| 1 | /** |
| 2 | * Copyright (c) 2006-2023 LOVE Development Team |
| 3 | * |
| 4 | * This software is provided 'as-is', without any express or implied |
| 5 | * warranty. In no event will the authors be held liable for any damages |
| 6 | * arising from the use of this software. |
| 7 | * |
| 8 | * Permission is granted to anyone to use this software for any purpose, |
| 9 | * including commercial applications, and to alter it and redistribute it |
| 10 | * freely, subject to the following restrictions: |
| 11 | * |
| 12 | * 1. The origin of this software must not be misrepresented; you must not |
| 13 | * claim that you wrote the original software. If you use this software |
| 14 | * in a product, an acknowledgment in the product documentation would be |
| 15 | * appreciated but is not required. |
| 16 | * 2. Altered source versions must be plainly marked as such, and must not be |
| 17 | * misrepresented as being the original software. |
| 18 | * 3. This notice may not be removed or altered from any source distribution. |
| 19 | **/ |
| 20 | |
| 21 | // LOVE |
| 22 | #include "Compressor.h" |
| 23 | #include "common/config.h" |
| 24 | #include "common/int.h" |
| 25 | |
| 26 | #include "libraries/lz4/lz4.h" |
| 27 | #include "libraries/lz4/lz4hc.h" |
| 28 | |
| 29 | #include <zlib.h> |
| 30 | |
| 31 | namespace love |
| 32 | { |
| 33 | namespace data |
| 34 | { |
| 35 | |
| 36 | class LZ4Compressor : public Compressor |
| 37 | { |
| 38 | public: |
| 39 | |
| 40 | char *compress(Format format, const char *data, size_t dataSize, int level, size_t &compressedSize) override |
| 41 | { |
| 42 | if (format != FORMAT_LZ4) |
| 43 | throw love::Exception("Invalid format (expecting LZ4)" ); |
| 44 | |
| 45 | if (dataSize > LZ4_MAX_INPUT_SIZE) |
| 46 | throw love::Exception("Data is too large for LZ4 compressor." ); |
| 47 | |
| 48 | // We use a custom header to store some info with the compressed data. |
| 49 | const size_t = sizeof(uint32); |
| 50 | |
| 51 | int maxdestsize = LZ4_compressBound((int) dataSize); |
| 52 | size_t maxsize = headersize + (size_t) maxdestsize; |
| 53 | char *compressedbytes = nullptr; |
| 54 | |
| 55 | try |
| 56 | { |
| 57 | compressedbytes = new char[maxsize]; |
| 58 | } |
| 59 | catch (std::bad_alloc &) |
| 60 | { |
| 61 | throw love::Exception("Out of memory." ); |
| 62 | } |
| 63 | |
| 64 | // Store the size of the uncompressed data as a header. |
| 65 | #ifdef LOVE_BIG_ENDIAN |
| 66 | // Make sure it's little-endian for storage. |
| 67 | *(uint32 *) compressedbytes = swapuint32((uint32) dataSize); |
| 68 | #else |
| 69 | *(uint32 *) compressedbytes = (uint32) dataSize; |
| 70 | #endif |
| 71 | |
| 72 | // Use LZ4-HC for compression level 9 and higher. |
| 73 | int csize = 0; |
| 74 | if (level > 8) |
| 75 | csize = LZ4_compress_HC(data, compressedbytes + headersize, (int) dataSize, maxdestsize, LZ4HC_CLEVEL_DEFAULT); |
| 76 | else |
| 77 | csize = LZ4_compress_default(data, compressedbytes + headersize, (int) dataSize, maxdestsize); |
| 78 | |
| 79 | if (csize <= 0) |
| 80 | { |
| 81 | delete[] compressedbytes; |
| 82 | throw love::Exception("Could not LZ4-compress data." ); |
| 83 | } |
| 84 | |
| 85 | // We allocated space for the maximum possible amount of data, but the |
| 86 | // actual compressed size might be much smaller, so we should shrink the |
| 87 | // data buffer if so. |
| 88 | if ((double) maxsize / (double) (csize + headersize) >= 1.2) |
| 89 | { |
| 90 | char *cbytes = new (std::nothrow) char[csize + headersize]; |
| 91 | if (cbytes) |
| 92 | { |
| 93 | memcpy(cbytes, compressedbytes, csize + headersize); |
| 94 | delete[] compressedbytes; |
| 95 | compressedbytes = cbytes; |
| 96 | } |
| 97 | } |
| 98 | |
| 99 | compressedSize = (size_t) csize + headersize; |
| 100 | return compressedbytes; |
| 101 | } |
| 102 | |
| 103 | char *decompress(Format format, const char *data, size_t dataSize, size_t &decompressedSize) override |
| 104 | { |
| 105 | if (format != FORMAT_LZ4) |
| 106 | throw love::Exception("Invalid format (expecting LZ4)" ); |
| 107 | |
| 108 | const size_t = sizeof(uint32); |
| 109 | char *rawbytes = nullptr; |
| 110 | |
| 111 | if (dataSize < headersize) |
| 112 | throw love::Exception("Invalid LZ4-compressed data size." ); |
| 113 | |
| 114 | // Extract the original uncompressed size (stored in our custom header.) |
| 115 | #ifdef LOVE_BIG_ENDIAN |
| 116 | // Convert from stored little-endian to big-endian. |
| 117 | uint32 rawsize = swapuint32(*(uint32 *) data); |
| 118 | #else |
| 119 | uint32 rawsize = *(uint32 *) data; |
| 120 | #endif |
| 121 | |
| 122 | try |
| 123 | { |
| 124 | rawbytes = new char[rawsize]; |
| 125 | } |
| 126 | catch (std::bad_alloc &) |
| 127 | { |
| 128 | throw love::Exception("Out of memory." ); |
| 129 | } |
| 130 | |
| 131 | // If the uncompressed size is passed in as an argument (non-zero) and |
| 132 | // it matches the header's stored size, then we assume it's 100% accurate |
| 133 | // and we use a more efficient decompression function. |
| 134 | if (decompressedSize > 0 && decompressedSize == (size_t) rawsize) |
| 135 | { |
| 136 | // We don't use the header here, but we need to account for its size. |
| 137 | if (LZ4_decompress_fast(data + headersize, rawbytes, (int) decompressedSize) < 0) |
| 138 | { |
| 139 | delete[] rawbytes; |
| 140 | throw love::Exception("Could not decompress LZ4-compressed data." ); |
| 141 | } |
| 142 | } |
| 143 | else |
| 144 | { |
| 145 | // Account for our custom header's size in the decompress arguments. |
| 146 | int result = LZ4_decompress_safe(data + headersize, rawbytes, |
| 147 | (int) (dataSize - headersize), rawsize); |
| 148 | |
| 149 | if (result < 0) |
| 150 | { |
| 151 | delete[] rawbytes; |
| 152 | throw love::Exception("Could not decompress LZ4-compressed data." ); |
| 153 | } |
| 154 | |
| 155 | decompressedSize = (size_t) result; |
| 156 | } |
| 157 | |
| 158 | return rawbytes; |
| 159 | } |
| 160 | |
| 161 | bool isSupported(Format format) const override |
| 162 | { |
| 163 | return format == FORMAT_LZ4; |
| 164 | } |
| 165 | |
| 166 | }; // LZ4Compressor |
| 167 | |
| 168 | |
| 169 | class zlibCompressor : public Compressor |
| 170 | { |
| 171 | private: |
| 172 | |
| 173 | // The following three functions are mostly copied from the zlib source |
| 174 | // (compressBound, compress2, and uncompress), but modified to support both |
| 175 | // zlib and gzip. |
| 176 | |
| 177 | uLong zlibCompressBound(Format format, uLong sourceLen) |
| 178 | { |
| 179 | uLong size = sourceLen + (sourceLen >> 12) + (sourceLen >> 14) + (sourceLen >> 25) + 13; |
| 180 | |
| 181 | // The gzip header is slightly larger than the zlib header. |
| 182 | if (format == FORMAT_GZIP) |
| 183 | size += 18 - 6; |
| 184 | |
| 185 | return size; |
| 186 | } |
| 187 | |
| 188 | int zlibCompress(Format format, Bytef *dest, uLongf *destLen, const Bytef *source, uLong sourceLen, int level) |
| 189 | { |
| 190 | z_stream stream = {}; |
| 191 | |
| 192 | stream.next_in = (Bytef *) source; |
| 193 | stream.avail_in = (uInt) sourceLen; |
| 194 | |
| 195 | stream.next_out = dest; |
| 196 | stream.avail_out = (uInt) (*destLen); |
| 197 | |
| 198 | int windowbits = 15; |
| 199 | if (format == FORMAT_GZIP) |
| 200 | windowbits += 16; // This tells zlib to use a gzip header. |
| 201 | else if (format == FORMAT_DEFLATE) |
| 202 | windowbits = -windowbits; |
| 203 | |
| 204 | int err = deflateInit2(&stream, level, Z_DEFLATED, windowbits, 8, Z_DEFAULT_STRATEGY); |
| 205 | |
| 206 | if (err != Z_OK) |
| 207 | return err; |
| 208 | |
| 209 | err = deflate(&stream, Z_FINISH); |
| 210 | |
| 211 | if (err != Z_STREAM_END) |
| 212 | { |
| 213 | deflateEnd(&stream); |
| 214 | return err == Z_OK ? Z_BUF_ERROR : err; |
| 215 | } |
| 216 | |
| 217 | *destLen = stream.total_out; |
| 218 | |
| 219 | return deflateEnd(&stream); |
| 220 | } |
| 221 | |
| 222 | int zlibDecompress(Format format, Bytef *dest, uLongf *destLen, const Bytef *source, uLong sourceLen) |
| 223 | { |
| 224 | z_stream stream = {}; |
| 225 | |
| 226 | stream.next_in = (Bytef *) source; |
| 227 | stream.avail_in = (uInt) sourceLen; |
| 228 | |
| 229 | stream.next_out = dest; |
| 230 | stream.avail_out = (uInt) (*destLen); |
| 231 | |
| 232 | // 15 is the default. Adding 32 makes zlib auto-detect the header type. |
| 233 | int windowbits = 15 + 32; |
| 234 | |
| 235 | if (format == FORMAT_DEFLATE) |
| 236 | windowbits = -15; |
| 237 | |
| 238 | int err = inflateInit2(&stream, windowbits); |
| 239 | |
| 240 | if (err != Z_OK) |
| 241 | return err; |
| 242 | |
| 243 | err = inflate(&stream, Z_FINISH); |
| 244 | |
| 245 | if (err != Z_STREAM_END) |
| 246 | { |
| 247 | inflateEnd(&stream); |
| 248 | if (err == Z_NEED_DICT || (err == Z_BUF_ERROR && stream.avail_in == 0)) |
| 249 | return Z_DATA_ERROR; |
| 250 | return err; |
| 251 | } |
| 252 | |
| 253 | *destLen = stream.total_out; |
| 254 | |
| 255 | return inflateEnd(&stream); |
| 256 | } |
| 257 | |
| 258 | public: |
| 259 | |
| 260 | char *compress(Format format, const char *data, size_t dataSize, int level, size_t &compressedSize) override |
| 261 | { |
| 262 | if (!isSupported(format)) |
| 263 | throw love::Exception("Invalid format (expecting zlib or gzip)" ); |
| 264 | |
| 265 | if (level < 0) |
| 266 | level = Z_DEFAULT_COMPRESSION; |
| 267 | else if (level > 9) |
| 268 | level = 9; |
| 269 | |
| 270 | uLong maxsize = zlibCompressBound(format, (uLong) dataSize); |
| 271 | char *compressedbytes = nullptr; |
| 272 | |
| 273 | try |
| 274 | { |
| 275 | compressedbytes = new char[maxsize]; |
| 276 | } |
| 277 | catch (std::bad_alloc &) |
| 278 | { |
| 279 | throw love::Exception("Out of memory." ); |
| 280 | } |
| 281 | |
| 282 | uLongf destlen = maxsize; |
| 283 | int status = zlibCompress(format, (Bytef *) compressedbytes, &destlen, (const Bytef *) data, (uLong) dataSize, level); |
| 284 | |
| 285 | if (status != Z_OK) |
| 286 | { |
| 287 | delete[] compressedbytes; |
| 288 | throw love::Exception("Could not zlib/gzip-compress data." ); |
| 289 | } |
| 290 | |
| 291 | // We allocated space for the maximum possible amount of data, but the |
| 292 | // actual compressed size might be much smaller, so we should shrink the |
| 293 | // data buffer if so. |
| 294 | if ((double) maxsize / (double) destlen >= 1.3) |
| 295 | { |
| 296 | char *cbytes = new (std::nothrow) char[destlen]; |
| 297 | if (cbytes) |
| 298 | { |
| 299 | memcpy(cbytes, compressedbytes, destlen); |
| 300 | delete[] compressedbytes; |
| 301 | compressedbytes = cbytes; |
| 302 | } |
| 303 | } |
| 304 | |
| 305 | compressedSize = (size_t) destlen; |
| 306 | return compressedbytes; |
| 307 | } |
| 308 | |
| 309 | char *decompress(Format format, const char *data, size_t dataSize, size_t &decompressedSize) override |
| 310 | { |
| 311 | if (!isSupported(format)) |
| 312 | throw love::Exception("Invalid format (expecting zlib or gzip)" ); |
| 313 | |
| 314 | char *rawbytes = nullptr; |
| 315 | |
| 316 | // We might know the output size before decompression. If not, we guess. |
| 317 | size_t rawsize = decompressedSize > 0 ? decompressedSize : dataSize * 2; |
| 318 | |
| 319 | // Repeatedly try to decompress with an increasingly large output buffer. |
| 320 | while (true) |
| 321 | { |
| 322 | try |
| 323 | { |
| 324 | rawbytes = new char[rawsize]; |
| 325 | } |
| 326 | catch (std::bad_alloc &) |
| 327 | { |
| 328 | throw love::Exception("Out of memory." ); |
| 329 | } |
| 330 | |
| 331 | uLongf destLen = (uLongf) rawsize; |
| 332 | int status = zlibDecompress(format, (Bytef *) rawbytes, &destLen, (const Bytef *) data, (uLong) dataSize); |
| 333 | |
| 334 | if (status == Z_OK) |
| 335 | { |
| 336 | decompressedSize = (size_t) destLen; |
| 337 | break; |
| 338 | } |
| 339 | else if (status != Z_BUF_ERROR) |
| 340 | { |
| 341 | // For any error other than "not enough room", throw an exception. |
| 342 | delete[] rawbytes; |
| 343 | throw love::Exception("Could not decompress zlib/gzip-compressed data." ); |
| 344 | } |
| 345 | |
| 346 | // Not enough room in the output buffer: try again with a larger size. |
| 347 | delete[] rawbytes; |
| 348 | rawsize *= 2; |
| 349 | } |
| 350 | |
| 351 | return rawbytes; |
| 352 | } |
| 353 | |
| 354 | bool isSupported(Format format) const override |
| 355 | { |
| 356 | return format == FORMAT_ZLIB || format == FORMAT_GZIP || format == FORMAT_DEFLATE; |
| 357 | } |
| 358 | |
| 359 | }; // zlibCompressor |
| 360 | |
| 361 | Compressor *Compressor::getCompressor(Format format) |
| 362 | { |
| 363 | static LZ4Compressor lz4compressor; |
| 364 | static zlibCompressor zlibcompressor; |
| 365 | |
| 366 | Compressor *compressors[] = {&lz4compressor, &zlibcompressor}; |
| 367 | |
| 368 | for (Compressor *c : compressors) |
| 369 | { |
| 370 | if (c->isSupported(format)) |
| 371 | return c; |
| 372 | } |
| 373 | |
| 374 | return nullptr; |
| 375 | } |
| 376 | |
| 377 | bool Compressor::getConstant(const char *in, Format &out) |
| 378 | { |
| 379 | return formatNames.find(in, out); |
| 380 | } |
| 381 | |
| 382 | bool Compressor::getConstant(Format in, const char *&out) |
| 383 | { |
| 384 | return formatNames.find(in, out); |
| 385 | } |
| 386 | |
| 387 | std::vector<std::string> Compressor::getConstants(Format) |
| 388 | { |
| 389 | return formatNames.getNames(); |
| 390 | } |
| 391 | |
| 392 | StringMap<Compressor::Format, Compressor::FORMAT_MAX_ENUM>::Entry Compressor::formatEntries[] = |
| 393 | { |
| 394 | { "lz4" , FORMAT_LZ4 }, |
| 395 | { "zlib" , FORMAT_ZLIB }, |
| 396 | { "gzip" , FORMAT_GZIP }, |
| 397 | { "deflate" , FORMAT_DEFLATE }, |
| 398 | }; |
| 399 | |
| 400 | StringMap<Compressor::Format, Compressor::FORMAT_MAX_ENUM> Compressor::formatNames(Compressor::formatEntries, sizeof(Compressor::formatEntries)); |
| 401 | |
| 402 | } // data |
| 403 | } // love |
| 404 | |