1 /* SPDX-License-Identifier: 0BSD */ 2 3 /** 4 * \file lzma/container.h 5 * \brief File formats 6 * \note Never include this file directly. Use <lzma.h> instead. 7 */ 8 9 /* 10 * Author: Lasse Collin 11 */ 12 13 #ifndef LZMA_H_INTERNAL 14 # error Never include this file directly. Use <lzma.h> instead. 15 #endif 16 17 18 /************ 19 * Encoding * 20 ************/ 21 22 /** 23 * \brief Default compression preset 24 * 25 * It's not straightforward to recommend a default preset, because in some 26 * cases keeping the resource usage relatively low is more important that 27 * getting the maximum compression ratio. 28 */ 29 #define LZMA_PRESET_DEFAULT UINT32_C(6) 30 31 32 /** 33 * \brief Mask for preset level 34 * 35 * This is useful only if you need to extract the level from the preset 36 * variable. That should be rare. 37 */ 38 #define LZMA_PRESET_LEVEL_MASK UINT32_C(0x1F) 39 40 41 /* 42 * Preset flags 43 * 44 * Currently only one flag is defined. 45 */ 46 47 /** 48 * \brief Extreme compression preset 49 * 50 * This flag modifies the preset to make the encoding significantly slower 51 * while improving the compression ratio only marginally. This is useful 52 * when you don't mind spending time to get as small result as possible. 53 * 54 * This flag doesn't affect the memory usage requirements of the decoder (at 55 * least not significantly). The memory usage of the encoder may be increased 56 * a little but only at the lowest preset levels (0-3). 57 */ 58 #define LZMA_PRESET_EXTREME (UINT32_C(1) << 31) 59 60 61 /** 62 * \brief Multithreading options 63 */ 64 typedef struct { 65 /** 66 * \brief Flags 67 * 68 * Set this to zero if no flags are wanted. 69 * 70 * Encoder: No flags are currently supported. 71 * 72 * Decoder: Bitwise-or of zero or more of the decoder flags: 73 * - LZMA_TELL_NO_CHECK 74 * - LZMA_TELL_UNSUPPORTED_CHECK 75 * - LZMA_TELL_ANY_CHECK 76 * - LZMA_IGNORE_CHECK 77 * - LZMA_CONCATENATED 78 * - LZMA_FAIL_FAST 79 */ 80 uint32_t flags; 81 82 /** 83 * \brief Number of worker threads to use 84 */ 85 uint32_t threads; 86 87 /** 88 * \brief Encoder only: Maximum uncompressed size of a Block 89 * 90 * The encoder will start a new .xz Block every block_size bytes. 91 * Using LZMA_FULL_FLUSH or LZMA_FULL_BARRIER with lzma_code() 92 * the caller may tell liblzma to start a new Block earlier. 93 * 94 * With LZMA2, a recommended block size is 2-4 times the LZMA2 95 * dictionary size. With very small dictionaries, it is recommended 96 * to use at least 1 MiB block size for good compression ratio, even 97 * if this is more than four times the dictionary size. Note that 98 * these are only recommendations for typical use cases; feel free 99 * to use other values. Just keep in mind that using a block size 100 * less than the LZMA2 dictionary size is waste of RAM. 101 * 102 * Set this to 0 to let liblzma choose the block size depending 103 * on the compression options. For LZMA2 it will be 3*dict_size 104 * or 1 MiB, whichever is more. 105 * 106 * For each thread, about 3 * block_size bytes of memory will be 107 * allocated. This may change in later liblzma versions. If so, 108 * the memory usage will probably be reduced, not increased. 109 */ 110 uint64_t block_size; 111 112 /** 113 * \brief Timeout to allow lzma_code() to return early 114 * 115 * Multithreading can make liblzma consume input and produce 116 * output in a very bursty way: it may first read a lot of input 117 * to fill internal buffers, then no input or output occurs for 118 * a while. 119 * 120 * In single-threaded mode, lzma_code() won't return until it has 121 * either consumed all the input or filled the output buffer. If 122 * this is done in multithreaded mode, it may cause a call 123 * lzma_code() to take even tens of seconds, which isn't acceptable 124 * in all applications. 125 * 126 * To avoid very long blocking times in lzma_code(), a timeout 127 * (in milliseconds) may be set here. If lzma_code() would block 128 * longer than this number of milliseconds, it will return with 129 * LZMA_OK. Reasonable values are 100 ms or more. The xz command 130 * line tool uses 300 ms. 131 * 132 * If long blocking times are acceptable, set timeout to a special 133 * value of 0. This will disable the timeout mechanism and will make 134 * lzma_code() block until all the input is consumed or the output 135 * buffer has been filled. 136 * 137 * \note Even with a timeout, lzma_code() might sometimes take 138 * a long time to return. No timing guarantees are made. 139 */ 140 uint32_t timeout; 141 142 /** 143 * \brief Encoder only: Compression preset 144 * 145 * The preset is set just like with lzma_easy_encoder(). 146 * The preset is ignored if filters below is non-NULL. 147 */ 148 uint32_t preset; 149 150 /** 151 * \brief Encoder only: Filter chain (alternative to a preset) 152 * 153 * If this is NULL, the preset above is used. Otherwise the preset 154 * is ignored and the filter chain specified here is used. 155 */ 156 const lzma_filter *filters; 157 158 /** 159 * \brief Encoder only: Integrity check type 160 * 161 * See check.h for available checks. The xz command line tool 162 * defaults to LZMA_CHECK_CRC64, which is a good choice if you 163 * are unsure. 164 */ 165 lzma_check check; 166 167 /* 168 * Reserved space to allow possible future extensions without 169 * breaking the ABI. You should not touch these, because the names 170 * of these variables may change. These are and will never be used 171 * with the currently supported options, so it is safe to leave these 172 * uninitialized. 173 */ 174 /** \private Reserved member. */ 175 lzma_reserved_enum reserved_enum1; 176 177 /** \private Reserved member. */ 178 lzma_reserved_enum reserved_enum2; 179 180 /** \private Reserved member. */ 181 lzma_reserved_enum reserved_enum3; 182 183 /** \private Reserved member. */ 184 uint32_t reserved_int1; 185 186 /** \private Reserved member. */ 187 uint32_t reserved_int2; 188 189 /** \private Reserved member. */ 190 uint32_t reserved_int3; 191 192 /** \private Reserved member. */ 193 uint32_t reserved_int4; 194 195 /** 196 * \brief Memory usage limit to reduce the number of threads 197 * 198 * Encoder: Ignored. 199 * 200 * Decoder: 201 * 202 * If the number of threads has been set so high that more than 203 * memlimit_threading bytes of memory would be needed, the number 204 * of threads will be reduced so that the memory usage will not exceed 205 * memlimit_threading bytes. However, if memlimit_threading cannot 206 * be met even in single-threaded mode, then decoding will continue 207 * in single-threaded mode and memlimit_threading may be exceeded 208 * even by a large amount. That is, memlimit_threading will never make 209 * lzma_code() return LZMA_MEMLIMIT_ERROR. To truly cap the memory 210 * usage, see memlimit_stop below. 211 * 212 * Setting memlimit_threading to UINT64_MAX or a similar huge value 213 * means that liblzma is allowed to keep the whole compressed file 214 * and the whole uncompressed file in memory in addition to the memory 215 * needed by the decompressor data structures used by each thread! 216 * In other words, a reasonable value limit must be set here or it 217 * will cause problems sooner or later. If you have no idea what 218 * a reasonable value could be, try lzma_physmem() / 4 as a starting 219 * point. Setting this limit will never prevent decompression of 220 * a file; this will only reduce the number of threads. 221 * 222 * If memlimit_threading is greater than memlimit_stop, then the value 223 * of memlimit_stop will be used for both. 224 */ 225 uint64_t memlimit_threading; 226 227 /** 228 * \brief Memory usage limit that should never be exceeded 229 * 230 * Encoder: Ignored. 231 * 232 * Decoder: If decompressing will need more than this amount of 233 * memory even in the single-threaded mode, then lzma_code() will 234 * return LZMA_MEMLIMIT_ERROR. 235 */ 236 uint64_t memlimit_stop; 237 238 /** \private Reserved member. */ 239 uint64_t reserved_int7; 240 241 /** \private Reserved member. */ 242 uint64_t reserved_int8; 243 244 /** \private Reserved member. */ 245 void *reserved_ptr1; 246 247 /** \private Reserved member. */ 248 void *reserved_ptr2; 249 250 /** \private Reserved member. */ 251 void *reserved_ptr3; 252 253 /** \private Reserved member. */ 254 void *reserved_ptr4; 255 256 } lzma_mt; 257 258 259 /** 260 * \brief Calculate approximate memory usage of easy encoder 261 * 262 * This function is a wrapper for lzma_raw_encoder_memusage(). 263 * 264 * \param preset Compression preset (level and possible flags) 265 * 266 * \return Number of bytes of memory required for the given 267 * preset when encoding or UINT64_MAX on error. 268 */ 269 extern LZMA_API(uint64_t) lzma_easy_encoder_memusage(uint32_t preset) 270 lzma_nothrow lzma_attr_pure; 271 272 273 /** 274 * \brief Calculate approximate decoder memory usage of a preset 275 * 276 * This function is a wrapper for lzma_raw_decoder_memusage(). 277 * 278 * \param preset Compression preset (level and possible flags) 279 * 280 * \return Number of bytes of memory required to decompress a file 281 * that was compressed using the given preset or UINT64_MAX 282 * on error. 283 */ 284 extern LZMA_API(uint64_t) lzma_easy_decoder_memusage(uint32_t preset) 285 lzma_nothrow lzma_attr_pure; 286 287 288 /** 289 * \brief Initialize .xz Stream encoder using a preset number 290 * 291 * This function is intended for those who just want to use the basic features 292 * of liblzma (that is, most developers out there). 293 * 294 * If initialization fails (return value is not LZMA_OK), all the memory 295 * allocated for *strm by liblzma is always freed. Thus, there is no need 296 * to call lzma_end() after failed initialization. 297 * 298 * If initialization succeeds, use lzma_code() to do the actual encoding. 299 * Valid values for 'action' (the second argument of lzma_code()) are 300 * LZMA_RUN, LZMA_SYNC_FLUSH, LZMA_FULL_FLUSH, and LZMA_FINISH. In future, 301 * there may be compression levels or flags that don't support LZMA_SYNC_FLUSH. 302 * 303 * \param strm Pointer to lzma_stream that is at least initialized 304 * with LZMA_STREAM_INIT. 305 * \param preset Compression preset to use. A preset consist of level 306 * number and zero or more flags. Usually flags aren't 307 * used, so preset is simply a number [0, 9] which match 308 * the options -0 ... -9 of the xz command line tool. 309 * Additional flags can be be set using bitwise-or with 310 * the preset level number, e.g. 6 | LZMA_PRESET_EXTREME. 311 * \param check Integrity check type to use. See check.h for available 312 * checks. The xz command line tool defaults to 313 * LZMA_CHECK_CRC64, which is a good choice if you are 314 * unsure. LZMA_CHECK_CRC32 is good too as long as the 315 * uncompressed file is not many gigabytes. 316 * 317 * \return Possible lzma_ret values: 318 * - LZMA_OK: Initialization succeeded. Use lzma_code() to 319 * encode your data. 320 * - LZMA_MEM_ERROR: Memory allocation failed. 321 * - LZMA_OPTIONS_ERROR: The given compression preset is not 322 * supported by this build of liblzma. 323 * - LZMA_UNSUPPORTED_CHECK: The given check type is not 324 * supported by this liblzma build. 325 * - LZMA_PROG_ERROR: One or more of the parameters have values 326 * that will never be valid. For example, strm == NULL. 327 */ 328 extern LZMA_API(lzma_ret) lzma_easy_encoder( 329 lzma_stream *strm, uint32_t preset, lzma_check check) 330 lzma_nothrow lzma_attr_warn_unused_result; 331 332 333 /** 334 * \brief Single-call .xz Stream encoding using a preset number 335 * 336 * The maximum required output buffer size can be calculated with 337 * lzma_stream_buffer_bound(). 338 * 339 * \param preset Compression preset to use. See the description 340 * in lzma_easy_encoder(). 341 * \param check Type of the integrity check to calculate from 342 * uncompressed data. 343 * \param allocator lzma_allocator for custom allocator functions. 344 * Set to NULL to use malloc() and free(). 345 * \param in Beginning of the input buffer 346 * \param in_size Size of the input buffer 347 * \param[out] out Beginning of the output buffer 348 * \param[out] out_pos The next byte will be written to out[*out_pos]. 349 * *out_pos is updated only if encoding succeeds. 350 * \param out_size Size of the out buffer; the first byte into 351 * which no data is written to is out[out_size]. 352 * 353 * \return Possible lzma_ret values: 354 * - LZMA_OK: Encoding was successful. 355 * - LZMA_BUF_ERROR: Not enough output buffer space. 356 * - LZMA_UNSUPPORTED_CHECK 357 * - LZMA_OPTIONS_ERROR 358 * - LZMA_MEM_ERROR 359 * - LZMA_DATA_ERROR 360 * - LZMA_PROG_ERROR 361 */ 362 extern LZMA_API(lzma_ret) lzma_easy_buffer_encode( 363 uint32_t preset, lzma_check check, 364 const lzma_allocator *allocator, 365 const uint8_t *in, size_t in_size, 366 uint8_t *out, size_t *out_pos, size_t out_size) lzma_nothrow; 367 368 369 /** 370 * \brief Initialize .xz Stream encoder using a custom filter chain 371 * 372 * \param strm Pointer to lzma_stream that is at least initialized 373 * with LZMA_STREAM_INIT. 374 * \param filters Array of filters terminated with 375 * .id == LZMA_VLI_UNKNOWN. See filters.h for more 376 * information. 377 * \param check Type of the integrity check to calculate from 378 * uncompressed data. 379 * 380 * \return Possible lzma_ret values: 381 * - LZMA_OK: Initialization was successful. 382 * - LZMA_MEM_ERROR 383 * - LZMA_UNSUPPORTED_CHECK 384 * - LZMA_OPTIONS_ERROR 385 * - LZMA_PROG_ERROR 386 */ 387 extern LZMA_API(lzma_ret) lzma_stream_encoder(lzma_stream *strm, 388 const lzma_filter *filters, lzma_check check) 389 lzma_nothrow lzma_attr_warn_unused_result; 390 391 392 /** 393 * \brief Calculate approximate memory usage of multithreaded .xz encoder 394 * 395 * Since doing the encoding in threaded mode doesn't affect the memory 396 * requirements of single-threaded decompressor, you can use 397 * lzma_easy_decoder_memusage(options->preset) or 398 * lzma_raw_decoder_memusage(options->filters) to calculate 399 * the decompressor memory requirements. 400 * 401 * \param options Compression options 402 * 403 * \return Number of bytes of memory required for encoding with the 404 * given options. If an error occurs, for example due to 405 * unsupported preset or filter chain, UINT64_MAX is returned. 406 */ 407 extern LZMA_API(uint64_t) lzma_stream_encoder_mt_memusage( 408 const lzma_mt *options) lzma_nothrow lzma_attr_pure; 409 410 411 /** 412 * \brief Initialize multithreaded .xz Stream encoder 413 * 414 * This provides the functionality of lzma_easy_encoder() and 415 * lzma_stream_encoder() as a single function for multithreaded use. 416 * 417 * The supported actions for lzma_code() are LZMA_RUN, LZMA_FULL_FLUSH, 418 * LZMA_FULL_BARRIER, and LZMA_FINISH. Support for LZMA_SYNC_FLUSH might be 419 * added in the future. 420 * 421 * \param strm Pointer to lzma_stream that is at least initialized 422 * with LZMA_STREAM_INIT. 423 * \param options Pointer to multithreaded compression options 424 * 425 * \return Possible lzma_ret values: 426 * - LZMA_OK 427 * - LZMA_MEM_ERROR 428 * - LZMA_UNSUPPORTED_CHECK 429 * - LZMA_OPTIONS_ERROR 430 * - LZMA_PROG_ERROR 431 */ 432 extern LZMA_API(lzma_ret) lzma_stream_encoder_mt( 433 lzma_stream *strm, const lzma_mt *options) 434 lzma_nothrow lzma_attr_warn_unused_result; 435 436 437 /** 438 * \brief Calculate recommended Block size for multithreaded .xz encoder 439 * 440 * This calculates a recommended Block size for multithreaded encoding given 441 * a filter chain. This is used internally by lzma_stream_encoder_mt() to 442 * determine the Block size if the block_size member is not set to the 443 * special value of 0 in the lzma_mt options struct. 444 * 445 * If one wishes to change the filters between Blocks, this function is 446 * helpful to set the block_size member of the lzma_mt struct before calling 447 * lzma_stream_encoder_mt(). Since the block_size member represents the 448 * maximum possible Block size for the multithreaded .xz encoder, one can 449 * use this function to find the maximum recommended Block size based on 450 * all planned filter chains. Otherwise, the multithreaded encoder will 451 * base its maximum Block size on the first filter chain used (if the 452 * block_size member is not set), which may unnecessarily limit the Block 453 * size for a later filter chain. 454 * 455 * \param filters Array of filters terminated with 456 * .id == LZMA_VLI_UNKNOWN. 457 * 458 * \return Recommended Block size in bytes, or UINT64_MAX if 459 * an error occurred. 460 */ 461 extern LZMA_API(uint64_t) lzma_mt_block_size(const lzma_filter *filters) 462 lzma_nothrow; 463 464 465 /** 466 * \brief Initialize .lzma encoder (legacy file format) 467 * 468 * The .lzma format is sometimes called the LZMA_Alone format, which is the 469 * reason for the name of this function. The .lzma format supports only the 470 * LZMA1 filter. There is no support for integrity checks like CRC32. 471 * 472 * Use this function if and only if you need to create files readable by 473 * legacy LZMA tools such as LZMA Utils 4.32.x. Moving to the .xz format 474 * is strongly recommended. 475 * 476 * The valid action values for lzma_code() are LZMA_RUN and LZMA_FINISH. 477 * No kind of flushing is supported, because the file format doesn't make 478 * it possible. 479 * 480 * \param strm Pointer to lzma_stream that is at least initialized 481 * with LZMA_STREAM_INIT. 482 * \param options Pointer to encoder options 483 * 484 * \return Possible lzma_ret values: 485 * - LZMA_OK 486 * - LZMA_MEM_ERROR 487 * - LZMA_OPTIONS_ERROR 488 * - LZMA_PROG_ERROR 489 */ 490 extern LZMA_API(lzma_ret) lzma_alone_encoder( 491 lzma_stream *strm, const lzma_options_lzma *options) 492 lzma_nothrow lzma_attr_warn_unused_result; 493 494 495 /** 496 * \brief Calculate output buffer size for single-call Stream encoder 497 * 498 * When trying to compress incompressible data, the encoded size will be 499 * slightly bigger than the input data. This function calculates how much 500 * output buffer space is required to be sure that lzma_stream_buffer_encode() 501 * doesn't return LZMA_BUF_ERROR. 502 * 503 * The calculated value is not exact, but it is guaranteed to be big enough. 504 * The actual maximum output space required may be slightly smaller (up to 505 * about 100 bytes). This should not be a problem in practice. 506 * 507 * If the calculated maximum size doesn't fit into size_t or would make the 508 * Stream grow past LZMA_VLI_MAX (which should never happen in practice), 509 * zero is returned to indicate the error. 510 * 511 * \note The limit calculated by this function applies only to 512 * single-call encoding. Multi-call encoding may (and probably 513 * will) have larger maximum expansion when encoding 514 * incompressible data. Currently there is no function to 515 * calculate the maximum expansion of multi-call encoding. 516 * 517 * \param uncompressed_size Size in bytes of the uncompressed 518 * input data 519 * 520 * \return Maximum number of bytes needed to store the compressed data. 521 */ 522 extern LZMA_API(size_t) lzma_stream_buffer_bound(size_t uncompressed_size) 523 lzma_nothrow; 524 525 526 /** 527 * \brief Single-call .xz Stream encoder 528 * 529 * \param filters Array of filters terminated with 530 * .id == LZMA_VLI_UNKNOWN. See filters.h for more 531 * information. 532 * \param check Type of the integrity check to calculate from 533 * uncompressed data. 534 * \param allocator lzma_allocator for custom allocator functions. 535 * Set to NULL to use malloc() and free(). 536 * \param in Beginning of the input buffer 537 * \param in_size Size of the input buffer 538 * \param[out] out Beginning of the output buffer 539 * \param[out] out_pos The next byte will be written to out[*out_pos]. 540 * *out_pos is updated only if encoding succeeds. 541 * \param out_size Size of the out buffer; the first byte into 542 * which no data is written to is out[out_size]. 543 * 544 * \return Possible lzma_ret values: 545 * - LZMA_OK: Encoding was successful. 546 * - LZMA_BUF_ERROR: Not enough output buffer space. 547 * - LZMA_UNSUPPORTED_CHECK 548 * - LZMA_OPTIONS_ERROR 549 * - LZMA_MEM_ERROR 550 * - LZMA_DATA_ERROR 551 * - LZMA_PROG_ERROR 552 */ 553 extern LZMA_API(lzma_ret) lzma_stream_buffer_encode( 554 lzma_filter *filters, lzma_check check, 555 const lzma_allocator *allocator, 556 const uint8_t *in, size_t in_size, 557 uint8_t *out, size_t *out_pos, size_t out_size) 558 lzma_nothrow lzma_attr_warn_unused_result; 559 560 561 /** 562 * \brief MicroLZMA encoder 563 * 564 * The MicroLZMA format is a raw LZMA stream whose first byte (always 0x00) 565 * has been replaced with bitwise-negation of the LZMA properties (lc/lp/pb). 566 * This encoding ensures that the first byte of MicroLZMA stream is never 567 * 0x00. There is no end of payload marker and thus the uncompressed size 568 * must be stored separately. For the best error detection the dictionary 569 * size should be stored separately as well but alternatively one may use 570 * the uncompressed size as the dictionary size when decoding. 571 * 572 * With the MicroLZMA encoder, lzma_code() behaves slightly unusually. 573 * The action argument must be LZMA_FINISH and the return value will never be 574 * LZMA_OK. Thus the encoding is always done with a single lzma_code() after 575 * the initialization. The benefit of the combination of initialization 576 * function and lzma_code() is that memory allocations can be re-used for 577 * better performance. 578 * 579 * lzma_code() will try to encode as much input as is possible to fit into 580 * the given output buffer. If not all input can be encoded, the stream will 581 * be finished without encoding all the input. The caller must check both 582 * input and output buffer usage after lzma_code() (total_in and total_out 583 * in lzma_stream can be convenient). Often lzma_code() can fill the output 584 * buffer completely if there is a lot of input, but sometimes a few bytes 585 * may remain unused because the next LZMA symbol would require more space. 586 * 587 * lzma_stream.avail_out must be at least 6. Otherwise LZMA_PROG_ERROR 588 * will be returned. 589 * 590 * The LZMA dictionary should be reasonably low to speed up the encoder 591 * re-initialization. A good value is bigger than the resulting 592 * uncompressed size of most of the output chunks. For example, if output 593 * size is 4 KiB, dictionary size of 32 KiB or 64 KiB is good. If the 594 * data compresses extremely well, even 128 KiB may be useful. 595 * 596 * The MicroLZMA format and this encoder variant were made with the EROFS 597 * file system in mind. This format may be convenient in other embedded 598 * uses too where many small streams are needed. XZ Embedded includes a 599 * decoder for this format. 600 * 601 * \param strm Pointer to lzma_stream that is at least initialized 602 * with LZMA_STREAM_INIT. 603 * \param options Pointer to encoder options 604 * 605 * \return Possible lzma_ret values: 606 * - LZMA_STREAM_END: All good. Check the amounts of input used 607 * and output produced. Store the amount of input used 608 * (uncompressed size) as it needs to be known to decompress 609 * the data. 610 * - LZMA_OPTIONS_ERROR 611 * - LZMA_MEM_ERROR 612 * - LZMA_PROG_ERROR: In addition to the generic reasons for this 613 * error code, this may also be returned if there isn't enough 614 * output space (6 bytes) to create a valid MicroLZMA stream. 615 */ 616 extern LZMA_API(lzma_ret) lzma_microlzma_encoder( 617 lzma_stream *strm, const lzma_options_lzma *options) 618 lzma_nothrow; 619 620 621 /************ 622 * Decoding * 623 ************/ 624 625 /** 626 * This flag makes lzma_code() return LZMA_NO_CHECK if the input stream 627 * being decoded has no integrity check. Note that when used with 628 * lzma_auto_decoder(), all .lzma files will trigger LZMA_NO_CHECK 629 * if LZMA_TELL_NO_CHECK is used. 630 */ 631 #define LZMA_TELL_NO_CHECK UINT32_C(0x01) 632 633 634 /** 635 * This flag makes lzma_code() return LZMA_UNSUPPORTED_CHECK if the input 636 * stream has an integrity check, but the type of the integrity check is not 637 * supported by this liblzma version or build. Such files can still be 638 * decoded, but the integrity check cannot be verified. 639 */ 640 #define LZMA_TELL_UNSUPPORTED_CHECK UINT32_C(0x02) 641 642 643 /** 644 * This flag makes lzma_code() return LZMA_GET_CHECK as soon as the type 645 * of the integrity check is known. The type can then be got with 646 * lzma_get_check(). 647 */ 648 #define LZMA_TELL_ANY_CHECK UINT32_C(0x04) 649 650 651 /** 652 * This flag makes lzma_code() not calculate and verify the integrity check 653 * of the compressed data in .xz files. This means that invalid integrity 654 * check values won't be detected and LZMA_DATA_ERROR won't be returned in 655 * such cases. 656 * 657 * This flag only affects the checks of the compressed data itself; the CRC32 658 * values in the .xz headers will still be verified normally. 659 * 660 * Don't use this flag unless you know what you are doing. Possible reasons 661 * to use this flag: 662 * 663 * - Trying to recover data from a corrupt .xz file. 664 * 665 * - Speeding up decompression, which matters mostly with SHA-256 666 * or with files that have compressed extremely well. It's recommended 667 * to not use this flag for this purpose unless the file integrity is 668 * verified externally in some other way. 669 * 670 * Support for this flag was added in liblzma 5.1.4beta. 671 */ 672 #define LZMA_IGNORE_CHECK UINT32_C(0x10) 673 674 675 /** 676 * This flag enables decoding of concatenated files with file formats that 677 * allow concatenating compressed files as is. From the formats currently 678 * supported by liblzma, only the .xz and .lz formats allow concatenated 679 * files. Concatenated files are not allowed with the legacy .lzma format. 680 * 681 * This flag also affects the usage of the 'action' argument for lzma_code(). 682 * When LZMA_CONCATENATED is used, lzma_code() won't return LZMA_STREAM_END 683 * unless LZMA_FINISH is used as 'action'. Thus, the application has to set 684 * LZMA_FINISH in the same way as it does when encoding. 685 * 686 * If LZMA_CONCATENATED is not used, the decoders still accept LZMA_FINISH 687 * as 'action' for lzma_code(), but the usage of LZMA_FINISH isn't required. 688 */ 689 #define LZMA_CONCATENATED UINT32_C(0x08) 690 691 692 /** 693 * This flag makes the threaded decoder report errors (like LZMA_DATA_ERROR) 694 * as soon as they are detected. This saves time when the application has no 695 * interest in a partially decompressed truncated or corrupt file. Note that 696 * due to timing randomness, if the same truncated or corrupt input is 697 * decompressed multiple times with this flag, a different amount of output 698 * may be produced by different runs, and even the error code might vary. 699 * 700 * When using LZMA_FAIL_FAST, it is recommended to use LZMA_FINISH to tell 701 * the decoder when no more input will be coming because it can help fast 702 * detection and reporting of truncated files. Note that in this situation 703 * truncated files might be diagnosed with LZMA_DATA_ERROR instead of 704 * LZMA_OK or LZMA_BUF_ERROR! 705 * 706 * Without this flag the threaded decoder will provide as much output as 707 * possible at first and then report the pending error. This default behavior 708 * matches the single-threaded decoder and provides repeatable behavior 709 * with truncated or corrupt input. There are a few special cases where the 710 * behavior can still differ like memory allocation failures (LZMA_MEM_ERROR). 711 * 712 * Single-threaded decoders currently ignore this flag. 713 * 714 * Support for this flag was added in liblzma 5.3.3alpha. Note that in older 715 * versions this flag isn't supported (LZMA_OPTIONS_ERROR) even by functions 716 * that ignore this flag in newer liblzma versions. 717 */ 718 #define LZMA_FAIL_FAST UINT32_C(0x20) 719 720 721 /** 722 * \brief Initialize .xz Stream decoder 723 * 724 * \param strm Pointer to lzma_stream that is at least initialized 725 * with LZMA_STREAM_INIT. 726 * \param memlimit Memory usage limit as bytes. Use UINT64_MAX 727 * to effectively disable the limiter. liblzma 728 * 5.2.3 and earlier don't allow 0 here and return 729 * LZMA_PROG_ERROR; later versions treat 0 as if 1 730 * had been specified. 731 * \param flags Bitwise-or of zero or more of the decoder flags: 732 * LZMA_TELL_NO_CHECK, LZMA_TELL_UNSUPPORTED_CHECK, 733 * LZMA_TELL_ANY_CHECK, LZMA_IGNORE_CHECK, 734 * LZMA_CONCATENATED, LZMA_FAIL_FAST 735 * 736 * \return Possible lzma_ret values: 737 * - LZMA_OK: Initialization was successful. 738 * - LZMA_MEM_ERROR: Cannot allocate memory. 739 * - LZMA_OPTIONS_ERROR: Unsupported flags 740 * - LZMA_PROG_ERROR 741 */ 742 extern LZMA_API(lzma_ret) lzma_stream_decoder( 743 lzma_stream *strm, uint64_t memlimit, uint32_t flags) 744 lzma_nothrow lzma_attr_warn_unused_result; 745 746 747 /** 748 * \brief Initialize multithreaded .xz Stream decoder 749 * 750 * The decoder can decode multiple Blocks in parallel. This requires that each 751 * Block Header contains the Compressed Size and Uncompressed size fields 752 * which are added by the multi-threaded encoder, see lzma_stream_encoder_mt(). 753 * 754 * A Stream with one Block will only utilize one thread. A Stream with multiple 755 * Blocks but without size information in Block Headers will be processed in 756 * single-threaded mode in the same way as done by lzma_stream_decoder(). 757 * Concatenated Streams are processed one Stream at a time; no inter-Stream 758 * parallelization is done. 759 * 760 * This function behaves like lzma_stream_decoder() when options->threads == 1 761 * and options->memlimit_threading <= 1. 762 * 763 * \param strm Pointer to lzma_stream that is at least initialized 764 * with LZMA_STREAM_INIT. 765 * \param options Pointer to multithreaded compression options 766 * 767 * \return Possible lzma_ret values: 768 * - LZMA_OK: Initialization was successful. 769 * - LZMA_MEM_ERROR: Cannot allocate memory. 770 * - LZMA_MEMLIMIT_ERROR: Memory usage limit was reached. 771 * - LZMA_OPTIONS_ERROR: Unsupported flags. 772 * - LZMA_PROG_ERROR 773 */ 774 extern LZMA_API(lzma_ret) lzma_stream_decoder_mt( 775 lzma_stream *strm, const lzma_mt *options) 776 lzma_nothrow lzma_attr_warn_unused_result; 777 778 779 /** 780 * \brief Decode .xz, .lzma, and .lz (lzip) files with autodetection 781 * 782 * This decoder autodetects between the .xz, .lzma, and .lz file formats, 783 * and calls lzma_stream_decoder(), lzma_alone_decoder(), or 784 * lzma_lzip_decoder() once the type of the input file has been detected. 785 * 786 * Support for .lz was added in 5.4.0. 787 * 788 * If the flag LZMA_CONCATENATED is used and the input is a .lzma file: 789 * For historical reasons concatenated .lzma files aren't supported. 790 * If there is trailing data after one .lzma stream, lzma_code() will 791 * return LZMA_DATA_ERROR. (lzma_alone_decoder() doesn't have such a check 792 * as it doesn't support any decoder flags. It will return LZMA_STREAM_END 793 * after one .lzma stream.) 794 * 795 * \param strm Pointer to lzma_stream that is at least initialized 796 * with LZMA_STREAM_INIT. 797 * \param memlimit Memory usage limit as bytes. Use UINT64_MAX 798 * to effectively disable the limiter. liblzma 799 * 5.2.3 and earlier don't allow 0 here and return 800 * LZMA_PROG_ERROR; later versions treat 0 as if 1 801 * had been specified. 802 * \param flags Bitwise-or of zero or more of the decoder flags: 803 * LZMA_TELL_NO_CHECK, LZMA_TELL_UNSUPPORTED_CHECK, 804 * LZMA_TELL_ANY_CHECK, LZMA_IGNORE_CHECK, 805 * LZMA_CONCATENATED, LZMA_FAIL_FAST 806 * 807 * \return Possible lzma_ret values: 808 * - LZMA_OK: Initialization was successful. 809 * - LZMA_MEM_ERROR: Cannot allocate memory. 810 * - LZMA_OPTIONS_ERROR: Unsupported flags 811 * - LZMA_PROG_ERROR 812 */ 813 extern LZMA_API(lzma_ret) lzma_auto_decoder( 814 lzma_stream *strm, uint64_t memlimit, uint32_t flags) 815 lzma_nothrow lzma_attr_warn_unused_result; 816 817 818 /** 819 * \brief Initialize .lzma decoder (legacy file format) 820 * 821 * Valid 'action' arguments to lzma_code() are LZMA_RUN and LZMA_FINISH. 822 * There is no need to use LZMA_FINISH, but it's allowed because it may 823 * simplify certain types of applications. 824 * 825 * \param strm Pointer to lzma_stream that is at least initialized 826 * with LZMA_STREAM_INIT. 827 * \param memlimit Memory usage limit as bytes. Use UINT64_MAX 828 * to effectively disable the limiter. liblzma 829 * 5.2.3 and earlier don't allow 0 here and return 830 * LZMA_PROG_ERROR; later versions treat 0 as if 1 831 * had been specified. 832 * 833 * \return Possible lzma_ret values: 834 * - LZMA_OK 835 * - LZMA_MEM_ERROR 836 * - LZMA_PROG_ERROR 837 */ 838 extern LZMA_API(lzma_ret) lzma_alone_decoder( 839 lzma_stream *strm, uint64_t memlimit) 840 lzma_nothrow lzma_attr_warn_unused_result; 841 842 843 /** 844 * \brief Initialize .lz (lzip) decoder (a foreign file format) 845 * 846 * This decoder supports the .lz format version 0 and the unextended .lz 847 * format version 1: 848 * 849 * - Files in the format version 0 were produced by lzip 1.3 and older. 850 * Such files aren't common but may be found from file archives 851 * as a few source packages were released in this format. People 852 * might have old personal files in this format too. Decompression 853 * support for the format version 0 was removed in lzip 1.18. 854 * 855 * - lzip 1.3 added decompression support for .lz format version 1 files. 856 * Compression support was added in lzip 1.4. In lzip 1.6 the .lz format 857 * version 1 was extended to support the Sync Flush marker. This extension 858 * is not supported by liblzma. lzma_code() will return LZMA_DATA_ERROR 859 * at the location of the Sync Flush marker. In practice files with 860 * the Sync Flush marker are very rare and thus liblzma can decompress 861 * almost all .lz files. 862 * 863 * Just like with lzma_stream_decoder() for .xz files, LZMA_CONCATENATED 864 * should be used when decompressing normal standalone .lz files. 865 * 866 * The .lz format allows putting non-.lz data at the end of a file after at 867 * least one valid .lz member. That is, one can append custom data at the end 868 * of a .lz file and the decoder is required to ignore it. In liblzma this 869 * is relevant only when LZMA_CONCATENATED is used. In that case lzma_code() 870 * will return LZMA_STREAM_END and leave lzma_stream.next_in pointing to 871 * the first byte of the non-.lz data. An exception to this is if the first 872 * 1-3 bytes of the non-.lz data are identical to the .lz magic bytes 873 * (0x4C, 0x5A, 0x49, 0x50; "LZIP" in US-ASCII). In such a case the 1-3 bytes 874 * will have been ignored by lzma_code(). If one wishes to locate the non-.lz 875 * data reliably, one must ensure that the first byte isn't 0x4C. Actually 876 * one should ensure that none of the first four bytes of trailing data are 877 * equal to the magic bytes because lzip >= 1.20 requires it by default. 878 * 879 * \param strm Pointer to lzma_stream that is at least initialized 880 * with LZMA_STREAM_INIT. 881 * \param memlimit Memory usage limit as bytes. Use UINT64_MAX 882 * to effectively disable the limiter. 883 * \param flags Bitwise-or of flags, or zero for no flags. 884 * All decoder flags listed above are supported 885 * although only LZMA_CONCATENATED and (in very rare 886 * cases) LZMA_IGNORE_CHECK are actually useful. 887 * LZMA_TELL_NO_CHECK, LZMA_TELL_UNSUPPORTED_CHECK, 888 * and LZMA_FAIL_FAST do nothing. LZMA_TELL_ANY_CHECK 889 * is supported for consistency only as CRC32 is 890 * always used in the .lz format. 891 * 892 * \return Possible lzma_ret values: 893 * - LZMA_OK: Initialization was successful. 894 * - LZMA_MEM_ERROR: Cannot allocate memory. 895 * - LZMA_OPTIONS_ERROR: Unsupported flags 896 * - LZMA_PROG_ERROR 897 */ 898 extern LZMA_API(lzma_ret) lzma_lzip_decoder( 899 lzma_stream *strm, uint64_t memlimit, uint32_t flags) 900 lzma_nothrow lzma_attr_warn_unused_result; 901 902 903 /** 904 * \brief Single-call .xz Stream decoder 905 * 906 * \param memlimit Pointer to how much memory the decoder is allowed 907 * to allocate. The value pointed by this pointer is 908 * modified if and only if LZMA_MEMLIMIT_ERROR is 909 * returned. 910 * \param flags Bitwise-or of zero or more of the decoder flags: 911 * LZMA_TELL_NO_CHECK, LZMA_TELL_UNSUPPORTED_CHECK, 912 * LZMA_IGNORE_CHECK, LZMA_CONCATENATED, 913 * LZMA_FAIL_FAST. Note that LZMA_TELL_ANY_CHECK 914 * is not allowed and will return LZMA_PROG_ERROR. 915 * \param allocator lzma_allocator for custom allocator functions. 916 * Set to NULL to use malloc() and free(). 917 * \param in Beginning of the input buffer 918 * \param in_pos The next byte will be read from in[*in_pos]. 919 * *in_pos is updated only if decoding succeeds. 920 * \param in_size Size of the input buffer; the first byte that 921 * won't be read is in[in_size]. 922 * \param[out] out Beginning of the output buffer 923 * \param[out] out_pos The next byte will be written to out[*out_pos]. 924 * *out_pos is updated only if decoding succeeds. 925 * \param out_size Size of the out buffer; the first byte into 926 * which no data is written to is out[out_size]. 927 * 928 * \return Possible lzma_ret values: 929 * - LZMA_OK: Decoding was successful. 930 * - LZMA_FORMAT_ERROR 931 * - LZMA_OPTIONS_ERROR 932 * - LZMA_DATA_ERROR 933 * - LZMA_NO_CHECK: This can be returned only if using 934 * the LZMA_TELL_NO_CHECK flag. 935 * - LZMA_UNSUPPORTED_CHECK: This can be returned only if using 936 * the LZMA_TELL_UNSUPPORTED_CHECK flag. 937 * - LZMA_MEM_ERROR 938 * - LZMA_MEMLIMIT_ERROR: Memory usage limit was reached. 939 * The minimum required memlimit value was stored to *memlimit. 940 * - LZMA_BUF_ERROR: Output buffer was too small. 941 * - LZMA_PROG_ERROR 942 */ 943 extern LZMA_API(lzma_ret) lzma_stream_buffer_decode( 944 uint64_t *memlimit, uint32_t flags, 945 const lzma_allocator *allocator, 946 const uint8_t *in, size_t *in_pos, size_t in_size, 947 uint8_t *out, size_t *out_pos, size_t out_size) 948 lzma_nothrow lzma_attr_warn_unused_result; 949 950 951 /** 952 * \brief MicroLZMA decoder 953 * 954 * See lzma_microlzma_encoder() for more information. 955 * 956 * The lzma_code() usage with this decoder is completely normal. The 957 * special behavior of lzma_code() applies to lzma_microlzma_encoder() only. 958 * 959 * \param strm Pointer to lzma_stream that is at least initialized 960 * with LZMA_STREAM_INIT. 961 * \param comp_size Compressed size of the MicroLZMA stream. 962 * The caller must somehow know this exactly. 963 * \param uncomp_size Uncompressed size of the MicroLZMA stream. 964 * If the exact uncompressed size isn't known, this 965 * can be set to a value that is at most as big as 966 * the exact uncompressed size would be, but then the 967 * next argument uncomp_size_is_exact must be false. 968 * \param uncomp_size_is_exact 969 * If true, uncomp_size must be exactly correct. 970 * This will improve error detection at the end of 971 * the stream. If the exact uncompressed size isn't 972 * known, this must be false. uncomp_size must still 973 * be at most as big as the exact uncompressed size 974 * is. Setting this to false when the exact size is 975 * known will work but error detection at the end of 976 * the stream will be weaker. 977 * \param dict_size LZMA dictionary size that was used when 978 * compressing the data. It is OK to use a bigger 979 * value too but liblzma will then allocate more 980 * memory than would actually be required and error 981 * detection will be slightly worse. (Note that with 982 * the implementation in XZ Embedded it doesn't 983 * affect the memory usage if one specifies bigger 984 * dictionary than actually required.) 985 * 986 * \return Possible lzma_ret values: 987 * - LZMA_OK 988 * - LZMA_MEM_ERROR 989 * - LZMA_OPTIONS_ERROR 990 * - LZMA_PROG_ERROR 991 */ 992 extern LZMA_API(lzma_ret) lzma_microlzma_decoder( 993 lzma_stream *strm, uint64_t comp_size, 994 uint64_t uncomp_size, lzma_bool uncomp_size_is_exact, 995 uint32_t dict_size) lzma_nothrow; 996