com.ning.compress.lzf.LZFUncompressor Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of compress-lzf Show documentation
Show all versions of compress-lzf Show documentation
Compression codec for LZF encoding for particularly encoding/decoding, with reasonable compression.
Compressor is basic Lempel-Ziv codec, without Huffman (deflate/gzip) or statistical post-encoding.
See "http://oldhome.schmorp.de/marc/liblzf.html" for more on original LZF package.
package com.ning.compress.lzf;
import java.io.IOException;
import com.ning.compress.BufferRecycler;
import com.ning.compress.DataHandler;
import com.ning.compress.Uncompressor;
import com.ning.compress.lzf.util.ChunkDecoderFactory;
/**
* {@link com.ning.compress.Uncompressor} implementation for uncompressing
* LZF encoded data in "push" mode, in which input is not
* read using {@link java.io.InputStream} but rather pushed to
* uncompressor in variable length chunks.
*/
public class LZFUncompressor extends Uncompressor
{
/*
///////////////////////////////////////////////////////////////////////
// State constants
///////////////////////////////////////////////////////////////////////
*/
/**
* State in which a new block or end-of-stream is expected.
*/
protected final static int STATE_INITIAL = 0;
protected final static int STATE_HEADER_Z_GOTTEN = 1;
protected final static int STATE_HEADER_ZV_GOTTEN = 2;
protected final static int STATE_HEADER_COMPRESSED_0 = 3;
protected final static int STATE_HEADER_COMPRESSED_1 = 4;
protected final static int STATE_HEADER_COMPRESSED_2 = 5;
protected final static int STATE_HEADER_COMPRESSED_3 = 6;
protected final static int STATE_HEADER_COMPRESSED_BUFFERING = 7;
protected final static int STATE_HEADER_UNCOMPRESSED_0 = 8;
protected final static int STATE_HEADER_UNCOMPRESSED_1 = 9;
protected final static int STATE_HEADER_UNCOMPRESSED_STREAMING = 10;
/*
///////////////////////////////////////////////////////////////////////
// Configuration, helper objects
///////////////////////////////////////////////////////////////////////
*/
/**
* Handler that will receive uncompressed data.
*/
protected final DataHandler _handler;
/**
* Underlying decompressor we use for chunk decompression.
*/
protected final ChunkDecoder _decoder;
protected final BufferRecycler _recycler;
/*
///////////////////////////////////////////////////////////////////////
// Decoder state
///////////////////////////////////////////////////////////////////////
*/
/**
* Current decoding state, which determines meaning of following byte(s).
*/
protected int _state = STATE_INITIAL;
/**
* Flag set if {@link DataHandler} indicates that processing should be
* terminated.
*/
protected boolean _terminated;
/**
* Number of bytes in current, compressed block
*/
protected int _compressedLength;
/**
* Number of bytes from current block, either after uncompressing data
* (for compressed blocks), or included in stream (for uncompressed).
*/
protected int _uncompressedLength;
/**
* Buffer in which compressed input is buffered if necessary, to get
* full chunks for decoding.
*/
protected byte[] _inputBuffer;
/**
* Buffer used for data uncompressed from _inputBuffer.
*/
protected byte[] _decodeBuffer;
/**
* Number of bytes that have been buffered in {@link #_inputBuffer} to be
* uncompressed; or copied directly from uncompressed block.
*/
protected int _bytesReadFromBlock;
/*
///////////////////////////////////////////////////////////////////////
// Instance creation
///////////////////////////////////////////////////////////////////////
*/
public LZFUncompressor(DataHandler handler) {
this(handler, ChunkDecoderFactory.optimalInstance(), BufferRecycler.instance());
}
public LZFUncompressor(DataHandler handler, BufferRecycler bufferRecycler) {
this(handler, ChunkDecoderFactory.optimalInstance(), bufferRecycler);
}
public LZFUncompressor(DataHandler handler, ChunkDecoder dec)
{
this(handler, dec, BufferRecycler.instance());
}
public LZFUncompressor(DataHandler handler, ChunkDecoder dec, BufferRecycler bufferRecycler)
{
_handler = handler;
_decoder = dec;
_recycler = bufferRecycler;
}
/*
///////////////////////////////////////////////////////////////////////
// Uncompressor API implementation
///////////////////////////////////////////////////////////////////////
*/
@Override
public boolean feedCompressedData(byte[] comp, int offset, int len) throws IOException
{
final int end = offset + len;
while (offset < end) {
byte b = comp[offset++];
switch (_state) {
case STATE_INITIAL:
if (b != LZFChunk.BYTE_Z) {
_reportBadHeader(comp, offset, len, 0);
}
if (offset >= end) {
_state = STATE_HEADER_Z_GOTTEN;
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_Z_GOTTEN:
if (b != LZFChunk.BYTE_V) {
_reportBadHeader(comp, offset, len, 1);
}
if (offset >= end) {
_state = STATE_HEADER_ZV_GOTTEN;
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_ZV_GOTTEN:
_bytesReadFromBlock = 0;
{
int type = b & 0xFF;
if (type != LZFChunk.BLOCK_TYPE_COMPRESSED) {
if (type == LZFChunk.BLOCK_TYPE_NON_COMPRESSED) {
_state = STATE_HEADER_UNCOMPRESSED_0;
continue;
}
_reportBadBlockType(comp, offset, len, type);
}
}
_state = STATE_HEADER_COMPRESSED_0;
if (offset >= end) {
break;
}
b = comp[offset++];
// fall through for compressed blocks
case STATE_HEADER_COMPRESSED_0: // first byte of compressed-length
_compressedLength = b & 0xFF;
if (offset >= end) {
_state = STATE_HEADER_COMPRESSED_1;
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_COMPRESSED_1:
_compressedLength = (_compressedLength << 8) + (b & 0xFF);
if (offset >= end) {
_state = STATE_HEADER_COMPRESSED_2;
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_COMPRESSED_2:
_uncompressedLength = b & 0xFF;
if (offset >= end) {
_state = STATE_HEADER_COMPRESSED_3;
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_COMPRESSED_3:
_uncompressedLength = (_uncompressedLength << 8) + (b & 0xFF);
_state = STATE_HEADER_COMPRESSED_BUFFERING;
if (offset >= end) {
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_COMPRESSED_BUFFERING:
offset = _handleCompressed(comp, --offset, end);
// either state changes, or we run out of input...
break;
case STATE_HEADER_UNCOMPRESSED_0:
_uncompressedLength = b & 0xFF;
if (offset >= end) {
_state = STATE_HEADER_UNCOMPRESSED_1;
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_UNCOMPRESSED_1:
_uncompressedLength = (_uncompressedLength << 8) + (b & 0xFF);
_state = STATE_HEADER_UNCOMPRESSED_STREAMING;
if (offset >= end) {
break;
}
b = comp[offset++];
// fall through
case STATE_HEADER_UNCOMPRESSED_STREAMING:
offset = _handleUncompressed(comp, --offset, end);
if (_terminated) {
break;
}
// All done?
if (_bytesReadFromBlock == _uncompressedLength) {
_state = STATE_INITIAL;
}
break;
}
}
return !_terminated;
}
@Override
public void complete() throws IOException
{
byte[] b = _inputBuffer;
if (b != null) {
_inputBuffer = null;
_recycler.releaseInputBuffer(b);
}
b = _decodeBuffer;
if (b != null) {
_decodeBuffer = null;
_recycler.releaseDecodeBuffer(b);
}
// 24-May-2012, tatu: Should we call this here; or fail with exception?
_handler.allDataHandled();
if (!_terminated) {
if (_state != STATE_INITIAL) {
if (_state == STATE_HEADER_COMPRESSED_BUFFERING) {
throw new LZFException("Incomplete compressed LZF block; only got "+_bytesReadFromBlock
+" bytes, needed "+_compressedLength);
}
if (_state == STATE_HEADER_UNCOMPRESSED_STREAMING) {
throw new LZFException("Incomplete uncompressed LZF block; only got "+_bytesReadFromBlock
+" bytes, needed "+_uncompressedLength);
}
throw new LZFException("Incomplete LZF block; decoding state = "+_state);
}
}
}
/*
///////////////////////////////////////////////////////////////////////
// Helper methods, decompression
///////////////////////////////////////////////////////////////////////
*/
private final int _handleUncompressed(byte[] comp, int offset, int end) throws IOException
{
// Simple, we just do pass through...
int amount = Math.min(end-offset, _uncompressedLength-_bytesReadFromBlock);
if (!_handler.handleData(comp, offset, amount)) {
_terminated = true;
}
_bytesReadFromBlock += amount;
return offset + amount;
}
private final int _handleCompressed(byte[] comp, int offset, int end) throws IOException
{
// One special case: if we get the whole block, can avoid buffering:
int available = end-offset;
if (_bytesReadFromBlock == 0 && available >= _compressedLength) {
_uncompress(comp, offset, _compressedLength);
offset += _compressedLength;
_state = STATE_INITIAL;
return offset;
}
// otherwise need to buffer
if (_inputBuffer == null) {
_inputBuffer = _recycler.allocInputBuffer(LZFChunk.MAX_CHUNK_LEN);
}
int amount = Math.min(available, _compressedLength - _bytesReadFromBlock);
System.arraycopy(comp, offset, _inputBuffer, _bytesReadFromBlock, amount);
offset += amount;
_bytesReadFromBlock += amount;
// Got it all?
if (_bytesReadFromBlock == _compressedLength) {
_uncompress(_inputBuffer, 0, _compressedLength);
_state = STATE_INITIAL;
}
return offset;
}
private final void _uncompress(byte[] src, int srcOffset, int len) throws IOException
{
if (_decodeBuffer == null) {
_decodeBuffer = _recycler.allocDecodeBuffer(LZFChunk.MAX_CHUNK_LEN);
}
_decoder.decodeChunk(src, srcOffset, _decodeBuffer, 0, _uncompressedLength);
_handler.handleData(_decodeBuffer, 0, _uncompressedLength);
}
/*
///////////////////////////////////////////////////////////////////////
// Helper methods, error reporting
///////////////////////////////////////////////////////////////////////
*/
protected void _reportBadHeader(byte[] comp, int nextOffset, int len, int relative)
throws IOException
{
char exp = (relative == 0) ? 'Z' : 'V';
--nextOffset;
throw new LZFException("Bad block: byte #"+relative+" of block header not '"
+exp+"' (0x"+Integer.toHexString(exp)
+") but 0x"+Integer.toHexString(comp[nextOffset] & 0xFF)
+" (at "+(nextOffset-1)+"/"+(len)+")");
}
protected void _reportBadBlockType(byte[] comp, int nextOffset, int len, int type)
throws IOException
{
throw new LZFException("Bad block: unrecognized type 0x"+Integer.toHexString(type & 0xFF)
+" (at "+(nextOffset-1)+"/"+len+")");
}
}
© 2015 - 2024 Weber Informatics LLC | Privacy Policy