github.com/insionng/yougam@v0.0.0-20170714101924-2bc18d833463/libraries/klauspost/compress/snappy/snappy.go (about)

     1  // Copyright 2011 The Snappy-Go Authors. All rights reserved.
     2  // Copyright 2016 Klaus Post. All rights reserved.
     3  // Use of this source code is governed by a BSD-style
     4  // license that can be found in the LICENSE file.
     5  
     6  // Package snappy implements the snappy block-based compression format.
     7  // It aims for very high speeds and reasonable compression.
     8  //
     9  // The C++ snappy implementation is at https://yougam/libraries/google/snappy
    10  package snappy // import "github.com/insionng/yougam/libraries/klauspost/compress/snappy"
    11  
    12  import (
    13  	"github.com/insionng/yougam/libraries/klauspost/crc32"
    14  )
    15  
    16  /*
    17  Each encoded block begins with the varint-encoded length of the decoded data,
    18  followed by a sequence of chunks. Chunks begin and end on byte boundaries. The
    19  first byte of each chunk is broken into its 2 least and 6 most significant bits
    20  called l and m: l ranges in [0, 4) and m ranges in [0, 64). l is the chunk tag.
    21  Zero means a literal tag. All other values mean a copy tag.
    22  
    23  For literal tags:
    24    - If m < 60, the next 1 + m bytes are literal bytes.
    25    - Otherwise, let n be the little-endian unsigned integer denoted by the next
    26      m - 59 bytes. The next 1 + n bytes after that are literal bytes.
    27  
    28  For copy tags, length bytes are copied from offset bytes ago, in the style of
    29  Lempel-Ziv compression algorithms. In particular:
    30    - For l == 1, the offset ranges in [0, 1<<11) and the length in [4, 12).
    31      The length is 4 + the low 3 bits of m. The high 3 bits of m form bits 8-10
    32      of the offset. The next byte is bits 0-7 of the offset.
    33    - For l == 2, the offset ranges in [0, 1<<16) and the length in [1, 65).
    34      The length is 1 + m. The offset is the little-endian unsigned integer
    35      denoted by the next 2 bytes.
    36    - For l == 3, this tag is a legacy format that is no longer supported.
    37  */
    38  const (
    39  	tagLiteral = 0x00
    40  	tagCopy1   = 0x01
    41  	tagCopy2   = 0x02
    42  	tagCopy4   = 0x03
    43  )
    44  
    45  const (
    46  	checksumSize    = 4
    47  	chunkHeaderSize = 4
    48  	magicChunk      = "\xff\x06\x00\x00" + magicBody
    49  	magicBody       = "sNaPpY"
    50  	// https://yougam/libraries/google/snappy/blob/master/framing_format.txt says
    51  	// that "the uncompressed data in a chunk must be no longer than 65536 bytes".
    52  	maxUncompressedChunkLen = 65536
    53  
    54  	// maxEncodedLenOfMaxUncompressedChunkLen equals
    55  	// MaxEncodedLen(maxUncompressedChunkLen), but is hard coded to be a const
    56  	// instead of a variable, so that obufLen can also be a const. Their
    57  	// equivalence is confirmed by TestMaxEncodedLenOfMaxUncompressedChunkLen.
    58  	maxEncodedLenOfMaxUncompressedChunkLen = 76490
    59  
    60  	obufHeaderLen = len(magicChunk) + checksumSize + chunkHeaderSize
    61  	obufLen       = obufHeaderLen + maxEncodedLenOfMaxUncompressedChunkLen
    62  )
    63  
    64  const (
    65  	chunkTypeCompressedData   = 0x00
    66  	chunkTypeUncompressedData = 0x01
    67  	chunkTypePadding          = 0xfe
    68  	chunkTypeStreamIdentifier = 0xff
    69  )
    70  
    71  var crcTable = crc32.MakeTable(crc32.Castagnoli)
    72  
    73  // crc implements the checksum specified in section 3 of
    74  // https://yougam/libraries/google/snappy/blob/master/framing_format.txt
    75  func crc(b []byte) uint32 {
    76  	c := crc32.Update(0, crcTable, b)
    77  	return uint32(c>>15|c<<17) + 0xa282ead8
    78  }