Mercurial > hg > orthanc
annotate Core/Compression/GzipCompressor.cpp @ 3060:4e43e67f8ecf
preparing for 2019
| author | Sebastien Jodogne <s.jodogne@gmail.com> |
|---|---|
| date | Mon, 24 Dec 2018 11:39:11 +0100 |
| parents | d924f9bb61cc |
| children | 596cfabd72c5 |
| rev | line source |
|---|---|
| 1513 | 1 /** |
| 2 * Orthanc - A Lightweight, RESTful DICOM Store | |
| 1900 | 3 * Copyright (C) 2012-2016 Sebastien Jodogne, Medical Physics |
| 1513 | 4 * Department, University Hospital of Liege, Belgium |
|
3060
4e43e67f8ecf
preparing for 2019
Sebastien Jodogne <s.jodogne@gmail.com>
parents:
2954
diff
changeset
|
5 * Copyright (C) 2017-2019 Osimis S.A., Belgium |
| 1513 | 6 * |
| 7 * This program is free software: you can redistribute it and/or | |
| 8 * modify it under the terms of the GNU General Public License as | |
| 9 * published by the Free Software Foundation, either version 3 of the | |
| 10 * License, or (at your option) any later version. | |
| 11 * | |
| 12 * In addition, as a special exception, the copyright holders of this | |
| 13 * program give permission to link the code of its release with the | |
| 14 * OpenSSL project's "OpenSSL" library (or with modified versions of it | |
| 15 * that use the same license as the "OpenSSL" library), and distribute | |
| 16 * the linked executables. You must obey the GNU General Public License | |
| 17 * in all respects for all of the code used other than "OpenSSL". If you | |
| 18 * modify file(s) with this exception, you may extend this exception to | |
| 19 * your version of the file(s), but you are not obligated to do so. If | |
| 20 * you do not wish to do so, delete this exception statement from your | |
| 21 * version. If you delete this exception statement from all source files | |
| 22 * in the program, then also delete it here. | |
| 23 * | |
| 24 * This program is distributed in the hope that it will be useful, but | |
| 25 * WITHOUT ANY WARRANTY; without even the implied warranty of | |
| 26 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
| 27 * General Public License for more details. | |
| 28 * | |
| 29 * You should have received a copy of the GNU General Public License | |
| 30 * along with this program. If not, see <http://www.gnu.org/licenses/>. | |
| 31 **/ | |
| 32 | |
| 33 | |
| 34 #include "../PrecompiledHeaders.h" | |
| 35 #include "GzipCompressor.h" | |
| 36 | |
| 37 #include <stdio.h> | |
| 38 #include <string.h> | |
| 39 #include <zlib.h> | |
| 40 | |
| 41 #include "../OrthancException.h" | |
| 42 #include "../Logging.h" | |
| 43 | |
| 44 namespace Orthanc | |
| 45 { | |
| 46 uint64_t GzipCompressor::GuessUncompressedSize(const void* compressed, | |
| 47 size_t compressedSize) | |
| 48 { | |
| 49 /** | |
| 50 * "Is there a way to find out the size of the original file which | |
| 51 * is inside a GZIP file? [...] There is no truly reliable way, | |
| 52 * other than gunzipping the stream. You do not need to save the | |
| 53 * result of the decompression, so you can determine the size by | |
| 54 * simply reading and decoding the entire file without taking up | |
| 55 * space with the decompressed result. | |
| 56 * | |
| 57 * There is an unreliable way to determine the uncompressed size, | |
| 58 * which is to look at the last four bytes of the gzip file, which | |
| 59 * is the uncompressed length of that entry modulo 232 in little | |
| 60 * endian order. | |
| 61 * | |
| 62 * It is unreliable because a) the uncompressed data may be longer | |
| 63 * than 2^32 bytes, and b) the gzip file may consist of multiple | |
| 64 * gzip streams, in which case you would find the length of only | |
| 65 * the last of those streams. | |
| 66 * | |
| 67 * If you are in control of the source of the gzip files, you know | |
| 68 * that they consist of single gzip streams, and you know that | |
| 69 * they are less than 2^32 bytes uncompressed, then and only then | |
| 70 * can you use those last four bytes with confidence." | |
| 71 * | |
| 72 * http://stackoverflow.com/a/9727599/881731 | |
| 73 **/ | |
| 74 | |
| 75 if (compressedSize < 4) | |
| 76 { | |
| 77 throw OrthancException(ErrorCode_BadFileFormat); | |
| 78 } | |
| 79 | |
| 80 const uint8_t* p = reinterpret_cast<const uint8_t*>(compressed) + compressedSize - 4; | |
| 81 | |
| 82 return ((static_cast<uint32_t>(p[0]) << 0) + | |
| 83 (static_cast<uint32_t>(p[1]) << 8) + | |
| 84 (static_cast<uint32_t>(p[2]) << 16) + | |
| 85 (static_cast<uint32_t>(p[3]) << 24)); | |
| 86 } | |
| 87 | |
| 88 | |
| 89 | |
| 90 void GzipCompressor::Compress(std::string& compressed, | |
| 91 const void* uncompressed, | |
| 92 size_t uncompressedSize) | |
| 93 { | |
| 94 uLongf compressedSize = compressBound(uncompressedSize) + 1024 /* security margin */; | |
| 95 if (compressedSize == 0) | |
| 96 { | |
| 97 compressedSize = 1; | |
| 98 } | |
| 99 | |
| 100 uint8_t* target; | |
| 101 if (HasPrefixWithUncompressedSize()) | |
| 102 { | |
| 103 compressed.resize(compressedSize + sizeof(uint64_t)); | |
| 104 target = reinterpret_cast<uint8_t*>(&compressed[0]) + sizeof(uint64_t); | |
| 105 } | |
| 106 else | |
| 107 { | |
| 108 compressed.resize(compressedSize); | |
| 109 target = reinterpret_cast<uint8_t*>(&compressed[0]); | |
| 110 } | |
| 111 | |
| 112 z_stream stream; | |
| 113 memset(&stream, 0, sizeof(stream)); | |
| 114 | |
| 115 stream.next_in = const_cast<Bytef*>(reinterpret_cast<const Bytef*>(uncompressed)); | |
| 116 stream.next_out = reinterpret_cast<Bytef*>(target); | |
| 117 | |
| 118 stream.avail_in = static_cast<uInt>(uncompressedSize); | |
| 119 stream.avail_out = static_cast<uInt>(compressedSize); | |
| 120 | |
| 121 // Ensure no overflow (if the buffer is too large for the current archicture) | |
| 122 if (static_cast<size_t>(stream.avail_in) != uncompressedSize || | |
| 123 static_cast<size_t>(stream.avail_out) != compressedSize) | |
| 124 { | |
| 125 throw OrthancException(ErrorCode_NotEnoughMemory); | |
| 126 } | |
| 127 | |
| 128 // Initialize the compression engine | |
| 129 int error = deflateInit2(&stream, | |
| 130 GetCompressionLevel(), | |
| 131 Z_DEFLATED, | |
| 132 MAX_WBITS + 16, // ask for gzip output | |
| 133 8, // default memory level | |
| 134 Z_DEFAULT_STRATEGY); | |
| 135 | |
| 136 if (error != Z_OK) | |
| 137 { | |
| 138 // Cannot initialize zlib | |
| 139 compressed.clear(); | |
| 140 throw OrthancException(ErrorCode_InternalError); | |
| 141 } | |
| 142 | |
| 143 // Compress the input buffer | |
| 144 error = deflate(&stream, Z_FINISH); | |
| 145 | |
| 146 if (error != Z_STREAM_END) | |
| 147 { | |
| 148 deflateEnd(&stream); | |
| 149 compressed.clear(); | |
| 150 | |
| 151 switch (error) | |
| 152 { | |
| 153 case Z_MEM_ERROR: | |
| 154 throw OrthancException(ErrorCode_NotEnoughMemory); | |
| 155 | |
| 156 default: | |
| 157 throw OrthancException(ErrorCode_InternalError); | |
| 158 } | |
| 159 } | |
| 160 | |
| 161 size_t size = stream.total_out; | |
| 162 | |
| 163 if (deflateEnd(&stream) != Z_OK) | |
| 164 { | |
| 165 throw OrthancException(ErrorCode_InternalError); | |
| 166 } | |
| 167 | |
| 168 // The compression was successful | |
| 169 if (HasPrefixWithUncompressedSize()) | |
| 170 { | |
| 171 uint64_t s = static_cast<uint64_t>(uncompressedSize); | |
| 172 memcpy(&compressed[0], &s, sizeof(uint64_t)); | |
| 173 compressed.resize(size + sizeof(uint64_t)); | |
| 174 } | |
| 175 else | |
| 176 { | |
| 177 compressed.resize(size); | |
| 178 } | |
| 179 } | |
| 180 | |
| 181 | |
| 182 void GzipCompressor::Uncompress(std::string& uncompressed, | |
| 183 const void* compressed, | |
| 184 size_t compressedSize) | |
| 185 { | |
| 186 uint64_t uncompressedSize; | |
| 187 const uint8_t* source = reinterpret_cast<const uint8_t*>(compressed); | |
| 188 | |
| 189 if (HasPrefixWithUncompressedSize()) | |
| 190 { | |
| 191 uncompressedSize = ReadUncompressedSizePrefix(compressed, compressedSize); | |
| 192 source += sizeof(uint64_t); | |
| 193 compressedSize -= sizeof(uint64_t); | |
| 194 } | |
| 195 else | |
| 196 { | |
| 197 uncompressedSize = GuessUncompressedSize(compressed, compressedSize); | |
| 198 } | |
| 199 | |
| 200 try | |
| 201 { | |
| 1545 | 202 uncompressed.resize(static_cast<size_t>(uncompressedSize)); |
| 1513 | 203 } |
| 204 catch (...) | |
| 205 { | |
| 206 throw OrthancException(ErrorCode_NotEnoughMemory); | |
| 207 } | |
| 208 | |
| 209 z_stream stream; | |
| 210 memset(&stream, 0, sizeof(stream)); | |
| 211 | |
| 212 char dummy = '\0'; // zlib does not like NULL output buffers (even if the uncompressed data is empty) | |
| 213 stream.next_in = const_cast<Bytef*>(source); | |
| 214 stream.next_out = reinterpret_cast<Bytef*>(uncompressedSize == 0 ? &dummy : &uncompressed[0]); | |
| 215 | |
| 216 stream.avail_in = static_cast<uInt>(compressedSize); | |
| 217 stream.avail_out = static_cast<uInt>(uncompressedSize); | |
| 218 | |
| 219 // Ensure no overflow (if the buffer is too large for the current archicture) | |
| 220 if (static_cast<size_t>(stream.avail_in) != compressedSize || | |
| 221 static_cast<size_t>(stream.avail_out) != uncompressedSize) | |
| 222 { | |
| 223 throw OrthancException(ErrorCode_NotEnoughMemory); | |
| 224 } | |
| 225 | |
| 226 // Initialize the compression engine | |
| 227 int error = inflateInit2(&stream, | |
| 228 MAX_WBITS + 16); // this is a gzip input | |
| 229 | |
| 230 if (error != Z_OK) | |
| 231 { | |
| 232 // Cannot initialize zlib | |
| 233 uncompressed.clear(); | |
| 234 throw OrthancException(ErrorCode_InternalError); | |
| 235 } | |
| 236 | |
| 237 // Uncompress the input buffer | |
| 238 error = inflate(&stream, Z_FINISH); | |
| 239 | |
| 240 if (error != Z_STREAM_END) | |
| 241 { | |
| 242 inflateEnd(&stream); | |
| 243 uncompressed.clear(); | |
| 244 | |
| 245 switch (error) | |
| 246 { | |
| 247 case Z_MEM_ERROR: | |
| 248 throw OrthancException(ErrorCode_NotEnoughMemory); | |
| 249 | |
| 250 case Z_BUF_ERROR: | |
| 251 case Z_NEED_DICT: | |
| 252 throw OrthancException(ErrorCode_BadFileFormat); | |
| 253 | |
| 254 default: | |
| 255 throw OrthancException(ErrorCode_InternalError); | |
| 256 } | |
| 257 } | |
| 258 | |
| 259 size_t size = stream.total_out; | |
| 260 | |
| 261 if (inflateEnd(&stream) != Z_OK) | |
| 262 { | |
| 263 uncompressed.clear(); | |
| 264 throw OrthancException(ErrorCode_InternalError); | |
| 265 } | |
| 266 | |
| 267 if (size != uncompressedSize) | |
| 268 { | |
| 269 uncompressed.clear(); | |
| 270 | |
| 271 // The uncompressed size was not that properly guess, presumably | |
| 272 // because of a file size over 4GB. Should fallback to | |
| 273 // stream-based decompression. | |
|
2954
d924f9bb61cc
taking advantage of details in OrthancException
Sebastien Jodogne <s.jodogne@gmail.com>
parents:
2447
diff
changeset
|
274 throw OrthancException(ErrorCode_NotImplemented, |
|
d924f9bb61cc
taking advantage of details in OrthancException
Sebastien Jodogne <s.jodogne@gmail.com>
parents:
2447
diff
changeset
|
275 "The uncompressed size of a gzip-encoded buffer was not properly guessed"); |
| 1513 | 276 } |
| 277 } | |
| 278 } |