1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/other-licenses/7zstub/src/DOC/lzma.txt Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,630 @@
1.4 +LZMA SDK 4.40
1.5 +-------------
1.6 +
1.7 +LZMA SDK Copyright (C) 1999-2006 Igor Pavlov
1.8 +
1.9 +LZMA SDK provides the documentation, samples, header files, libraries,
1.10 +and tools you need to develop applications that use LZMA compression.
1.11 +
1.12 +LZMA is default and general compression method of 7z format
1.13 +in 7-Zip compression program (www.7-zip.org). LZMA provides high
1.14 +compression ratio and very fast decompression.
1.15 +
1.16 +LZMA is an improved version of famous LZ77 compression algorithm.
1.17 +It was improved in way of maximum increasing of compression ratio,
1.18 +keeping high decompression speed and low memory requirements for
1.19 +decompressing.
1.20 +
1.21 +
1.22 +
1.23 +LICENSE
1.24 +-------
1.25 +
1.26 +LZMA SDK is available under any of the following licenses:
1.27 +
1.28 +1) GNU Lesser General Public License (GNU LGPL)
1.29 +2) Common Public License (CPL)
1.30 +3) Simplified license for unmodified code (read SPECIAL EXCEPTION)
1.31 +4) Proprietary license
1.32 +
1.33 +It means that you can select one of these four options and follow rules of that license.
1.34 +
1.35 +
1.36 +1,2) GNU LGPL and CPL licenses are pretty similar and both these
1.37 +licenses are classified as
1.38 + - "Free software licenses" at http://www.gnu.org/
1.39 + - "OSI-approved" at http://www.opensource.org/
1.40 +
1.41 +
1.42 +3) SPECIAL EXCEPTION
1.43 +
1.44 +Igor Pavlov, as the author of this code, expressly permits you
1.45 +to statically or dynamically link your code (or bind by name)
1.46 +to the files from LZMA SDK without subjecting your linked
1.47 +code to the terms of the CPL or GNU LGPL.
1.48 +Any modifications or additions to files from LZMA SDK, however,
1.49 +are subject to the GNU LGPL or CPL terms.
1.50 +
1.51 +SPECIAL EXCEPTION allows you to use LZMA SDK in applications with closed code,
1.52 +while you keep LZMA SDK code unmodified.
1.53 +
1.54 +
1.55 +SPECIAL EXCEPTION #2: Igor Pavlov, as the author of this code, expressly permits
1.56 +you to use this code under the same terms and conditions contained in the License
1.57 +Agreement you have for any previous version of LZMA SDK developed by Igor Pavlov.
1.58 +
1.59 +SPECIAL EXCEPTION #2 allows owners of proprietary licenses to use latest version
1.60 +of LZMA SDK as update for previous versions.
1.61 +
1.62 +
1.63 +SPECIAL EXCEPTION #3: Igor Pavlov, as the author of this code, expressly permits
1.64 +you to use code of the following files:
1.65 +BranchTypes.h, LzmaTypes.h, LzmaTest.c, LzmaStateTest.c, LzmaAlone.cpp,
1.66 +LzmaAlone.cs, LzmaAlone.java
1.67 +as public domain code.
1.68 +
1.69 +
1.70 +4) Proprietary license
1.71 +
1.72 +LZMA SDK also can be available under a proprietary license which
1.73 +can include:
1.74 +
1.75 +1) Right to modify code without subjecting modified code to the
1.76 +terms of the CPL or GNU LGPL
1.77 +2) Technical support for code
1.78 +
1.79 +To request such proprietary license or any additional consultations,
1.80 +send email message from that page:
1.81 +http://www.7-zip.org/support.html
1.82 +
1.83 +
1.84 +You should have received a copy of the GNU Lesser General Public
1.85 +License along with this library; if not, write to the Free Software
1.86 +Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
1.87 +
1.88 +You should have received a copy of the Common Public License
1.89 +along with this library.
1.90 +
1.91 +
1.92 +LZMA SDK Contents
1.93 +-----------------
1.94 +
1.95 +LZMA SDK includes:
1.96 +
1.97 + - C++ source code of LZMA compressing and decompressing
1.98 + - ANSI-C compatible source code for LZMA decompressing
1.99 + - C# source code for LZMA compressing and decompressing
1.100 + - Java source code for LZMA compressing and decompressing
1.101 + - Compiled file->file LZMA compressing/decompressing program for Windows system
1.102 +
1.103 +ANSI-C LZMA decompression code was ported from original C++ sources to C.
1.104 +Also it was simplified and optimized for code size.
1.105 +But it is fully compatible with LZMA from 7-Zip.
1.106 +
1.107 +
1.108 +UNIX/Linux version
1.109 +------------------
1.110 +To compile C++ version of file->file LZMA, go to directory
1.111 +C/7zip/Compress/LZMA_Alone
1.112 +and type "make" or "make clean all" to recompile all.
1.113 +
1.114 +In some UNIX/Linux versions you must compile LZMA with static libraries.
1.115 +To compile with static libraries, change string in makefile
1.116 +LIB = -lm
1.117 +to string
1.118 +LIB = -lm -static
1.119 +
1.120 +
1.121 +Files
1.122 +---------------------
1.123 +C - C / CPP source code
1.124 +CS - C# source code
1.125 +Java - Java source code
1.126 +lzma.txt - LZMA SDK description (this file)
1.127 +7zFormat.txt - 7z Format description
1.128 +7zC.txt - 7z ANSI-C Decoder description (this file)
1.129 +methods.txt - Compression method IDs for .7z
1.130 +LGPL.txt - GNU Lesser General Public License
1.131 +CPL.html - Common Public License
1.132 +lzma.exe - Compiled file->file LZMA encoder/decoder for Windows
1.133 +history.txt - history of the LZMA SDK
1.134 +
1.135 +
1.136 +Source code structure
1.137 +---------------------
1.138 +
1.139 +C - C / CPP files
1.140 + Common - common files for C++ projects
1.141 + Windows - common files for Windows related code
1.142 + 7zip - files related to 7-Zip Project
1.143 + Common - common files for 7-Zip
1.144 + Compress - files related to compression/decompression
1.145 + LZ - files related to LZ (Lempel-Ziv) compression algorithm
1.146 + BinTree - Binary Tree Match Finder for LZ algorithm
1.147 + HashChain - Hash Chain Match Finder for LZ algorithm
1.148 + Patricia - Patricia Match Finder for LZ algorithm
1.149 + RangeCoder - Range Coder (special code of compression/decompression)
1.150 + LZMA - LZMA compression/decompression on C++
1.151 + LZMA_Alone - file->file LZMA compression/decompression
1.152 + LZMA_C - ANSI-C compatible LZMA decompressor
1.153 + LzmaDecode.h - interface for LZMA decoding on ANSI-C
1.154 + LzmaDecode.c - LZMA decoding on ANSI-C (new fastest version)
1.155 + LzmaDecodeSize.c - LZMA decoding on ANSI-C (old size-optimized version)
1.156 + LzmaTest.c - test application that decodes LZMA encoded file
1.157 + LzmaTypes.h - basic types for LZMA Decoder
1.158 + LzmaStateDecode.h - interface for LZMA decoding (State version)
1.159 + LzmaStateDecode.c - LZMA decoding on ANSI-C (State version)
1.160 + LzmaStateTest.c - test application (State version)
1.161 + Branch - Filters for x86, IA-64, ARM, ARM-Thumb, PowerPC and SPARC code
1.162 + Archive - files related to archiving
1.163 + 7z_C - 7z ANSI-C Decoder
1.164 +
1.165 +CS - C# files
1.166 + 7zip
1.167 + Common - some common files for 7-Zip
1.168 + Compress - files related to compression/decompression
1.169 + LZ - files related to LZ (Lempel-Ziv) compression algorithm
1.170 + LZMA - LZMA compression/decompression
1.171 + LzmaAlone - file->file LZMA compression/decompression
1.172 + RangeCoder - Range Coder (special code of compression/decompression)
1.173 +
1.174 +Java - Java files
1.175 + SevenZip
1.176 + Compression - files related to compression/decompression
1.177 + LZ - files related to LZ (Lempel-Ziv) compression algorithm
1.178 + LZMA - LZMA compression/decompression
1.179 + RangeCoder - Range Coder (special code of compression/decompression)
1.180 +
1.181 +C/C++ source code of LZMA SDK is part of 7-Zip project.
1.182 +
1.183 +You can find ANSI-C LZMA decompressing code at folder
1.184 + C/7zip/Compress/LZMA_C
1.185 +7-Zip doesn't use that ANSI-C LZMA code and that code was developed
1.186 +specially for this SDK. And files from LZMA_C do not need files from
1.187 +other directories of SDK for compiling.
1.188 +
1.189 +7-Zip source code can be downloaded from 7-Zip's SourceForge page:
1.190 +
1.191 + http://sourceforge.net/projects/sevenzip/
1.192 +
1.193 +
1.194 +LZMA features
1.195 +-------------
1.196 + - Variable dictionary size (up to 1 GB)
1.197 + - Estimated compressing speed: about 1 MB/s on 1 GHz CPU
1.198 + - Estimated decompressing speed:
1.199 + - 8-12 MB/s on 1 GHz Intel Pentium 3 or AMD Athlon
1.200 + - 500-1000 KB/s on 100 MHz ARM, MIPS, PowerPC or other simple RISC
1.201 + - Small memory requirements for decompressing (8-32 KB + DictionarySize)
1.202 + - Small code size for decompressing: 2-8 KB (depending from
1.203 + speed optimizations)
1.204 +
1.205 +LZMA decoder uses only integer operations and can be
1.206 +implemented in any modern 32-bit CPU (or on 16-bit CPU with some conditions).
1.207 +
1.208 +Some critical operations that affect to speed of LZMA decompression:
1.209 + 1) 32*16 bit integer multiply
1.210 + 2) Misspredicted branches (penalty mostly depends from pipeline length)
1.211 + 3) 32-bit shift and arithmetic operations
1.212 +
1.213 +Speed of LZMA decompressing mostly depends from CPU speed.
1.214 +Memory speed has no big meaning. But if your CPU has small data cache,
1.215 +overall weight of memory speed will slightly increase.
1.216 +
1.217 +
1.218 +How To Use
1.219 +----------
1.220 +
1.221 +Using LZMA encoder/decoder executable
1.222 +--------------------------------------
1.223 +
1.224 +Usage: LZMA <e|d> inputFile outputFile [<switches>...]
1.225 +
1.226 + e: encode file
1.227 +
1.228 + d: decode file
1.229 +
1.230 + b: Benchmark. There are two tests: compressing and decompressing
1.231 + with LZMA method. Benchmark shows rating in MIPS (million
1.232 + instructions per second). Rating value is calculated from
1.233 + measured speed and it is normalized with AMD Athlon 64 X2 CPU
1.234 + results. Also Benchmark checks possible hardware errors (RAM
1.235 + errors in most cases). Benchmark uses these settings:
1.236 + (-a1, -d21, -fb32, -mfbt4). You can change only -d. Also you
1.237 + can change number of iterations. Example for 30 iterations:
1.238 + LZMA b 30
1.239 + Default number of iterations is 10.
1.240 +
1.241 +<Switches>
1.242 +
1.243 +
1.244 + -a{N}: set compression mode 0 = fast, 1 = normal
1.245 + default: 1 (normal)
1.246 +
1.247 + d{N}: Sets Dictionary size - [0, 30], default: 23 (8MB)
1.248 + The maximum value for dictionary size is 1 GB = 2^30 bytes.
1.249 + Dictionary size is calculated as DictionarySize = 2^N bytes.
1.250 + For decompressing file compressed by LZMA method with dictionary
1.251 + size D = 2^N you need about D bytes of memory (RAM).
1.252 +
1.253 + -fb{N}: set number of fast bytes - [5, 273], default: 128
1.254 + Usually big number gives a little bit better compression ratio
1.255 + and slower compression process.
1.256 +
1.257 + -lc{N}: set number of literal context bits - [0, 8], default: 3
1.258 + Sometimes lc=4 gives gain for big files.
1.259 +
1.260 + -lp{N}: set number of literal pos bits - [0, 4], default: 0
1.261 + lp switch is intended for periodical data when period is
1.262 + equal 2^N. For example, for 32-bit (4 bytes)
1.263 + periodical data you can use lp=2. Often it's better to set lc0,
1.264 + if you change lp switch.
1.265 +
1.266 + -pb{N}: set number of pos bits - [0, 4], default: 2
1.267 + pb switch is intended for periodical data
1.268 + when period is equal 2^N.
1.269 +
1.270 + -mf{MF_ID}: set Match Finder. Default: bt4.
1.271 + Algorithms from hc* group doesn't provide good compression
1.272 + ratio, but they often works pretty fast in combination with
1.273 + fast mode (-a0).
1.274 +
1.275 + Memory requirements depend from dictionary size
1.276 + (parameter "d" in table below).
1.277 +
1.278 + MF_ID Memory Description
1.279 +
1.280 + bt2 d * 9.5 + 4MB Binary Tree with 2 bytes hashing.
1.281 + bt3 d * 11.5 + 4MB Binary Tree with 3 bytes hashing.
1.282 + bt4 d * 11.5 + 4MB Binary Tree with 4 bytes hashing.
1.283 + hc4 d * 7.5 + 4MB Hash Chain with 4 bytes hashing.
1.284 +
1.285 + -eos: write End Of Stream marker. By default LZMA doesn't write
1.286 + eos marker, since LZMA decoder knows uncompressed size
1.287 + stored in .lzma file header.
1.288 +
1.289 + -si: Read data from stdin (it will write End Of Stream marker).
1.290 + -so: Write data to stdout
1.291 +
1.292 +
1.293 +Examples:
1.294 +
1.295 +1) LZMA e file.bin file.lzma -d16 -lc0
1.296 +
1.297 +compresses file.bin to file.lzma with 64 KB dictionary (2^16=64K)
1.298 +and 0 literal context bits. -lc0 allows to reduce memory requirements
1.299 +for decompression.
1.300 +
1.301 +
1.302 +2) LZMA e file.bin file.lzma -lc0 -lp2
1.303 +
1.304 +compresses file.bin to file.lzma with settings suitable
1.305 +for 32-bit periodical data (for example, ARM or MIPS code).
1.306 +
1.307 +3) LZMA d file.lzma file.bin
1.308 +
1.309 +decompresses file.lzma to file.bin.
1.310 +
1.311 +
1.312 +Compression ratio hints
1.313 +-----------------------
1.314 +
1.315 +Recommendations
1.316 +---------------
1.317 +
1.318 +To increase compression ratio for LZMA compressing it's desirable
1.319 +to have aligned data (if it's possible) and also it's desirable to locate
1.320 +data in such order, where code is grouped in one place and data is
1.321 +grouped in other place (it's better than such mixing: code, data, code,
1.322 +data, ...).
1.323 +
1.324 +
1.325 +Using Filters
1.326 +-------------
1.327 +You can increase compression ratio for some data types, using
1.328 +special filters before compressing. For example, it's possible to
1.329 +increase compression ratio on 5-10% for code for those CPU ISAs:
1.330 +x86, IA-64, ARM, ARM-Thumb, PowerPC, SPARC.
1.331 +
1.332 +You can find C/C++ source code of such filters in folder "7zip/Compress/Branch"
1.333 +
1.334 +You can check compression ratio gain of these filters with such
1.335 +7-Zip commands (example for ARM code):
1.336 +No filter:
1.337 + 7z a a1.7z a.bin -m0=lzma
1.338 +
1.339 +With filter for little-endian ARM code:
1.340 + 7z a a2.7z a.bin -m0=bc_arm -m1=lzma
1.341 +
1.342 +With filter for big-endian ARM code (using additional Swap4 filter):
1.343 + 7z a a3.7z a.bin -m0=swap4 -m1=bc_arm -m2=lzma
1.344 +
1.345 +It works in such manner:
1.346 +Compressing = Filter_encoding + LZMA_encoding
1.347 +Decompressing = LZMA_decoding + Filter_decoding
1.348 +
1.349 +Compressing and decompressing speed of such filters is very high,
1.350 +so it will not increase decompressing time too much.
1.351 +Moreover, it reduces decompression time for LZMA_decoding,
1.352 +since compression ratio with filtering is higher.
1.353 +
1.354 +These filters convert CALL (calling procedure) instructions
1.355 +from relative offsets to absolute addresses, so such data becomes more
1.356 +compressible. Source code of these CALL filters is pretty simple
1.357 +(about 20 lines of C++), so you can convert it from C++ version yourself.
1.358 +
1.359 +For some ISAs (for example, for MIPS) it's impossible to get gain from such filter.
1.360 +
1.361 +
1.362 +LZMA compressed file format
1.363 +---------------------------
1.364 +Offset Size Description
1.365 + 0 1 Special LZMA properties for compressed data
1.366 + 1 4 Dictionary size (little endian)
1.367 + 5 8 Uncompressed size (little endian). -1 means unknown size
1.368 + 13 Compressed data
1.369 +
1.370 +
1.371 +ANSI-C LZMA Decoder
1.372 +~~~~~~~~~~~~~~~~~~~
1.373 +
1.374 +To compile ANSI-C LZMA Decoder you can use one of the following files sets:
1.375 +1) LzmaDecode.h + LzmaDecode.c + LzmaTest.c (fastest version)
1.376 +2) LzmaDecode.h + LzmaDecodeSize.c + LzmaTest.c (old size-optimized version)
1.377 +3) LzmaStateDecode.h + LzmaStateDecode.c + LzmaStateTest.c (zlib-like interface)
1.378 +
1.379 +
1.380 +Memory requirements for LZMA decoding
1.381 +-------------------------------------
1.382 +
1.383 +LZMA decoder doesn't allocate memory itself, so you must
1.384 +allocate memory and send it to LZMA.
1.385 +
1.386 +Stack usage of LZMA decoding function for local variables is not
1.387 +larger than 200 bytes.
1.388 +
1.389 +How To decompress data
1.390 +----------------------
1.391 +
1.392 +LZMA Decoder (ANSI-C version) now supports 5 interfaces:
1.393 +1) Single-call Decompressing
1.394 +2) Single-call Decompressing with input stream callback
1.395 +3) Multi-call Decompressing with output buffer
1.396 +4) Multi-call Decompressing with input callback and output buffer
1.397 +5) Multi-call State Decompressing (zlib-like interface)
1.398 +
1.399 +Variant-5 is similar to Variant-4, but Variant-5 doesn't use callback functions.
1.400 +
1.401 +Decompressing steps
1.402 +-------------------
1.403 +
1.404 +1) read LZMA properties (5 bytes):
1.405 + unsigned char properties[LZMA_PROPERTIES_SIZE];
1.406 +
1.407 +2) read uncompressed size (8 bytes, little-endian)
1.408 +
1.409 +3) Decode properties:
1.410 +
1.411 + CLzmaDecoderState state; /* it's 24-140 bytes structure, if int is 32-bit */
1.412 +
1.413 + if (LzmaDecodeProperties(&state.Properties, properties, LZMA_PROPERTIES_SIZE) != LZMA_RESULT_OK)
1.414 + return PrintError(rs, "Incorrect stream properties");
1.415 +
1.416 +4) Allocate memory block for internal Structures:
1.417 +
1.418 + state.Probs = (CProb *)malloc(LzmaGetNumProbs(&state.Properties) * sizeof(CProb));
1.419 + if (state.Probs == 0)
1.420 + return PrintError(rs, kCantAllocateMessage);
1.421 +
1.422 + LZMA decoder uses array of CProb variables as internal structure.
1.423 + By default, CProb is unsigned_short. But you can define _LZMA_PROB32 to make
1.424 + it unsigned_int. It can increase speed on some 32-bit CPUs, but memory
1.425 + usage will be doubled in that case.
1.426 +
1.427 +
1.428 +5) Main Decompressing
1.429 +
1.430 +You must use one of the following interfaces:
1.431 +
1.432 +5.1 Single-call Decompressing
1.433 +-----------------------------
1.434 +When to use: RAM->RAM decompressing
1.435 +Compile files: LzmaDecode.h, LzmaDecode.c
1.436 +Compile defines: no defines
1.437 +Memory Requirements:
1.438 + - Input buffer: compressed size
1.439 + - Output buffer: uncompressed size
1.440 + - LZMA Internal Structures (~16 KB for default settings)
1.441 +
1.442 +Interface:
1.443 + int res = LzmaDecode(&state,
1.444 + inStream, compressedSize, &inProcessed,
1.445 + outStream, outSize, &outProcessed);
1.446 +
1.447 +
1.448 +5.2 Single-call Decompressing with input stream callback
1.449 +--------------------------------------------------------
1.450 +When to use: File->RAM or Flash->RAM decompressing.
1.451 +Compile files: LzmaDecode.h, LzmaDecode.c
1.452 +Compile defines: _LZMA_IN_CB
1.453 +Memory Requirements:
1.454 + - Buffer for input stream: any size (for example, 16 KB)
1.455 + - Output buffer: uncompressed size
1.456 + - LZMA Internal Structures (~16 KB for default settings)
1.457 +
1.458 +Interface:
1.459 + typedef struct _CBuffer
1.460 + {
1.461 + ILzmaInCallback InCallback;
1.462 + FILE *File;
1.463 + unsigned char Buffer[kInBufferSize];
1.464 + } CBuffer;
1.465 +
1.466 + int LzmaReadCompressed(void *object, const unsigned char **buffer, SizeT *size)
1.467 + {
1.468 + CBuffer *bo = (CBuffer *)object;
1.469 + *buffer = bo->Buffer;
1.470 + *size = MyReadFile(bo->File, bo->Buffer, kInBufferSize);
1.471 + return LZMA_RESULT_OK;
1.472 + }
1.473 +
1.474 + CBuffer g_InBuffer;
1.475 +
1.476 + g_InBuffer.File = inFile;
1.477 + g_InBuffer.InCallback.Read = LzmaReadCompressed;
1.478 + int res = LzmaDecode(&state,
1.479 + &g_InBuffer.InCallback,
1.480 + outStream, outSize, &outProcessed);
1.481 +
1.482 +
1.483 +5.3 Multi-call decompressing with output buffer
1.484 +-----------------------------------------------
1.485 +When to use: RAM->File decompressing
1.486 +Compile files: LzmaDecode.h, LzmaDecode.c
1.487 +Compile defines: _LZMA_OUT_READ
1.488 +Memory Requirements:
1.489 + - Input buffer: compressed size
1.490 + - Buffer for output stream: any size (for example, 16 KB)
1.491 + - LZMA Internal Structures (~16 KB for default settings)
1.492 + - LZMA dictionary (dictionary size is encoded in stream properties)
1.493 +
1.494 +Interface:
1.495 +
1.496 + state.Dictionary = (unsigned char *)malloc(state.Properties.DictionarySize);
1.497 +
1.498 + LzmaDecoderInit(&state);
1.499 + do
1.500 + {
1.501 + LzmaDecode(&state,
1.502 + inBuffer, inAvail, &inProcessed,
1.503 + g_OutBuffer, outAvail, &outProcessed);
1.504 + inAvail -= inProcessed;
1.505 + inBuffer += inProcessed;
1.506 + }
1.507 + while you need more bytes
1.508 +
1.509 + see LzmaTest.c for more details.
1.510 +
1.511 +
1.512 +5.4 Multi-call decompressing with input callback and output buffer
1.513 +------------------------------------------------------------------
1.514 +When to use: File->File decompressing
1.515 +Compile files: LzmaDecode.h, LzmaDecode.c
1.516 +Compile defines: _LZMA_IN_CB, _LZMA_OUT_READ
1.517 +Memory Requirements:
1.518 + - Buffer for input stream: any size (for example, 16 KB)
1.519 + - Buffer for output stream: any size (for example, 16 KB)
1.520 + - LZMA Internal Structures (~16 KB for default settings)
1.521 + - LZMA dictionary (dictionary size is encoded in stream properties)
1.522 +
1.523 +Interface:
1.524 +
1.525 + state.Dictionary = (unsigned char *)malloc(state.Properties.DictionarySize);
1.526 +
1.527 + LzmaDecoderInit(&state);
1.528 + do
1.529 + {
1.530 + LzmaDecode(&state,
1.531 + &bo.InCallback,
1.532 + g_OutBuffer, outAvail, &outProcessed);
1.533 + }
1.534 + while you need more bytes
1.535 +
1.536 + see LzmaTest.c for more details:
1.537 +
1.538 +
1.539 +5.5 Multi-call State Decompressing (zlib-like interface)
1.540 +------------------------------------------------------------------
1.541 +When to use: file->file decompressing
1.542 +Compile files: LzmaStateDecode.h, LzmaStateDecode.c
1.543 +Compile defines:
1.544 +Memory Requirements:
1.545 + - Buffer for input stream: any size (for example, 16 KB)
1.546 + - Buffer for output stream: any size (for example, 16 KB)
1.547 + - LZMA Internal Structures (~16 KB for default settings)
1.548 + - LZMA dictionary (dictionary size is encoded in stream properties)
1.549 +
1.550 +Interface:
1.551 +
1.552 + state.Dictionary = (unsigned char *)malloc(state.Properties.DictionarySize);
1.553 +
1.554 +
1.555 + LzmaDecoderInit(&state);
1.556 + do
1.557 + {
1.558 + res = LzmaDecode(&state,
1.559 + inBuffer, inAvail, &inProcessed,
1.560 + g_OutBuffer, outAvail, &outProcessed,
1.561 + finishDecoding);
1.562 + inAvail -= inProcessed;
1.563 + inBuffer += inProcessed;
1.564 + }
1.565 + while you need more bytes
1.566 +
1.567 + see LzmaStateTest.c for more details:
1.568 +
1.569 +
1.570 +6) Free all allocated blocks
1.571 +
1.572 +
1.573 +Note
1.574 +----
1.575 +LzmaDecodeSize.c is size-optimized version of LzmaDecode.c.
1.576 +But compiled code of LzmaDecodeSize.c can be larger than
1.577 +compiled code of LzmaDecode.c. So it's better to use
1.578 +LzmaDecode.c in most cases.
1.579 +
1.580 +
1.581 +EXIT codes
1.582 +-----------
1.583 +
1.584 +LZMA decoder can return one of the following codes:
1.585 +
1.586 +#define LZMA_RESULT_OK 0
1.587 +#define LZMA_RESULT_DATA_ERROR 1
1.588 +
1.589 +If you use callback function for input data and you return some
1.590 +error code, LZMA Decoder also returns that code.
1.591 +
1.592 +
1.593 +
1.594 +LZMA Defines
1.595 +------------
1.596 +
1.597 +_LZMA_IN_CB - Use callback for input data
1.598 +
1.599 +_LZMA_OUT_READ - Use read function for output data
1.600 +
1.601 +_LZMA_LOC_OPT - Enable local speed optimizations inside code.
1.602 + _LZMA_LOC_OPT is only for LzmaDecodeSize.c (size-optimized version).
1.603 + _LZMA_LOC_OPT doesn't affect LzmaDecode.c (speed-optimized version)
1.604 + and LzmaStateDecode.c
1.605 +
1.606 +_LZMA_PROB32 - It can increase speed on some 32-bit CPUs,
1.607 + but memory usage will be doubled in that case
1.608 +
1.609 +_LZMA_UINT32_IS_ULONG - Define it if int is 16-bit on your compiler
1.610 + and long is 32-bit.
1.611 +
1.612 +_LZMA_SYSTEM_SIZE_T - Define it if you want to use system's size_t.
1.613 + You can use it to enable 64-bit sizes supporting
1.614 +
1.615 +
1.616 +
1.617 +C++ LZMA Encoder/Decoder
1.618 +~~~~~~~~~~~~~~~~~~~~~~~~
1.619 +C++ LZMA code use COM-like interfaces. So if you want to use it,
1.620 +you can study basics of COM/OLE.
1.621 +
1.622 +By default, LZMA Encoder contains all Match Finders.
1.623 +But for compressing it's enough to have just one of them.
1.624 +So for reducing size of compressing code you can define:
1.625 + #define COMPRESS_MF_BT
1.626 + #define COMPRESS_MF_BT4
1.627 +and it will use only bt4 match finder.
1.628 +
1.629 +
1.630 +---
1.631 +
1.632 +http://www.7-zip.org
1.633 +http://www.7-zip.org/support.html
