/** * \file lzma/base.h * \brief Data types and functions used in many places of the public API * * \author Copyright (C) 1999-2006 Igor Pavlov * \author Copyright (C) 2007 Lasse Collin * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2.1 of the License, or (at your option) any later version. * * This library is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. */ #ifndef LZMA_H_INTERNAL # error Never include this file directly. Use instead. #endif /** * \brief Boolean * * This is here because C89 doesn't have stdbool.h. To set a value for * variables having type lzma_bool, you can use * - C99's `true' and `false' from stdbool.h; * - C++'s internal `true' and `false'; or * - integers one (true) and zero (false). */ typedef unsigned char lzma_bool; /** * \brief Type of reserved enumeration variable in structures * * To avoid breaking library ABI when new features are added, several * structures contain extra variables that may be used in future. Since * sizeof(enum) can be different than sizeof(int), and sizeof(enum) may * even vary depending on the range of enumeration constants, we specify * a separate type to be used for reserved enumeration variables. All * enumeration constants in liblzma API will be non-negative and less * than 128, which should guarantee that the ABI won't break even when * new constants are added to existing enumerations. */ typedef enum { LZMA_RESERVED_ENUM = 0 } lzma_reserved_enum; /** * \brief Return values used by several functions in liblzma * * Check the descriptions of specific functions to find out which return * values they can return. With some functions the return values may have * more specific meanings than described here; those differences are * described per-function basis. */ typedef enum { LZMA_OK = 0, /**< * \brief Operation completed successfully */ LZMA_STREAM_END = 1, /**< * \brief End of stream was reached * * In encoder, LZMA_SYNC_FLUSH, LZMA_FULL_FLUSH, or * LZMA_FINISH was finished. In decoder, this indicates * that all the data was successfully decoded. * * In all cases, when LZMA_STREAM_END is returned, the last * output bytes should be picked from strm->next_out. */ LZMA_NO_CHECK = 2, /**< * \brief Input stream has no integrity check * * This return value can be returned only if the * LZMA_TELL_NO_CHECK flag was used when initializing * the decoder. LZMA_NO_CHECK is just a warning, and * the decoding can be continued normally. * * It is possible to call lzma_get_check() immediatelly after * lzma_code has returned LZMA_NO_CHECK. The result will * naturally be LZMA_CHECK_NONE, but the possibility to call * lzma_get_check() may be convenient in some applications. */ LZMA_UNSUPPORTED_CHECK = 3, /**< * \brief Cannot calculate the integrity check * * The usage of this return value is slightly different in * encoders and decoders. * * Encoders can return this value only from the initialization * function. If initialization fails with this value, the * encoding cannot be done, because there's no way to produce * output with the correct integrity check. * * Decoders can return this value only from the lzma_code * function and only if the LZMA_TELL_UNSUPPORTED_CHECK flag * was used when initializing the decoder. The decoding can * still be continued normally even if the check type is * unsupported, but naturally the check will not be validated, * and possible errors may go undetected. * * With decoder, it is possible to call lzma_get_check() * immediatelly after lzma_code has returned * LZMA_UNSUPPORTED_CHECK. This way it is possible to find * out what the unsupported Check ID was. */ LZMA_GET_CHECK = 4, /**< * \brief Integrity check type is now available * * This value can be returned only by the lzma_code() function * and only if the decoder was initialized with the * LZMA_TELL_ANY_CHECK flag. LZMA_GET_CHECK tells the * application that it may now call lzma_get_check() to find * out the Check ID. This can be used, for example, to * implement a decoder that accepts only files that have * strong enough integrity check. */ LZMA_MEM_ERROR = 5, /**< * \brief Cannot allocate memory * * Memory allocation failed, or the size of the allocation * would be greater than SIZE_MAX. * * Due to lazy coding, the coding cannot be continued even * if more memory were made available after LZMA_MEM_ERROR. */ LZMA_MEMLIMIT_ERROR = 6, /** * \brief Memory usage limit was reached * * Decoder would need more memory than allowed by the * specified memory usage limit. To continue decoding, * the memory usage limit has to be increased. See functions * lzma_memlimit_get() and lzma_memlimit_set(). */ LZMA_FORMAT_ERROR = 7, /**< * \brief Unknown file format * * The decoder did not recognize the input as supported file * format. This error can occur, for example, when trying to * decode LZMA_Alone format file with lzma_stream_decoder, * because lzma_stream_decoder accepts only the new .lzma * format. */ LZMA_HEADER_ERROR = 8, /**< * \brief Invalid or unsupported options * * Invalid or unsupported options, for example * - unsupported filter(s) or filter options; or * - reserved bits set in headers (decoder only). * * Rebuilding liblzma with more features enabled, or * upgrading to a newer version of liblzma may help. */ LZMA_DATA_ERROR = 9, /**< * \brief Data is corrupt * * The usage of this return value is different in encoders * and decoders. In both encoder and decoder, the coding * cannot continue after this error. * * Encoders return this if size limits of the target file * format would be exceeded. These limits are huge, thus * getting this error from an encoder is mostly theoretical. * For example, the maximum compressed and uncompressed * size of a Stream created with lzma_stream_encoder is * 2^63 - 1 bytes (one byte less than 8 EiB). * * Decoders return this error if the input data is corrupt. * This can mean, for example, invalid CRC32 in headers * or invalid check of uncompressed data. */ LZMA_BUF_ERROR = 10, /**< * \brief No progress is possible * * This error code is returned when the coder cannot consume * any new input and produce any new output. The most common * reason for this error is that the input stream being * decoded is truncated or corrupt. * * This error is not fatal. Coding can be continued normally * by providing more input and/or more output space, if * possible. * * Typically the first call to lzma_code() that can do no * progress returns LZMA_OK instead of LZMA_BUF_ERROR. Only * the second consecutive call doing no progress will return * LZMA_BUF_ERROR. This is by design. * * With zlib, Z_BUF_ERROR may be returned even if the * application is doing nothing wrong. The above hack * guarantees that liblzma never returns LZMA_BUF_ERROR * to properly written applications unless the input file * is truncated or corrupt. This should simplify the * applications a little. */ LZMA_PROG_ERROR = 11, /**< * \brief Programming error * * This indicates that the arguments given to the function are * invalid or the internal state of the decoder is corrupt. * - Function arguments are invalid or the structures * pointed by the argument pointers are invalid * e.g. if strm->next_out has been set to NULL and * strm->avail_out > 0 when calling lzma_code(). * - lzma_* functions have been called in wrong order * e.g. lzma_code() was called right after lzma_end(). * - If errors occur randomly, the reason might be flaky * hardware. * * If you think that your code is correct, this error code * can be a sign of a bug in liblzma. See the documentation * how to report bugs. */ } lzma_ret; /** * \brief The `action' argument for lzma_code() */ typedef enum { LZMA_RUN = 0, /**< * Encoder: Encode as much input as possible. Some internal * buffering will probably be done (depends on the filter * chain in use), which causes latency: the input used won't * usually be decodeable from the output of the same * lzma_code() call. * * Decoder: Decode as much input as possible and produce as * much output as possible. This action provides best * throughput, but may introduce latency, because the * decoder may decode more data into its internal buffers * than that fits into next_out. */ LZMA_SYNC_FLUSH = 1, /**< * Encoder: Makes all the data given to liblzma via next_in * available in next_out without resetting the filters. Call * lzma_code() with LZMA_SYNC_FLUSH until it returns * LZMA_STREAM_END. Then continue encoding normally. * * \note Synchronous flushing is supported only by * some filters. Using LZMA_SYNC_FLUSH with * which such filters will make lzma_code() * return LZMA_HEADER_ERROR. * * Decoders don't support LZMA_SYNC_FLUSH. */ LZMA_FULL_FLUSH = 2, /**< * Finishes encoding of the current Data Block. All the input * data going to the current Data Block must have been given * to the encoder (the last bytes can still be pending in * next_in). Call lzma_code() with LZMA_FULL_FLUSH until * it returns LZMA_STREAM_END. Then continue normally with * LZMA_RUN or finish the Stream with LZMA_FINISH. * * This action is supported only by Stream encoder and easy * encoder (which uses Stream encoder). If there is no * unfinished Block, no empty Block is created. */ LZMA_FINISH = 3 /**< * Finishes the coding operation. All the input data must * have been given to the encoder (the last bytes can still * be pending in next_in). Call lzma_code() with LZMA_FINISH * until it returns LZMA_STREAM_END. Once LZMA_FINISH has * been used, the amount of input must no longer be changed * by the application. * * When decoding, using LZMA_FINISH is optional unless the * LZMA_CONCATENATED flag was used when the decoder was * initialized. When LZMA_CONCATENATED was not used, the only * effect of LZMA_FINISH is that the amount of input must not * be changed just like in the encoder. */ } lzma_action; /** * \brief Custom functions for memory handling * * A pointer to lzma_allocator may be passed via lzma_stream structure * to liblzma, and some advanced function take pointer lzma_allocator as * a separate function argument. The library will use the functions * specified in lzma_allocator for memory handling instead of the default * malloc() and free(). * * liblzma doesn't make an internal copy of lzma_allocator. Thus, it is * OK to change these function pointers in the middle of the coding * process, but obviously it must be done carefully to make sure that the * replacement `free' can deallocate memory allocated by the earlier * `alloc' function(s). */ typedef struct { /** * \brief Pointer to custom memory allocation function * * If you don't want a custom allocator, but still want * custom free(), set this to NULL and liblzma will use * the standard malloc(). * * \param opaque lzma_allocator.opaque (see below) * \param nmemb Number of elements like in calloc(). * liblzma will always set nmemb to 1. * This argument exists only for * compatibility with zlib and libbzip2. * \param size Size of an element in bytes. * liblzma never sets this to zero. * * \return Pointer to the beginning of a memory block of * size nmemb * size, or NULL if allocation fails * for some reason. When allocation fails, functions * of liblzma return LZMA_MEM_ERROR. * * For performance reasons, the allocator should not waste time * zeroing the allocated buffers. This is not only about speed, but * also memory usage, since the operating system kernel doesn't * necessarily allocate the requested memory until it is actually * used. With small input files liblzma may actually need only a * fraction of the memory that it requested for allocation. * * \note LZMA_MEM_ERROR is also used when the size of the * allocation would be greater than SIZE_MAX. Thus, * don't assume that the custom allocator must have * returned NULL if some function from liblzma * returns LZMA_MEM_ERROR. */ void *(*alloc)(void *opaque, size_t nmemb, size_t size); /** * \brief Pointer to custom memory freeing function * * If you don't want a custom freeing function, but still * want a custom allocator, set this to NULL and liblzma * will use the standard free(). * * \param opaque lzma_allocator.opaque (see below) * \param ptr Pointer returned by lzma_allocator.alloc(), * or when it is set to NULL, a pointer returned * by the standard malloc(). */ void (*free)(void *opaque, void *ptr); /** * \brief Pointer passed to .alloc() and .free() * * opaque is passed as the first argument to lzma_allocator.alloc() * and lzma_allocator.free(). This intended to ease implementing * custom memory allocation functions for use with liblzma. * * If you don't need this, you should set this to NULL. */ void *opaque; } lzma_allocator; /** * \brief Internal data structure * * The contents of this structure is not visible outside the library. */ typedef struct lzma_internal_s lzma_internal; /** * \brief Passing data to and from liblzma * * The lzma_stream structure is used for * - passing pointers to input and output buffers to liblzma; * - defining custom memory hander functions; and * - holding a pointer to coder-specific internal data structures. * * When a new lzma_stream structure is allocated (either as automatic variable * on stack or dynamically with malloc()), the new lzma_stream structure must * be initialized to LZMA_STREAM_INIT. * * Before initializing a coder (for example, with lzma_stream_decoder()), * * * Before calling any of the lzma_*_init() functions the first time, * the application must reset lzma_stream to LZMA_STREAM_INIT. The * lzma_*_init() function will verify the options, allocate internal * data structures and store pointer to them into `internal'. Finally * total_in and total_out are reset to zero. In contrast to zlib, * next_in and avail_in are ignored by the initialization functions. * * The actual coding is done with the lzma_code() function. Application * must update next_in, avail_in, next_out, and avail_out between * calls to lzma_decode() just like with zlib. * * In contrast to zlib, even the decoder requires that there always * is at least one byte space in next_out; if avail_out == 0, * LZMA_BUF_ERROR is returned immediatelly. This shouldn't be a problem * for most applications that already use zlib, but it's still worth * checking your application. * * Application may modify values of total_in and total_out as it wants. * They are updated by liblzma to match the amount of data read and * written, but liblzma doesn't use the values internally. * * Application must not touch the `internal' pointer. */ typedef struct { const uint8_t *next_in; /**< Pointer to the next input byte. */ size_t avail_in; /**< Number of available input bytes in next_in. */ uint64_t total_in; /**< Total number of bytes read by liblzma. */ uint8_t *next_out; /**< Pointer to the next output position. */ size_t avail_out; /**< Amount of free space in next_out. */ uint64_t total_out; /**< Total number of bytes written by liblzma. */ /** * Custom memory allocation functions. Set to NULL to use * the standard malloc() and free(). */ lzma_allocator *allocator; /** Internal state is not visible to applications. */ lzma_internal *internal; /** * Reserved space to allow possible future extensions without * breaking the ABI. Excluding the initialization of this structure, * you should not touch these, because the names of these variables * may change. */ void *reserved_ptr1; void *reserved_ptr2; uint64_t reserved_int1; uint64_t reserved_int2; lzma_reserved_enum reserved_enum1; lzma_reserved_enum reserved_enum2; } lzma_stream; /** * \brief Initialization for lzma_stream * * When you declare an instance of lzma_stream, you can immediatelly * initialize it so that initialization functions know that no memory * has been allocated yet: * * lzma_stream strm = LZMA_STREAM_INIT; * * If you need to initialize a dynamically allocated lzma_stream, you can use * memset(strm_pointer, 0, sizeof(lzma_stream)). Strictly speaking, this * violates the C standard since NULL may have different internal * representation than zero, but it should be portable enough in practice. * Anyway, for maximum portability, you can use something like this: * * lzma_stream tmp = LZMA_STREAM_INIT; * *strm = tmp; */ #define LZMA_STREAM_INIT \ { NULL, 0, 0, NULL, 0, 0, NULL, NULL, NULL, NULL, 0, 0, 0, 0 } /** * \brief Encodes or decodes data * * Once the lzma_stream has been successfully initialized (e.g. with * lzma_stream_encoder()), the actual encoding or decoding is done * using this function. The application has to update strm->next_in, * strm->avail_in, strm->next_out, and strm->avail_out to pass input * to and get output from liblzma. * * See the description of the coder-specific initialization function to find * out what `action' values are supported by the coder. See documentation of * lzma_ret for the possible return values. */ extern lzma_ret lzma_code(lzma_stream *strm, lzma_action action) lzma_attr_warn_unused_result; /** * \brief Frees memory allocated for the coder data structures * * \param strm Pointer to lzma_stream that is at least initialized * with LZMA_STREAM_INIT. * * \note zlib indicates an error if application end()s unfinished * stream. liblzma doesn't do this, and assumes that * application knows what it is doing. */ extern void lzma_end(lzma_stream *strm);