ClickHouse/dbms/src/IO/CompressedStream.h

61 lines
2.1 KiB
C++
Raw Normal View History

2012-01-05 10:22:02 +00:00
#pragma once
2011-06-17 21:19:39 +00:00
#include <cstdint>
2017-05-28 14:29:40 +00:00
/** Common Defines */
2011-06-17 21:19:39 +00:00
#define DBMS_MAX_COMPRESSED_SIZE 0x40000000ULL /// 1GB
2012-01-05 18:35:22 +00:00
2011-06-17 21:19:39 +00:00
#define QUICKLZ_ADDITIONAL_SPACE 400
2016-02-03 21:16:19 +00:00
#define COMPRESSED_BLOCK_HEADER_SIZE 9
2012-01-05 18:35:22 +00:00
namespace DB
{
2017-05-28 14:29:40 +00:00
/** Compression method */
enum class CompressionMethod
2012-01-05 18:35:22 +00:00
{
QuickLZ = 0,
LZ4 = 1,
2017-05-28 14:29:40 +00:00
LZ4HC = 2, /// The format is the same as for LZ4. The difference is only in compression.
ZSTD = 3, /// Experimental algorithm: https://github.com/Cyan4973/zstd
};
2017-05-28 14:29:40 +00:00
/** The compressed block format is as follows:
*
2017-05-28 14:29:40 +00:00
* The first 16 bytes are the checksum from all other bytes of the block. Now only CityHash128 is used.
* In the future, you can provide other checksums, although it will not be possible to make them different in size.
*
2017-05-28 14:29:40 +00:00
* The next byte specifies the compression algorithm. Then everything depends on the algorithm.
*
2017-05-28 14:29:40 +00:00
* The first 4 options are compatible with QuickLZ level 1.
* That is, if the value of the first byte is < 4, it is enough to use qlz_level1_decompress function to decompress.
*
2017-05-28 14:29:40 +00:00
* 0x00 - uncompressed data, small block. Next, one byte - compressed data size, including header; one byte - uncompressed data size.
* 0x01 - compressed data, QuickLZ level 1, small block. Then two bytes are similar.
* 0x02 - uncompressed data, large block. Then 4 bytes - compressed data size, including header; 4 bytes uncompressed data size.
* 0x03 - compressed data, QuickLZ level 1, large block. Then 8 bytes are similar.
*
2017-05-28 14:29:40 +00:00
* 0x82 - LZ4 or LZ4HC (they have the same format).
* Next 4 bytes - the size of the compressed data, taking into account the header; 4 bytes is the size of the uncompressed data.
*
2017-05-28 14:29:40 +00:00
* NOTE: Why is 0x82?
* Originally only QuickLZ was used. Then LZ4 was added.
* The high bit is set to distinguish from QuickLZ, and the second bit is set for compatibility,
* for the functions qlz_size_compressed, qlz_size_decompressed to work.
* Although now such compatibility is no longer relevant.
*
* 0x90 - ZSTD
*
2017-05-28 14:29:40 +00:00
* All sizes are little endian.
*/
enum class CompressionMethodByte : uint8_t
{
LZ4 = 0x82,
ZSTD = 0x90,
};
2012-01-05 18:35:22 +00:00
}