(entriesStream);
#pragma warning restore 612, 618
                    }
                }
                else
                {
                    // TODO remove once 3.x is not supported anymore
                    mapping = ReadLegacyEntries(stream, firstInt);
                }
                return(mapping);
            }
            catch (IOException ioe)
            {
                priorE = ioe;
            }
            finally
            {
                IOUtils.DisposeWhileHandlingException(priorE, stream, entriesStream);
            }
            // this is needed until Java 7's real try-with-resources:
            throw new InvalidOperationException("impossible to get here");
        }
示例#16
0
        public virtual void  TestDirectInstantiation()
        {
            System.IO.DirectoryInfo path = new System.IO.DirectoryInfo(AppSettings.Get("tempDir", System.IO.Path.GetTempPath()));

            int sz = 2;

            Directory[] dirs = new Directory[sz];

            dirs[0] = new SimpleFSDirectory(path, null);
            // dirs[1] = new NIOFSDirectory(path, null);
            System.Console.WriteLine("Skipping NIOFSDirectory() test under Lucene.Net");
            dirs[1] = new MMapDirectory(path, null);

            for (int i = 0; i < sz; i++)
            {
                Directory dir = dirs[i];
                dir.EnsureOpen();
                System.String fname       = "foo." + i;
                System.String lockname    = "foo" + i + ".lck";
                IndexOutput   out_Renamed = dir.CreateOutput(fname, null);
                out_Renamed.WriteByte((byte)i);
                out_Renamed.Close();

                for (int j = 0; j < sz; j++)
                {
                    Directory d2 = dirs[j];
                    d2.EnsureOpen();
                    Assert.IsTrue(d2.FileExists(fname, null));
                    Assert.AreEqual(1, d2.FileLength(fname, null));

                    // don't test read on MMapDirectory, since it can't really be
                    // closed and will cause a failure to delete the file.
                    if (d2 is MMapDirectory)
                    {
                        continue;
                    }

                    IndexInput input = d2.OpenInput(fname, null);
                    Assert.AreEqual((byte)i, input.ReadByte(null));
                    input.Close();
                }

                // delete with a different dir
                dirs[(i + 1) % sz].DeleteFile(fname, null);

                for (int j = 0; j < sz; j++)
                {
                    Directory d2 = dirs[j];
                    Assert.IsFalse(d2.FileExists(fname, null));
                }

                Lock lock_Renamed = dir.MakeLock(lockname);
                Assert.IsTrue(lock_Renamed.Obtain());

                for (int j = 0; j < sz; j++)
                {
                    Directory d2    = dirs[j];
                    Lock      lock2 = d2.MakeLock(lockname);
                    try
                    {
                        Assert.IsFalse(lock2.Obtain(1));
                    }
                    catch (LockObtainFailedException)
                    {
                        // OK
                    }
                }

                lock_Renamed.Release();

                // now lock with different dir
                lock_Renamed = dirs[(i + 1) % sz].MakeLock(lockname);
                Assert.IsTrue(lock_Renamed.Obtain());
                lock_Renamed.Release();
            }

            for (int i = 0; i < sz; i++)
            {
                Directory dir = dirs[i];
                dir.EnsureOpen();
                dir.Close();
                Assert.IsFalse(dir.isOpen_ForNUnit);
            }
        }
示例#17
0
        /// <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);
            }
        }