#include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include namespace DB { namespace ErrorCodes { extern const int TOO_LARGE_SIZE_COMPRESSED; extern const int BAD_ARGUMENTS; } } namespace { /// Outputs sizes of uncompressed and compressed blocks for compressed file. void checkAndWriteHeader(DB::ReadBuffer & in, DB::WriteBuffer & out) { while (!in.eof()) { in.ignore(16); /// checksum char header[COMPRESSED_BLOCK_HEADER_SIZE]; in.readStrict(header, COMPRESSED_BLOCK_HEADER_SIZE); UInt32 size_compressed = unalignedLoad(&header[1]); if (size_compressed > DBMS_MAX_COMPRESSED_SIZE) throw DB::Exception("Too large size_compressed. Most likely corrupted data.", DB::ErrorCodes::TOO_LARGE_SIZE_COMPRESSED); UInt32 size_decompressed = unalignedLoad(&header[5]); DB::writeText(size_decompressed, out); DB::writeChar('\t', out); DB::writeText(size_compressed, out); DB::writeChar('\n', out); in.ignore(size_compressed - COMPRESSED_BLOCK_HEADER_SIZE); } } } #pragma GCC diagnostic ignored "-Wunused-function" #pragma GCC diagnostic ignored "-Wmissing-declarations" int mainEntryClickHouseCompressor(int argc, char ** argv) { boost::program_options::options_description desc = createOptionsDescription("Allowed options", getTerminalWidth()); desc.add_options() ("help,h", "produce help message") ("decompress,d", "decompress") ("block-size,b", boost::program_options::value()->default_value(DBMS_DEFAULT_BUFFER_SIZE), "compress in blocks of specified size") ("hc", "use LZ4HC instead of LZ4") ("zstd", "use ZSTD instead of LZ4") ("codec", boost::program_options::value>()->multitoken(), "use codecs combination instead of LZ4") ("level", boost::program_options::value(), "compression level for codecs specified via flags") ("none", "use no compression instead of LZ4") ("stat", "print block statistics of compressed data") ; boost::program_options::variables_map options; boost::program_options::store(boost::program_options::parse_command_line(argc, argv, desc), options); if (options.count("help")) { std::cout << "Usage: " << argv[0] << " [options] < in > out" << std::endl; std::cout << desc << std::endl; return 1; } try { bool decompress = options.count("decompress"); bool use_lz4hc = options.count("hc"); bool use_zstd = options.count("zstd"); bool stat_mode = options.count("stat"); bool use_none = options.count("none"); unsigned block_size = options["block-size"].as(); std::vector codecs; if (options.count("codec")) codecs = options["codec"].as>(); if ((use_lz4hc || use_zstd || use_none) && !codecs.empty()) throw DB::Exception("Wrong options, codec flags like --zstd and --codec options are mutually exclusive", DB::ErrorCodes::BAD_ARGUMENTS); if (!codecs.empty() && options.count("level")) throw DB::Exception("Wrong options, --level is not compatible with --codec list", DB::ErrorCodes::BAD_ARGUMENTS); std::string method_family = "LZ4"; if (use_lz4hc) method_family = "LZ4HC"; else if (use_zstd) method_family = "ZSTD"; else if (use_none) method_family = "NONE"; std::optional level = std::nullopt; if (options.count("level")) level = options["level"].as(); DB::CompressionCodecPtr codec; if (!codecs.empty()) { DB::ParserCodec codec_parser; std::string codecs_line = boost::algorithm::join(codecs, ","); auto ast = DB::parseQuery(codec_parser, "(" + codecs_line + ")", 0, DBMS_DEFAULT_MAX_PARSER_DEPTH); codec = DB::CompressionCodecFactory::instance().get(ast, nullptr); } else codec = DB::CompressionCodecFactory::instance().get(method_family, level); DB::ReadBufferFromFileDescriptor rb(STDIN_FILENO); DB::WriteBufferFromFileDescriptor wb(STDOUT_FILENO); if (stat_mode) { /// Output statistic for compressed file. checkAndWriteHeader(rb, wb); } else if (decompress) { /// Decompression DB::CompressedReadBuffer from(rb); DB::copyData(from, wb); } else { /// Compression DB::CompressedWriteBuffer to(wb, codec, block_size); DB::copyData(rb, to); } } catch (...) { std::cerr << DB::getCurrentExceptionMessage(true); return DB::getCurrentExceptionCode(); } return 0; }