1 | /* |
2 | * Copyright (c) Meta Platforms, Inc. and affiliates. |
3 | * All rights reserved. |
4 | * |
5 | * This source code is licensed under both the BSD-style license (found in the |
6 | * LICENSE file in the root directory of this source tree) and the GPLv2 (found |
7 | * in the COPYING file in the root directory of this source tree). |
8 | * You may select, at your option, one of the above-listed licenses. |
9 | */ |
10 | #if defined (__cplusplus) |
11 | extern "C" { |
12 | #endif |
13 | |
14 | #ifndef ZSTD_H_235446 |
15 | #define ZSTD_H_235446 |
16 | |
17 | /* ====== Dependencies ======*/ |
18 | #include <limits.h> /* INT_MAX */ |
19 | #include <stddef.h> /* size_t */ |
20 | |
21 | |
22 | /* ===== ZSTDLIB_API : control library symbols visibility ===== */ |
23 | #ifndef ZSTDLIB_VISIBLE |
24 | /* Backwards compatibility with old macro name */ |
25 | # ifdef ZSTDLIB_VISIBILITY |
26 | # define ZSTDLIB_VISIBLE ZSTDLIB_VISIBILITY |
27 | # elif defined(__GNUC__) && (__GNUC__ >= 4) && !defined(__MINGW32__) |
28 | # define ZSTDLIB_VISIBLE __attribute__ ((visibility ("default"))) |
29 | # else |
30 | # define ZSTDLIB_VISIBLE |
31 | # endif |
32 | #endif |
33 | |
34 | #ifndef ZSTDLIB_HIDDEN |
35 | # if defined(__GNUC__) && (__GNUC__ >= 4) && !defined(__MINGW32__) |
36 | # define ZSTDLIB_HIDDEN __attribute__ ((visibility ("hidden"))) |
37 | # else |
38 | # define ZSTDLIB_HIDDEN |
39 | # endif |
40 | #endif |
41 | |
42 | #if defined(ZSTD_DLL_EXPORT) && (ZSTD_DLL_EXPORT==1) |
43 | # define ZSTDLIB_API __declspec(dllexport) ZSTDLIB_VISIBLE |
44 | #elif defined(ZSTD_DLL_IMPORT) && (ZSTD_DLL_IMPORT==1) |
45 | # define ZSTDLIB_API __declspec(dllimport) ZSTDLIB_VISIBLE /* It isn't required but allows to generate better code, saving a function pointer load from the IAT and an indirect jump.*/ |
46 | #else |
47 | # define ZSTDLIB_API ZSTDLIB_VISIBLE |
48 | #endif |
49 | |
50 | /* Deprecation warnings : |
51 | * Should these warnings be a problem, it is generally possible to disable them, |
52 | * typically with -Wno-deprecated-declarations for gcc or _CRT_SECURE_NO_WARNINGS in Visual. |
53 | * Otherwise, it's also possible to define ZSTD_DISABLE_DEPRECATE_WARNINGS. |
54 | */ |
55 | #ifdef ZSTD_DISABLE_DEPRECATE_WARNINGS |
56 | # define ZSTD_DEPRECATED(message) /* disable deprecation warnings */ |
57 | #else |
58 | # if defined (__cplusplus) && (__cplusplus >= 201402) /* C++14 or greater */ |
59 | # define ZSTD_DEPRECATED(message) [[deprecated(message)]] |
60 | # elif (defined(GNUC) && (GNUC > 4 || (GNUC == 4 && GNUC_MINOR >= 5))) || defined(__clang__) |
61 | # define ZSTD_DEPRECATED(message) __attribute__((deprecated(message))) |
62 | # elif defined(__GNUC__) && (__GNUC__ >= 3) |
63 | # define ZSTD_DEPRECATED(message) __attribute__((deprecated)) |
64 | # elif defined(_MSC_VER) |
65 | # define ZSTD_DEPRECATED(message) __declspec(deprecated(message)) |
66 | # else |
67 | # pragma message("WARNING: You need to implement ZSTD_DEPRECATED for this compiler") |
68 | # define ZSTD_DEPRECATED(message) |
69 | # endif |
70 | #endif /* ZSTD_DISABLE_DEPRECATE_WARNINGS */ |
71 | |
72 | |
73 | /******************************************************************************* |
74 | Introduction |
75 | |
76 | zstd, short for Zstandard, is a fast lossless compression algorithm, targeting |
77 | real-time compression scenarios at zlib-level and better compression ratios. |
78 | The zstd compression library provides in-memory compression and decompression |
79 | functions. |
80 | |
81 | The library supports regular compression levels from 1 up to ZSTD_maxCLevel(), |
82 | which is currently 22. Levels >= 20, labeled `--ultra`, should be used with |
83 | caution, as they require more memory. The library also offers negative |
84 | compression levels, which extend the range of speed vs. ratio preferences. |
85 | The lower the level, the faster the speed (at the cost of compression). |
86 | |
87 | Compression can be done in: |
88 | - a single step (described as Simple API) |
89 | - a single step, reusing a context (described as Explicit context) |
90 | - unbounded multiple steps (described as Streaming compression) |
91 | |
92 | The compression ratio achievable on small data can be highly improved using |
93 | a dictionary. Dictionary compression can be performed in: |
94 | - a single step (described as Simple dictionary API) |
95 | - a single step, reusing a dictionary (described as Bulk-processing |
96 | dictionary API) |
97 | |
98 | Advanced experimental functions can be accessed using |
99 | `#define ZSTD_STATIC_LINKING_ONLY` before including zstd.h. |
100 | |
101 | Advanced experimental APIs should never be used with a dynamically-linked |
102 | library. They are not "stable"; their definitions or signatures may change in |
103 | the future. Only static linking is allowed. |
104 | *******************************************************************************/ |
105 | |
106 | /*------ Version ------*/ |
107 | #define ZSTD_VERSION_MAJOR 1 |
108 | #define ZSTD_VERSION_MINOR 5 |
109 | #define ZSTD_VERSION_RELEASE 5 |
110 | #define ZSTD_VERSION_NUMBER (ZSTD_VERSION_MAJOR *100*100 + ZSTD_VERSION_MINOR *100 + ZSTD_VERSION_RELEASE) |
111 | |
112 | /*! ZSTD_versionNumber() : |
113 | * Return runtime library version, the value is (MAJOR*100*100 + MINOR*100 + RELEASE). */ |
114 | ZSTDLIB_API unsigned ZSTD_versionNumber(void); |
115 | |
116 | #define ZSTD_LIB_VERSION ZSTD_VERSION_MAJOR.ZSTD_VERSION_MINOR.ZSTD_VERSION_RELEASE |
117 | #define ZSTD_QUOTE(str) #str |
118 | #define ZSTD_EXPAND_AND_QUOTE(str) ZSTD_QUOTE(str) |
119 | #define ZSTD_VERSION_STRING ZSTD_EXPAND_AND_QUOTE(ZSTD_LIB_VERSION) |
120 | |
121 | /*! ZSTD_versionString() : |
122 | * Return runtime library version, like "1.4.5". Requires v1.3.0+. */ |
123 | ZSTDLIB_API const char* ZSTD_versionString(void); |
124 | |
125 | /* ************************************* |
126 | * Default constant |
127 | ***************************************/ |
128 | #ifndef ZSTD_CLEVEL_DEFAULT |
129 | # define ZSTD_CLEVEL_DEFAULT 3 |
130 | #endif |
131 | |
132 | /* ************************************* |
133 | * Constants |
134 | ***************************************/ |
135 | |
136 | /* All magic numbers are supposed read/written to/from files/memory using little-endian convention */ |
137 | #define ZSTD_MAGICNUMBER 0xFD2FB528 /* valid since v0.8.0 */ |
138 | #define ZSTD_MAGIC_DICTIONARY 0xEC30A437 /* valid since v0.7.0 */ |
139 | #define ZSTD_MAGIC_SKIPPABLE_START 0x184D2A50 /* all 16 values, from 0x184D2A50 to 0x184D2A5F, signal the beginning of a skippable frame */ |
140 | #define ZSTD_MAGIC_SKIPPABLE_MASK 0xFFFFFFF0 |
141 | |
142 | #define ZSTD_BLOCKSIZELOG_MAX 17 |
143 | #define ZSTD_BLOCKSIZE_MAX (1<<ZSTD_BLOCKSIZELOG_MAX) |
144 | |
145 | |
146 | /*************************************** |
147 | * Simple API |
148 | ***************************************/ |
149 | /*! ZSTD_compress() : |
150 | * Compresses `src` content as a single zstd compressed frame into already allocated `dst`. |
151 | * NOTE: Providing `dstCapacity >= ZSTD_compressBound(srcSize)` guarantees that zstd will have |
152 | * enough space to successfully compress the data. |
153 | * @return : compressed size written into `dst` (<= `dstCapacity), |
154 | * or an error code if it fails (which can be tested using ZSTD_isError()). */ |
155 | ZSTDLIB_API size_t ZSTD_compress( void* dst, size_t dstCapacity, |
156 | const void* src, size_t srcSize, |
157 | int compressionLevel); |
158 | |
159 | /*! ZSTD_decompress() : |
160 | * `compressedSize` : must be the _exact_ size of some number of compressed and/or skippable frames. |
161 | * `dstCapacity` is an upper bound of originalSize to regenerate. |
162 | * If user cannot imply a maximum upper bound, it's better to use streaming mode to decompress data. |
163 | * @return : the number of bytes decompressed into `dst` (<= `dstCapacity`), |
164 | * or an errorCode if it fails (which can be tested using ZSTD_isError()). */ |
165 | ZSTDLIB_API size_t ZSTD_decompress( void* dst, size_t dstCapacity, |
166 | const void* src, size_t compressedSize); |
167 | |
168 | /*! ZSTD_getFrameContentSize() : requires v1.3.0+ |
169 | * `src` should point to the start of a ZSTD encoded frame. |
170 | * `srcSize` must be at least as large as the frame header. |
171 | * hint : any size >= `ZSTD_frameHeaderSize_max` is large enough. |
172 | * @return : - decompressed size of `src` frame content, if known |
173 | * - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined |
174 | * - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) |
175 | * note 1 : a 0 return value means the frame is valid but "empty". |
176 | * note 2 : decompressed size is an optional field, it may not be present, typically in streaming mode. |
177 | * When `return==ZSTD_CONTENTSIZE_UNKNOWN`, data to decompress could be any size. |
178 | * In which case, it's necessary to use streaming mode to decompress data. |
179 | * Optionally, application can rely on some implicit limit, |
180 | * as ZSTD_decompress() only needs an upper bound of decompressed size. |
181 | * (For example, data could be necessarily cut into blocks <= 16 KB). |
182 | * note 3 : decompressed size is always present when compression is completed using single-pass functions, |
183 | * such as ZSTD_compress(), ZSTD_compressCCtx() ZSTD_compress_usingDict() or ZSTD_compress_usingCDict(). |
184 | * note 4 : decompressed size can be very large (64-bits value), |
185 | * potentially larger than what local system can handle as a single memory segment. |
186 | * In which case, it's necessary to use streaming mode to decompress data. |
187 | * note 5 : If source is untrusted, decompressed size could be wrong or intentionally modified. |
188 | * Always ensure return value fits within application's authorized limits. |
189 | * Each application can set its own limits. |
190 | * note 6 : This function replaces ZSTD_getDecompressedSize() */ |
191 | #define ZSTD_CONTENTSIZE_UNKNOWN (0ULL - 1) |
192 | #define ZSTD_CONTENTSIZE_ERROR (0ULL - 2) |
193 | ZSTDLIB_API unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize); |
194 | |
195 | /*! ZSTD_getDecompressedSize() : |
196 | * NOTE: This function is now obsolete, in favor of ZSTD_getFrameContentSize(). |
197 | * Both functions work the same way, but ZSTD_getDecompressedSize() blends |
198 | * "empty", "unknown" and "error" results to the same return value (0), |
199 | * while ZSTD_getFrameContentSize() gives them separate return values. |
200 | * @return : decompressed size of `src` frame content _if known and not empty_, 0 otherwise. */ |
201 | ZSTD_DEPRECATED("Replaced by ZSTD_getFrameContentSize" ) |
202 | ZSTDLIB_API |
203 | unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize); |
204 | |
205 | /*! ZSTD_findFrameCompressedSize() : Requires v1.4.0+ |
206 | * `src` should point to the start of a ZSTD frame or skippable frame. |
207 | * `srcSize` must be >= first frame size |
208 | * @return : the compressed size of the first frame starting at `src`, |
209 | * suitable to pass as `srcSize` to `ZSTD_decompress` or similar, |
210 | * or an error code if input is invalid */ |
211 | ZSTDLIB_API size_t ZSTD_findFrameCompressedSize(const void* src, size_t srcSize); |
212 | |
213 | |
214 | /*====== Helper functions ======*/ |
215 | /* ZSTD_compressBound() : |
216 | * maximum compressed size in worst case single-pass scenario. |
217 | * When invoking `ZSTD_compress()` or any other one-pass compression function, |
218 | * it's recommended to provide @dstCapacity >= ZSTD_compressBound(srcSize) |
219 | * as it eliminates one potential failure scenario, |
220 | * aka not enough room in dst buffer to write the compressed frame. |
221 | * Note : ZSTD_compressBound() itself can fail, if @srcSize > ZSTD_MAX_INPUT_SIZE . |
222 | * In which case, ZSTD_compressBound() will return an error code |
223 | * which can be tested using ZSTD_isError(). |
224 | * |
225 | * ZSTD_COMPRESSBOUND() : |
226 | * same as ZSTD_compressBound(), but as a macro. |
227 | * It can be used to produce constants, which can be useful for static allocation, |
228 | * for example to size a static array on stack. |
229 | * Will produce constant value 0 if srcSize too large. |
230 | */ |
231 | #define ZSTD_MAX_INPUT_SIZE ((sizeof(size_t)==8) ? 0xFF00FF00FF00FF00LLU : 0xFF00FF00U) |
232 | #define ZSTD_COMPRESSBOUND(srcSize) (((size_t)(srcSize) >= ZSTD_MAX_INPUT_SIZE) ? 0 : (srcSize) + ((srcSize)>>8) + (((srcSize) < (128<<10)) ? (((128<<10) - (srcSize)) >> 11) /* margin, from 64 to 0 */ : 0)) /* this formula ensures that bound(A) + bound(B) <= bound(A+B) as long as A and B >= 128 KB */ |
233 | ZSTDLIB_API size_t ZSTD_compressBound(size_t srcSize); /*!< maximum compressed size in worst case single-pass scenario */ |
234 | /* ZSTD_isError() : |
235 | * Most ZSTD_* functions returning a size_t value can be tested for error, |
236 | * using ZSTD_isError(). |
237 | * @return 1 if error, 0 otherwise |
238 | */ |
239 | ZSTDLIB_API unsigned ZSTD_isError(size_t code); /*!< tells if a `size_t` function result is an error code */ |
240 | ZSTDLIB_API const char* ZSTD_getErrorName(size_t code); /*!< provides readable string from an error code */ |
241 | ZSTDLIB_API int ZSTD_minCLevel(void); /*!< minimum negative compression level allowed, requires v1.4.0+ */ |
242 | ZSTDLIB_API int ZSTD_maxCLevel(void); /*!< maximum compression level available */ |
243 | ZSTDLIB_API int ZSTD_defaultCLevel(void); /*!< default compression level, specified by ZSTD_CLEVEL_DEFAULT, requires v1.5.0+ */ |
244 | |
245 | |
246 | /*************************************** |
247 | * Explicit context |
248 | ***************************************/ |
249 | /*= Compression context |
250 | * When compressing many times, |
251 | * it is recommended to allocate a context just once, |
252 | * and re-use it for each successive compression operation. |
253 | * This will make workload friendlier for system's memory. |
254 | * Note : re-using context is just a speed / resource optimization. |
255 | * It doesn't change the compression ratio, which remains identical. |
256 | * Note 2 : In multi-threaded environments, |
257 | * use one different context per thread for parallel execution. |
258 | */ |
259 | typedef struct ZSTD_CCtx_s ZSTD_CCtx; |
260 | ZSTDLIB_API ZSTD_CCtx* ZSTD_createCCtx(void); |
261 | ZSTDLIB_API size_t ZSTD_freeCCtx(ZSTD_CCtx* cctx); /* accept NULL pointer */ |
262 | |
263 | /*! ZSTD_compressCCtx() : |
264 | * Same as ZSTD_compress(), using an explicit ZSTD_CCtx. |
265 | * Important : in order to behave similarly to `ZSTD_compress()`, |
266 | * this function compresses at requested compression level, |
267 | * __ignoring any other parameter__ . |
268 | * If any advanced parameter was set using the advanced API, |
269 | * they will all be reset. Only `compressionLevel` remains. |
270 | */ |
271 | ZSTDLIB_API size_t ZSTD_compressCCtx(ZSTD_CCtx* cctx, |
272 | void* dst, size_t dstCapacity, |
273 | const void* src, size_t srcSize, |
274 | int compressionLevel); |
275 | |
276 | /*= Decompression context |
277 | * When decompressing many times, |
278 | * it is recommended to allocate a context only once, |
279 | * and re-use it for each successive compression operation. |
280 | * This will make workload friendlier for system's memory. |
281 | * Use one context per thread for parallel execution. */ |
282 | typedef struct ZSTD_DCtx_s ZSTD_DCtx; |
283 | ZSTDLIB_API ZSTD_DCtx* ZSTD_createDCtx(void); |
284 | ZSTDLIB_API size_t ZSTD_freeDCtx(ZSTD_DCtx* dctx); /* accept NULL pointer */ |
285 | |
286 | /*! ZSTD_decompressDCtx() : |
287 | * Same as ZSTD_decompress(), |
288 | * requires an allocated ZSTD_DCtx. |
289 | * Compatible with sticky parameters. |
290 | */ |
291 | ZSTDLIB_API size_t ZSTD_decompressDCtx(ZSTD_DCtx* dctx, |
292 | void* dst, size_t dstCapacity, |
293 | const void* src, size_t srcSize); |
294 | |
295 | |
296 | /********************************************* |
297 | * Advanced compression API (Requires v1.4.0+) |
298 | **********************************************/ |
299 | |
300 | /* API design : |
301 | * Parameters are pushed one by one into an existing context, |
302 | * using ZSTD_CCtx_set*() functions. |
303 | * Pushed parameters are sticky : they are valid for next compressed frame, and any subsequent frame. |
304 | * "sticky" parameters are applicable to `ZSTD_compress2()` and `ZSTD_compressStream*()` ! |
305 | * __They do not apply to "simple" one-shot variants such as ZSTD_compressCCtx()__ . |
306 | * |
307 | * It's possible to reset all parameters to "default" using ZSTD_CCtx_reset(). |
308 | * |
309 | * This API supersedes all other "advanced" API entry points in the experimental section. |
310 | * In the future, we expect to remove from experimental API entry points which are redundant with this API. |
311 | */ |
312 | |
313 | |
314 | /* Compression strategies, listed from fastest to strongest */ |
315 | typedef enum { ZSTD_fast=1, |
316 | ZSTD_dfast=2, |
317 | ZSTD_greedy=3, |
318 | ZSTD_lazy=4, |
319 | ZSTD_lazy2=5, |
320 | ZSTD_btlazy2=6, |
321 | ZSTD_btopt=7, |
322 | ZSTD_btultra=8, |
323 | ZSTD_btultra2=9 |
324 | /* note : new strategies _might_ be added in the future. |
325 | Only the order (from fast to strong) is guaranteed */ |
326 | } ZSTD_strategy; |
327 | |
328 | typedef enum { |
329 | |
330 | /* compression parameters |
331 | * Note: When compressing with a ZSTD_CDict these parameters are superseded |
332 | * by the parameters used to construct the ZSTD_CDict. |
333 | * See ZSTD_CCtx_refCDict() for more info (superseded-by-cdict). */ |
334 | ZSTD_c_compressionLevel=100, /* Set compression parameters according to pre-defined cLevel table. |
335 | * Note that exact compression parameters are dynamically determined, |
336 | * depending on both compression level and srcSize (when known). |
337 | * Default level is ZSTD_CLEVEL_DEFAULT==3. |
338 | * Special: value 0 means default, which is controlled by ZSTD_CLEVEL_DEFAULT. |
339 | * Note 1 : it's possible to pass a negative compression level. |
340 | * Note 2 : setting a level does not automatically set all other compression parameters |
341 | * to default. Setting this will however eventually dynamically impact the compression |
342 | * parameters which have not been manually set. The manually set |
343 | * ones will 'stick'. */ |
344 | /* Advanced compression parameters : |
345 | * It's possible to pin down compression parameters to some specific values. |
346 | * In which case, these values are no longer dynamically selected by the compressor */ |
347 | ZSTD_c_windowLog=101, /* Maximum allowed back-reference distance, expressed as power of 2. |
348 | * This will set a memory budget for streaming decompression, |
349 | * with larger values requiring more memory |
350 | * and typically compressing more. |
351 | * Must be clamped between ZSTD_WINDOWLOG_MIN and ZSTD_WINDOWLOG_MAX. |
352 | * Special: value 0 means "use default windowLog". |
353 | * Note: Using a windowLog greater than ZSTD_WINDOWLOG_LIMIT_DEFAULT |
354 | * requires explicitly allowing such size at streaming decompression stage. */ |
355 | ZSTD_c_hashLog=102, /* Size of the initial probe table, as a power of 2. |
356 | * Resulting memory usage is (1 << (hashLog+2)). |
357 | * Must be clamped between ZSTD_HASHLOG_MIN and ZSTD_HASHLOG_MAX. |
358 | * Larger tables improve compression ratio of strategies <= dFast, |
359 | * and improve speed of strategies > dFast. |
360 | * Special: value 0 means "use default hashLog". */ |
361 | ZSTD_c_chainLog=103, /* Size of the multi-probe search table, as a power of 2. |
362 | * Resulting memory usage is (1 << (chainLog+2)). |
363 | * Must be clamped between ZSTD_CHAINLOG_MIN and ZSTD_CHAINLOG_MAX. |
364 | * Larger tables result in better and slower compression. |
365 | * This parameter is useless for "fast" strategy. |
366 | * It's still useful when using "dfast" strategy, |
367 | * in which case it defines a secondary probe table. |
368 | * Special: value 0 means "use default chainLog". */ |
369 | ZSTD_c_searchLog=104, /* Number of search attempts, as a power of 2. |
370 | * More attempts result in better and slower compression. |
371 | * This parameter is useless for "fast" and "dFast" strategies. |
372 | * Special: value 0 means "use default searchLog". */ |
373 | ZSTD_c_minMatch=105, /* Minimum size of searched matches. |
374 | * Note that Zstandard can still find matches of smaller size, |
375 | * it just tweaks its search algorithm to look for this size and larger. |
376 | * Larger values increase compression and decompression speed, but decrease ratio. |
377 | * Must be clamped between ZSTD_MINMATCH_MIN and ZSTD_MINMATCH_MAX. |
378 | * Note that currently, for all strategies < btopt, effective minimum is 4. |
379 | * , for all strategies > fast, effective maximum is 6. |
380 | * Special: value 0 means "use default minMatchLength". */ |
381 | ZSTD_c_targetLength=106, /* Impact of this field depends on strategy. |
382 | * For strategies btopt, btultra & btultra2: |
383 | * Length of Match considered "good enough" to stop search. |
384 | * Larger values make compression stronger, and slower. |
385 | * For strategy fast: |
386 | * Distance between match sampling. |
387 | * Larger values make compression faster, and weaker. |
388 | * Special: value 0 means "use default targetLength". */ |
389 | ZSTD_c_strategy=107, /* See ZSTD_strategy enum definition. |
390 | * The higher the value of selected strategy, the more complex it is, |
391 | * resulting in stronger and slower compression. |
392 | * Special: value 0 means "use default strategy". */ |
393 | /* LDM mode parameters */ |
394 | ZSTD_c_enableLongDistanceMatching=160, /* Enable long distance matching. |
395 | * This parameter is designed to improve compression ratio |
396 | * for large inputs, by finding large matches at long distance. |
397 | * It increases memory usage and window size. |
398 | * Note: enabling this parameter increases default ZSTD_c_windowLog to 128 MB |
399 | * except when expressly set to a different value. |
400 | * Note: will be enabled by default if ZSTD_c_windowLog >= 128 MB and |
401 | * compression strategy >= ZSTD_btopt (== compression level 16+) */ |
402 | ZSTD_c_ldmHashLog=161, /* Size of the table for long distance matching, as a power of 2. |
403 | * Larger values increase memory usage and compression ratio, |
404 | * but decrease compression speed. |
405 | * Must be clamped between ZSTD_HASHLOG_MIN and ZSTD_HASHLOG_MAX |
406 | * default: windowlog - 7. |
407 | * Special: value 0 means "automatically determine hashlog". */ |
408 | ZSTD_c_ldmMinMatch=162, /* Minimum match size for long distance matcher. |
409 | * Larger/too small values usually decrease compression ratio. |
410 | * Must be clamped between ZSTD_LDM_MINMATCH_MIN and ZSTD_LDM_MINMATCH_MAX. |
411 | * Special: value 0 means "use default value" (default: 64). */ |
412 | ZSTD_c_ldmBucketSizeLog=163, /* Log size of each bucket in the LDM hash table for collision resolution. |
413 | * Larger values improve collision resolution but decrease compression speed. |
414 | * The maximum value is ZSTD_LDM_BUCKETSIZELOG_MAX. |
415 | * Special: value 0 means "use default value" (default: 3). */ |
416 | ZSTD_c_ldmHashRateLog=164, /* Frequency of inserting/looking up entries into the LDM hash table. |
417 | * Must be clamped between 0 and (ZSTD_WINDOWLOG_MAX - ZSTD_HASHLOG_MIN). |
418 | * Default is MAX(0, (windowLog - ldmHashLog)), optimizing hash table usage. |
419 | * Larger values improve compression speed. |
420 | * Deviating far from default value will likely result in a compression ratio decrease. |
421 | * Special: value 0 means "automatically determine hashRateLog". */ |
422 | |
423 | /* frame parameters */ |
424 | ZSTD_c_contentSizeFlag=200, /* Content size will be written into frame header _whenever known_ (default:1) |
425 | * Content size must be known at the beginning of compression. |
426 | * This is automatically the case when using ZSTD_compress2(), |
427 | * For streaming scenarios, content size must be provided with ZSTD_CCtx_setPledgedSrcSize() */ |
428 | ZSTD_c_checksumFlag=201, /* A 32-bits checksum of content is written at end of frame (default:0) */ |
429 | ZSTD_c_dictIDFlag=202, /* When applicable, dictionary's ID is written into frame header (default:1) */ |
430 | |
431 | /* multi-threading parameters */ |
432 | /* These parameters are only active if multi-threading is enabled (compiled with build macro ZSTD_MULTITHREAD). |
433 | * Otherwise, trying to set any other value than default (0) will be a no-op and return an error. |
434 | * In a situation where it's unknown if the linked library supports multi-threading or not, |
435 | * setting ZSTD_c_nbWorkers to any value >= 1 and consulting the return value provides a quick way to check this property. |
436 | */ |
437 | ZSTD_c_nbWorkers=400, /* Select how many threads will be spawned to compress in parallel. |
438 | * When nbWorkers >= 1, triggers asynchronous mode when invoking ZSTD_compressStream*() : |
439 | * ZSTD_compressStream*() consumes input and flush output if possible, but immediately gives back control to caller, |
440 | * while compression is performed in parallel, within worker thread(s). |
441 | * (note : a strong exception to this rule is when first invocation of ZSTD_compressStream2() sets ZSTD_e_end : |
442 | * in which case, ZSTD_compressStream2() delegates to ZSTD_compress2(), which is always a blocking call). |
443 | * More workers improve speed, but also increase memory usage. |
444 | * Default value is `0`, aka "single-threaded mode" : no worker is spawned, |
445 | * compression is performed inside Caller's thread, and all invocations are blocking */ |
446 | ZSTD_c_jobSize=401, /* Size of a compression job. This value is enforced only when nbWorkers >= 1. |
447 | * Each compression job is completed in parallel, so this value can indirectly impact the nb of active threads. |
448 | * 0 means default, which is dynamically determined based on compression parameters. |
449 | * Job size must be a minimum of overlap size, or ZSTDMT_JOBSIZE_MIN (= 512 KB), whichever is largest. |
450 | * The minimum size is automatically and transparently enforced. */ |
451 | ZSTD_c_overlapLog=402, /* Control the overlap size, as a fraction of window size. |
452 | * The overlap size is an amount of data reloaded from previous job at the beginning of a new job. |
453 | * It helps preserve compression ratio, while each job is compressed in parallel. |
454 | * This value is enforced only when nbWorkers >= 1. |
455 | * Larger values increase compression ratio, but decrease speed. |
456 | * Possible values range from 0 to 9 : |
457 | * - 0 means "default" : value will be determined by the library, depending on strategy |
458 | * - 1 means "no overlap" |
459 | * - 9 means "full overlap", using a full window size. |
460 | * Each intermediate rank increases/decreases load size by a factor 2 : |
461 | * 9: full window; 8: w/2; 7: w/4; 6: w/8; 5:w/16; 4: w/32; 3:w/64; 2:w/128; 1:no overlap; 0:default |
462 | * default value varies between 6 and 9, depending on strategy */ |
463 | |
464 | /* note : additional experimental parameters are also available |
465 | * within the experimental section of the API. |
466 | * At the time of this writing, they include : |
467 | * ZSTD_c_rsyncable |
468 | * ZSTD_c_format |
469 | * ZSTD_c_forceMaxWindow |
470 | * ZSTD_c_forceAttachDict |
471 | * ZSTD_c_literalCompressionMode |
472 | * ZSTD_c_targetCBlockSize |
473 | * ZSTD_c_srcSizeHint |
474 | * ZSTD_c_enableDedicatedDictSearch |
475 | * ZSTD_c_stableInBuffer |
476 | * ZSTD_c_stableOutBuffer |
477 | * ZSTD_c_blockDelimiters |
478 | * ZSTD_c_validateSequences |
479 | * ZSTD_c_useBlockSplitter |
480 | * ZSTD_c_useRowMatchFinder |
481 | * ZSTD_c_prefetchCDictTables |
482 | * ZSTD_c_enableSeqProducerFallback |
483 | * ZSTD_c_maxBlockSize |
484 | * Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them. |
485 | * note : never ever use experimentalParam? names directly; |
486 | * also, the enums values themselves are unstable and can still change. |
487 | */ |
488 | ZSTD_c_experimentalParam1=500, |
489 | ZSTD_c_experimentalParam2=10, |
490 | ZSTD_c_experimentalParam3=1000, |
491 | ZSTD_c_experimentalParam4=1001, |
492 | ZSTD_c_experimentalParam5=1002, |
493 | ZSTD_c_experimentalParam6=1003, |
494 | ZSTD_c_experimentalParam7=1004, |
495 | ZSTD_c_experimentalParam8=1005, |
496 | ZSTD_c_experimentalParam9=1006, |
497 | ZSTD_c_experimentalParam10=1007, |
498 | ZSTD_c_experimentalParam11=1008, |
499 | ZSTD_c_experimentalParam12=1009, |
500 | ZSTD_c_experimentalParam13=1010, |
501 | ZSTD_c_experimentalParam14=1011, |
502 | ZSTD_c_experimentalParam15=1012, |
503 | ZSTD_c_experimentalParam16=1013, |
504 | ZSTD_c_experimentalParam17=1014, |
505 | ZSTD_c_experimentalParam18=1015, |
506 | ZSTD_c_experimentalParam19=1016 |
507 | } ZSTD_cParameter; |
508 | |
509 | typedef struct { |
510 | size_t error; |
511 | int lowerBound; |
512 | int upperBound; |
513 | } ZSTD_bounds; |
514 | |
515 | /*! ZSTD_cParam_getBounds() : |
516 | * All parameters must belong to an interval with lower and upper bounds, |
517 | * otherwise they will either trigger an error or be automatically clamped. |
518 | * @return : a structure, ZSTD_bounds, which contains |
519 | * - an error status field, which must be tested using ZSTD_isError() |
520 | * - lower and upper bounds, both inclusive |
521 | */ |
522 | ZSTDLIB_API ZSTD_bounds ZSTD_cParam_getBounds(ZSTD_cParameter cParam); |
523 | |
524 | /*! ZSTD_CCtx_setParameter() : |
525 | * Set one compression parameter, selected by enum ZSTD_cParameter. |
526 | * All parameters have valid bounds. Bounds can be queried using ZSTD_cParam_getBounds(). |
527 | * Providing a value beyond bound will either clamp it, or trigger an error (depending on parameter). |
528 | * Setting a parameter is generally only possible during frame initialization (before starting compression). |
529 | * Exception : when using multi-threading mode (nbWorkers >= 1), |
530 | * the following parameters can be updated _during_ compression (within same frame): |
531 | * => compressionLevel, hashLog, chainLog, searchLog, minMatch, targetLength and strategy. |
532 | * new parameters will be active for next job only (after a flush()). |
533 | * @return : an error code (which can be tested using ZSTD_isError()). |
534 | */ |
535 | ZSTDLIB_API size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, int value); |
536 | |
537 | /*! ZSTD_CCtx_setPledgedSrcSize() : |
538 | * Total input data size to be compressed as a single frame. |
539 | * Value will be written in frame header, unless if explicitly forbidden using ZSTD_c_contentSizeFlag. |
540 | * This value will also be controlled at end of frame, and trigger an error if not respected. |
541 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
542 | * Note 1 : pledgedSrcSize==0 actually means zero, aka an empty frame. |
543 | * In order to mean "unknown content size", pass constant ZSTD_CONTENTSIZE_UNKNOWN. |
544 | * ZSTD_CONTENTSIZE_UNKNOWN is default value for any new frame. |
545 | * Note 2 : pledgedSrcSize is only valid once, for the next frame. |
546 | * It's discarded at the end of the frame, and replaced by ZSTD_CONTENTSIZE_UNKNOWN. |
547 | * Note 3 : Whenever all input data is provided and consumed in a single round, |
548 | * for example with ZSTD_compress2(), |
549 | * or invoking immediately ZSTD_compressStream2(,,,ZSTD_e_end), |
550 | * this value is automatically overridden by srcSize instead. |
551 | */ |
552 | ZSTDLIB_API size_t ZSTD_CCtx_setPledgedSrcSize(ZSTD_CCtx* cctx, unsigned long long pledgedSrcSize); |
553 | |
554 | typedef enum { |
555 | ZSTD_reset_session_only = 1, |
556 | ZSTD_reset_parameters = 2, |
557 | ZSTD_reset_session_and_parameters = 3 |
558 | } ZSTD_ResetDirective; |
559 | |
560 | /*! ZSTD_CCtx_reset() : |
561 | * There are 2 different things that can be reset, independently or jointly : |
562 | * - The session : will stop compressing current frame, and make CCtx ready to start a new one. |
563 | * Useful after an error, or to interrupt any ongoing compression. |
564 | * Any internal data not yet flushed is cancelled. |
565 | * Compression parameters and dictionary remain unchanged. |
566 | * They will be used to compress next frame. |
567 | * Resetting session never fails. |
568 | * - The parameters : changes all parameters back to "default". |
569 | * This also removes any reference to any dictionary or external sequence producer. |
570 | * Parameters can only be changed between 2 sessions (i.e. no compression is currently ongoing) |
571 | * otherwise the reset fails, and function returns an error value (which can be tested using ZSTD_isError()) |
572 | * - Both : similar to resetting the session, followed by resetting parameters. |
573 | */ |
574 | ZSTDLIB_API size_t ZSTD_CCtx_reset(ZSTD_CCtx* cctx, ZSTD_ResetDirective reset); |
575 | |
576 | /*! ZSTD_compress2() : |
577 | * Behave the same as ZSTD_compressCCtx(), but compression parameters are set using the advanced API. |
578 | * ZSTD_compress2() always starts a new frame. |
579 | * Should cctx hold data from a previously unfinished frame, everything about it is forgotten. |
580 | * - Compression parameters are pushed into CCtx before starting compression, using ZSTD_CCtx_set*() |
581 | * - The function is always blocking, returns when compression is completed. |
582 | * NOTE: Providing `dstCapacity >= ZSTD_compressBound(srcSize)` guarantees that zstd will have |
583 | * enough space to successfully compress the data, though it is possible it fails for other reasons. |
584 | * @return : compressed size written into `dst` (<= `dstCapacity), |
585 | * or an error code if it fails (which can be tested using ZSTD_isError()). |
586 | */ |
587 | ZSTDLIB_API size_t ZSTD_compress2( ZSTD_CCtx* cctx, |
588 | void* dst, size_t dstCapacity, |
589 | const void* src, size_t srcSize); |
590 | |
591 | |
592 | /*********************************************** |
593 | * Advanced decompression API (Requires v1.4.0+) |
594 | ************************************************/ |
595 | |
596 | /* The advanced API pushes parameters one by one into an existing DCtx context. |
597 | * Parameters are sticky, and remain valid for all following frames |
598 | * using the same DCtx context. |
599 | * It's possible to reset parameters to default values using ZSTD_DCtx_reset(). |
600 | * Note : This API is compatible with existing ZSTD_decompressDCtx() and ZSTD_decompressStream(). |
601 | * Therefore, no new decompression function is necessary. |
602 | */ |
603 | |
604 | typedef enum { |
605 | |
606 | ZSTD_d_windowLogMax=100, /* Select a size limit (in power of 2) beyond which |
607 | * the streaming API will refuse to allocate memory buffer |
608 | * in order to protect the host from unreasonable memory requirements. |
609 | * This parameter is only useful in streaming mode, since no internal buffer is allocated in single-pass mode. |
610 | * By default, a decompression context accepts window sizes <= (1 << ZSTD_WINDOWLOG_LIMIT_DEFAULT). |
611 | * Special: value 0 means "use default maximum windowLog". */ |
612 | |
613 | /* note : additional experimental parameters are also available |
614 | * within the experimental section of the API. |
615 | * At the time of this writing, they include : |
616 | * ZSTD_d_format |
617 | * ZSTD_d_stableOutBuffer |
618 | * ZSTD_d_forceIgnoreChecksum |
619 | * ZSTD_d_refMultipleDDicts |
620 | * ZSTD_d_disableHuffmanAssembly |
621 | * Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them. |
622 | * note : never ever use experimentalParam? names directly |
623 | */ |
624 | ZSTD_d_experimentalParam1=1000, |
625 | ZSTD_d_experimentalParam2=1001, |
626 | ZSTD_d_experimentalParam3=1002, |
627 | ZSTD_d_experimentalParam4=1003, |
628 | ZSTD_d_experimentalParam5=1004 |
629 | |
630 | } ZSTD_dParameter; |
631 | |
632 | /*! ZSTD_dParam_getBounds() : |
633 | * All parameters must belong to an interval with lower and upper bounds, |
634 | * otherwise they will either trigger an error or be automatically clamped. |
635 | * @return : a structure, ZSTD_bounds, which contains |
636 | * - an error status field, which must be tested using ZSTD_isError() |
637 | * - both lower and upper bounds, inclusive |
638 | */ |
639 | ZSTDLIB_API ZSTD_bounds ZSTD_dParam_getBounds(ZSTD_dParameter dParam); |
640 | |
641 | /*! ZSTD_DCtx_setParameter() : |
642 | * Set one compression parameter, selected by enum ZSTD_dParameter. |
643 | * All parameters have valid bounds. Bounds can be queried using ZSTD_dParam_getBounds(). |
644 | * Providing a value beyond bound will either clamp it, or trigger an error (depending on parameter). |
645 | * Setting a parameter is only possible during frame initialization (before starting decompression). |
646 | * @return : 0, or an error code (which can be tested using ZSTD_isError()). |
647 | */ |
648 | ZSTDLIB_API size_t ZSTD_DCtx_setParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int value); |
649 | |
650 | /*! ZSTD_DCtx_reset() : |
651 | * Return a DCtx to clean state. |
652 | * Session and parameters can be reset jointly or separately. |
653 | * Parameters can only be reset when no active frame is being decompressed. |
654 | * @return : 0, or an error code, which can be tested with ZSTD_isError() |
655 | */ |
656 | ZSTDLIB_API size_t ZSTD_DCtx_reset(ZSTD_DCtx* dctx, ZSTD_ResetDirective reset); |
657 | |
658 | |
659 | /**************************** |
660 | * Streaming |
661 | ****************************/ |
662 | |
663 | typedef struct ZSTD_inBuffer_s { |
664 | const void* src; /**< start of input buffer */ |
665 | size_t size; /**< size of input buffer */ |
666 | size_t pos; /**< position where reading stopped. Will be updated. Necessarily 0 <= pos <= size */ |
667 | } ZSTD_inBuffer; |
668 | |
669 | typedef struct ZSTD_outBuffer_s { |
670 | void* dst; /**< start of output buffer */ |
671 | size_t size; /**< size of output buffer */ |
672 | size_t pos; /**< position where writing stopped. Will be updated. Necessarily 0 <= pos <= size */ |
673 | } ZSTD_outBuffer; |
674 | |
675 | |
676 | |
677 | /*-*********************************************************************** |
678 | * Streaming compression - HowTo |
679 | * |
680 | * A ZSTD_CStream object is required to track streaming operation. |
681 | * Use ZSTD_createCStream() and ZSTD_freeCStream() to create/release resources. |
682 | * ZSTD_CStream objects can be reused multiple times on consecutive compression operations. |
683 | * It is recommended to re-use ZSTD_CStream since it will play nicer with system's memory, by re-using already allocated memory. |
684 | * |
685 | * For parallel execution, use one separate ZSTD_CStream per thread. |
686 | * |
687 | * note : since v1.3.0, ZSTD_CStream and ZSTD_CCtx are the same thing. |
688 | * |
689 | * Parameters are sticky : when starting a new compression on the same context, |
690 | * it will re-use the same sticky parameters as previous compression session. |
691 | * When in doubt, it's recommended to fully initialize the context before usage. |
692 | * Use ZSTD_CCtx_reset() to reset the context and ZSTD_CCtx_setParameter(), |
693 | * ZSTD_CCtx_setPledgedSrcSize(), or ZSTD_CCtx_loadDictionary() and friends to |
694 | * set more specific parameters, the pledged source size, or load a dictionary. |
695 | * |
696 | * Use ZSTD_compressStream2() with ZSTD_e_continue as many times as necessary to |
697 | * consume input stream. The function will automatically update both `pos` |
698 | * fields within `input` and `output`. |
699 | * Note that the function may not consume the entire input, for example, because |
700 | * the output buffer is already full, in which case `input.pos < input.size`. |
701 | * The caller must check if input has been entirely consumed. |
702 | * If not, the caller must make some room to receive more compressed data, |
703 | * and then present again remaining input data. |
704 | * note: ZSTD_e_continue is guaranteed to make some forward progress when called, |
705 | * but doesn't guarantee maximal forward progress. This is especially relevant |
706 | * when compressing with multiple threads. The call won't block if it can |
707 | * consume some input, but if it can't it will wait for some, but not all, |
708 | * output to be flushed. |
709 | * @return : provides a minimum amount of data remaining to be flushed from internal buffers |
710 | * or an error code, which can be tested using ZSTD_isError(). |
711 | * |
712 | * At any moment, it's possible to flush whatever data might remain stuck within internal buffer, |
713 | * using ZSTD_compressStream2() with ZSTD_e_flush. `output->pos` will be updated. |
714 | * Note that, if `output->size` is too small, a single invocation with ZSTD_e_flush might not be enough (return code > 0). |
715 | * In which case, make some room to receive more compressed data, and call again ZSTD_compressStream2() with ZSTD_e_flush. |
716 | * You must continue calling ZSTD_compressStream2() with ZSTD_e_flush until it returns 0, at which point you can change the |
717 | * operation. |
718 | * note: ZSTD_e_flush will flush as much output as possible, meaning when compressing with multiple threads, it will |
719 | * block until the flush is complete or the output buffer is full. |
720 | * @return : 0 if internal buffers are entirely flushed, |
721 | * >0 if some data still present within internal buffer (the value is minimal estimation of remaining size), |
722 | * or an error code, which can be tested using ZSTD_isError(). |
723 | * |
724 | * Calling ZSTD_compressStream2() with ZSTD_e_end instructs to finish a frame. |
725 | * It will perform a flush and write frame epilogue. |
726 | * The epilogue is required for decoders to consider a frame completed. |
727 | * flush operation is the same, and follows same rules as calling ZSTD_compressStream2() with ZSTD_e_flush. |
728 | * You must continue calling ZSTD_compressStream2() with ZSTD_e_end until it returns 0, at which point you are free to |
729 | * start a new frame. |
730 | * note: ZSTD_e_end will flush as much output as possible, meaning when compressing with multiple threads, it will |
731 | * block until the flush is complete or the output buffer is full. |
732 | * @return : 0 if frame fully completed and fully flushed, |
733 | * >0 if some data still present within internal buffer (the value is minimal estimation of remaining size), |
734 | * or an error code, which can be tested using ZSTD_isError(). |
735 | * |
736 | * *******************************************************************/ |
737 | |
738 | typedef ZSTD_CCtx ZSTD_CStream; /**< CCtx and CStream are now effectively same object (>= v1.3.0) */ |
739 | /* Continue to distinguish them for compatibility with older versions <= v1.2.0 */ |
740 | /*===== ZSTD_CStream management functions =====*/ |
741 | ZSTDLIB_API ZSTD_CStream* ZSTD_createCStream(void); |
742 | ZSTDLIB_API size_t ZSTD_freeCStream(ZSTD_CStream* zcs); /* accept NULL pointer */ |
743 | |
744 | /*===== Streaming compression functions =====*/ |
745 | typedef enum { |
746 | ZSTD_e_continue=0, /* collect more data, encoder decides when to output compressed result, for optimal compression ratio */ |
747 | ZSTD_e_flush=1, /* flush any data provided so far, |
748 | * it creates (at least) one new block, that can be decoded immediately on reception; |
749 | * frame will continue: any future data can still reference previously compressed data, improving compression. |
750 | * note : multithreaded compression will block to flush as much output as possible. */ |
751 | ZSTD_e_end=2 /* flush any remaining data _and_ close current frame. |
752 | * note that frame is only closed after compressed data is fully flushed (return value == 0). |
753 | * After that point, any additional data starts a new frame. |
754 | * note : each frame is independent (does not reference any content from previous frame). |
755 | : note : multithreaded compression will block to flush as much output as possible. */ |
756 | } ZSTD_EndDirective; |
757 | |
758 | /*! ZSTD_compressStream2() : Requires v1.4.0+ |
759 | * Behaves about the same as ZSTD_compressStream, with additional control on end directive. |
760 | * - Compression parameters are pushed into CCtx before starting compression, using ZSTD_CCtx_set*() |
761 | * - Compression parameters cannot be changed once compression is started (save a list of exceptions in multi-threading mode) |
762 | * - output->pos must be <= dstCapacity, input->pos must be <= srcSize |
763 | * - output->pos and input->pos will be updated. They are guaranteed to remain below their respective limit. |
764 | * - endOp must be a valid directive |
765 | * - When nbWorkers==0 (default), function is blocking : it completes its job before returning to caller. |
766 | * - When nbWorkers>=1, function is non-blocking : it copies a portion of input, distributes jobs to internal worker threads, flush to output whatever is available, |
767 | * and then immediately returns, just indicating that there is some data remaining to be flushed. |
768 | * The function nonetheless guarantees forward progress : it will return only after it reads or write at least 1+ byte. |
769 | * - Exception : if the first call requests a ZSTD_e_end directive and provides enough dstCapacity, the function delegates to ZSTD_compress2() which is always blocking. |
770 | * - @return provides a minimum amount of data remaining to be flushed from internal buffers |
771 | * or an error code, which can be tested using ZSTD_isError(). |
772 | * if @return != 0, flush is not fully completed, there is still some data left within internal buffers. |
773 | * This is useful for ZSTD_e_flush, since in this case more flushes are necessary to empty all buffers. |
774 | * For ZSTD_e_end, @return == 0 when internal buffers are fully flushed and frame is completed. |
775 | * - after a ZSTD_e_end directive, if internal buffer is not fully flushed (@return != 0), |
776 | * only ZSTD_e_end or ZSTD_e_flush operations are allowed. |
777 | * Before starting a new compression job, or changing compression parameters, |
778 | * it is required to fully flush internal buffers. |
779 | */ |
780 | ZSTDLIB_API size_t ZSTD_compressStream2( ZSTD_CCtx* cctx, |
781 | ZSTD_outBuffer* output, |
782 | ZSTD_inBuffer* input, |
783 | ZSTD_EndDirective endOp); |
784 | |
785 | |
786 | /* These buffer sizes are softly recommended. |
787 | * They are not required : ZSTD_compressStream*() happily accepts any buffer size, for both input and output. |
788 | * Respecting the recommended size just makes it a bit easier for ZSTD_compressStream*(), |
789 | * reducing the amount of memory shuffling and buffering, resulting in minor performance savings. |
790 | * |
791 | * However, note that these recommendations are from the perspective of a C caller program. |
792 | * If the streaming interface is invoked from some other language, |
793 | * especially managed ones such as Java or Go, through a foreign function interface such as jni or cgo, |
794 | * a major performance rule is to reduce crossing such interface to an absolute minimum. |
795 | * It's not rare that performance ends being spent more into the interface, rather than compression itself. |
796 | * In which cases, prefer using large buffers, as large as practical, |
797 | * for both input and output, to reduce the nb of roundtrips. |
798 | */ |
799 | ZSTDLIB_API size_t ZSTD_CStreamInSize(void); /**< recommended size for input buffer */ |
800 | ZSTDLIB_API size_t ZSTD_CStreamOutSize(void); /**< recommended size for output buffer. Guarantee to successfully flush at least one complete compressed block. */ |
801 | |
802 | |
803 | /* ***************************************************************************** |
804 | * This following is a legacy streaming API, available since v1.0+ . |
805 | * It can be replaced by ZSTD_CCtx_reset() and ZSTD_compressStream2(). |
806 | * It is redundant, but remains fully supported. |
807 | ******************************************************************************/ |
808 | |
809 | /*! |
810 | * Equivalent to: |
811 | * |
812 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
813 | * ZSTD_CCtx_refCDict(zcs, NULL); // clear the dictionary (if any) |
814 | * ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel); |
815 | * |
816 | * Note that ZSTD_initCStream() clears any previously set dictionary. Use the new API |
817 | * to compress with a dictionary. |
818 | */ |
819 | ZSTDLIB_API size_t ZSTD_initCStream(ZSTD_CStream* zcs, int compressionLevel); |
820 | /*! |
821 | * Alternative for ZSTD_compressStream2(zcs, output, input, ZSTD_e_continue). |
822 | * NOTE: The return value is different. ZSTD_compressStream() returns a hint for |
823 | * the next read size (if non-zero and not an error). ZSTD_compressStream2() |
824 | * returns the minimum nb of bytes left to flush (if non-zero and not an error). |
825 | */ |
826 | ZSTDLIB_API size_t ZSTD_compressStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output, ZSTD_inBuffer* input); |
827 | /*! Equivalent to ZSTD_compressStream2(zcs, output, &emptyInput, ZSTD_e_flush). */ |
828 | ZSTDLIB_API size_t ZSTD_flushStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output); |
829 | /*! Equivalent to ZSTD_compressStream2(zcs, output, &emptyInput, ZSTD_e_end). */ |
830 | ZSTDLIB_API size_t ZSTD_endStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output); |
831 | |
832 | |
833 | /*-*************************************************************************** |
834 | * Streaming decompression - HowTo |
835 | * |
836 | * A ZSTD_DStream object is required to track streaming operations. |
837 | * Use ZSTD_createDStream() and ZSTD_freeDStream() to create/release resources. |
838 | * ZSTD_DStream objects can be re-used multiple times. |
839 | * |
840 | * Use ZSTD_initDStream() to start a new decompression operation. |
841 | * @return : recommended first input size |
842 | * Alternatively, use advanced API to set specific properties. |
843 | * |
844 | * Use ZSTD_decompressStream() repetitively to consume your input. |
845 | * The function will update both `pos` fields. |
846 | * If `input.pos < input.size`, some input has not been consumed. |
847 | * It's up to the caller to present again remaining data. |
848 | * The function tries to flush all data decoded immediately, respecting output buffer size. |
849 | * If `output.pos < output.size`, decoder has flushed everything it could. |
850 | * But if `output.pos == output.size`, there might be some data left within internal buffers., |
851 | * In which case, call ZSTD_decompressStream() again to flush whatever remains in the buffer. |
852 | * Note : with no additional input provided, amount of data flushed is necessarily <= ZSTD_BLOCKSIZE_MAX. |
853 | * @return : 0 when a frame is completely decoded and fully flushed, |
854 | * or an error code, which can be tested using ZSTD_isError(), |
855 | * or any other value > 0, which means there is still some decoding or flushing to do to complete current frame : |
856 | * the return value is a suggested next input size (just a hint for better latency) |
857 | * that will never request more than the remaining frame size. |
858 | * *******************************************************************************/ |
859 | |
860 | typedef ZSTD_DCtx ZSTD_DStream; /**< DCtx and DStream are now effectively same object (>= v1.3.0) */ |
861 | /* For compatibility with versions <= v1.2.0, prefer differentiating them. */ |
862 | /*===== ZSTD_DStream management functions =====*/ |
863 | ZSTDLIB_API ZSTD_DStream* ZSTD_createDStream(void); |
864 | ZSTDLIB_API size_t ZSTD_freeDStream(ZSTD_DStream* zds); /* accept NULL pointer */ |
865 | |
866 | /*===== Streaming decompression functions =====*/ |
867 | |
868 | /*! ZSTD_initDStream() : |
869 | * Initialize/reset DStream state for new decompression operation. |
870 | * Call before new decompression operation using same DStream. |
871 | * |
872 | * Note : This function is redundant with the advanced API and equivalent to: |
873 | * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); |
874 | * ZSTD_DCtx_refDDict(zds, NULL); |
875 | */ |
876 | ZSTDLIB_API size_t ZSTD_initDStream(ZSTD_DStream* zds); |
877 | |
878 | /*! ZSTD_decompressStream() : |
879 | * Streaming decompression function. |
880 | * Call repetitively to consume full input updating it as necessary. |
881 | * Function will update both input and output `pos` fields exposing current state via these fields: |
882 | * - `input.pos < input.size`, some input remaining and caller should provide remaining input |
883 | * on the next call. |
884 | * - `output.pos < output.size`, decoder finished and flushed all remaining buffers. |
885 | * - `output.pos == output.size`, potentially uncflushed data present in the internal buffers, |
886 | * call ZSTD_decompressStream() again to flush remaining data to output. |
887 | * Note : with no additional input, amount of data flushed <= ZSTD_BLOCKSIZE_MAX. |
888 | * |
889 | * @return : 0 when a frame is completely decoded and fully flushed, |
890 | * or an error code, which can be tested using ZSTD_isError(), |
891 | * or any other value > 0, which means there is some decoding or flushing to do to complete current frame. |
892 | */ |
893 | ZSTDLIB_API size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inBuffer* input); |
894 | |
895 | ZSTDLIB_API size_t ZSTD_DStreamInSize(void); /*!< recommended size for input buffer */ |
896 | ZSTDLIB_API size_t ZSTD_DStreamOutSize(void); /*!< recommended size for output buffer. Guarantee to successfully flush at least one complete block in all circumstances. */ |
897 | |
898 | |
899 | /************************** |
900 | * Simple dictionary API |
901 | ***************************/ |
902 | /*! ZSTD_compress_usingDict() : |
903 | * Compression at an explicit compression level using a Dictionary. |
904 | * A dictionary can be any arbitrary data segment (also called a prefix), |
905 | * or a buffer with specified information (see zdict.h). |
906 | * Note : This function loads the dictionary, resulting in significant startup delay. |
907 | * It's intended for a dictionary used only once. |
908 | * Note 2 : When `dict == NULL || dictSize < 8` no dictionary is used. */ |
909 | ZSTDLIB_API size_t ZSTD_compress_usingDict(ZSTD_CCtx* ctx, |
910 | void* dst, size_t dstCapacity, |
911 | const void* src, size_t srcSize, |
912 | const void* dict,size_t dictSize, |
913 | int compressionLevel); |
914 | |
915 | /*! ZSTD_decompress_usingDict() : |
916 | * Decompression using a known Dictionary. |
917 | * Dictionary must be identical to the one used during compression. |
918 | * Note : This function loads the dictionary, resulting in significant startup delay. |
919 | * It's intended for a dictionary used only once. |
920 | * Note : When `dict == NULL || dictSize < 8` no dictionary is used. */ |
921 | ZSTDLIB_API size_t ZSTD_decompress_usingDict(ZSTD_DCtx* dctx, |
922 | void* dst, size_t dstCapacity, |
923 | const void* src, size_t srcSize, |
924 | const void* dict,size_t dictSize); |
925 | |
926 | |
927 | /*********************************** |
928 | * Bulk processing dictionary API |
929 | **********************************/ |
930 | typedef struct ZSTD_CDict_s ZSTD_CDict; |
931 | |
932 | /*! ZSTD_createCDict() : |
933 | * When compressing multiple messages or blocks using the same dictionary, |
934 | * it's recommended to digest the dictionary only once, since it's a costly operation. |
935 | * ZSTD_createCDict() will create a state from digesting a dictionary. |
936 | * The resulting state can be used for future compression operations with very limited startup cost. |
937 | * ZSTD_CDict can be created once and shared by multiple threads concurrently, since its usage is read-only. |
938 | * @dictBuffer can be released after ZSTD_CDict creation, because its content is copied within CDict. |
939 | * Note 1 : Consider experimental function `ZSTD_createCDict_byReference()` if you prefer to not duplicate @dictBuffer content. |
940 | * Note 2 : A ZSTD_CDict can be created from an empty @dictBuffer, |
941 | * in which case the only thing that it transports is the @compressionLevel. |
942 | * This can be useful in a pipeline featuring ZSTD_compress_usingCDict() exclusively, |
943 | * expecting a ZSTD_CDict parameter with any data, including those without a known dictionary. */ |
944 | ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict(const void* dictBuffer, size_t dictSize, |
945 | int compressionLevel); |
946 | |
947 | /*! ZSTD_freeCDict() : |
948 | * Function frees memory allocated by ZSTD_createCDict(). |
949 | * If a NULL pointer is passed, no operation is performed. */ |
950 | ZSTDLIB_API size_t ZSTD_freeCDict(ZSTD_CDict* CDict); |
951 | |
952 | /*! ZSTD_compress_usingCDict() : |
953 | * Compression using a digested Dictionary. |
954 | * Recommended when same dictionary is used multiple times. |
955 | * Note : compression level is _decided at dictionary creation time_, |
956 | * and frame parameters are hardcoded (dictID=yes, contentSize=yes, checksum=no) */ |
957 | ZSTDLIB_API size_t ZSTD_compress_usingCDict(ZSTD_CCtx* cctx, |
958 | void* dst, size_t dstCapacity, |
959 | const void* src, size_t srcSize, |
960 | const ZSTD_CDict* cdict); |
961 | |
962 | |
963 | typedef struct ZSTD_DDict_s ZSTD_DDict; |
964 | |
965 | /*! ZSTD_createDDict() : |
966 | * Create a digested dictionary, ready to start decompression operation without startup delay. |
967 | * dictBuffer can be released after DDict creation, as its content is copied inside DDict. */ |
968 | ZSTDLIB_API ZSTD_DDict* ZSTD_createDDict(const void* dictBuffer, size_t dictSize); |
969 | |
970 | /*! ZSTD_freeDDict() : |
971 | * Function frees memory allocated with ZSTD_createDDict() |
972 | * If a NULL pointer is passed, no operation is performed. */ |
973 | ZSTDLIB_API size_t ZSTD_freeDDict(ZSTD_DDict* ddict); |
974 | |
975 | /*! ZSTD_decompress_usingDDict() : |
976 | * Decompression using a digested Dictionary. |
977 | * Recommended when same dictionary is used multiple times. */ |
978 | ZSTDLIB_API size_t ZSTD_decompress_usingDDict(ZSTD_DCtx* dctx, |
979 | void* dst, size_t dstCapacity, |
980 | const void* src, size_t srcSize, |
981 | const ZSTD_DDict* ddict); |
982 | |
983 | |
984 | /******************************** |
985 | * Dictionary helper functions |
986 | *******************************/ |
987 | |
988 | /*! ZSTD_getDictID_fromDict() : Requires v1.4.0+ |
989 | * Provides the dictID stored within dictionary. |
990 | * if @return == 0, the dictionary is not conformant with Zstandard specification. |
991 | * It can still be loaded, but as a content-only dictionary. */ |
992 | ZSTDLIB_API unsigned ZSTD_getDictID_fromDict(const void* dict, size_t dictSize); |
993 | |
994 | /*! ZSTD_getDictID_fromCDict() : Requires v1.5.0+ |
995 | * Provides the dictID of the dictionary loaded into `cdict`. |
996 | * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. |
997 | * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ |
998 | ZSTDLIB_API unsigned ZSTD_getDictID_fromCDict(const ZSTD_CDict* cdict); |
999 | |
1000 | /*! ZSTD_getDictID_fromDDict() : Requires v1.4.0+ |
1001 | * Provides the dictID of the dictionary loaded into `ddict`. |
1002 | * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. |
1003 | * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ |
1004 | ZSTDLIB_API unsigned ZSTD_getDictID_fromDDict(const ZSTD_DDict* ddict); |
1005 | |
1006 | /*! ZSTD_getDictID_fromFrame() : Requires v1.4.0+ |
1007 | * Provides the dictID required to decompressed the frame stored within `src`. |
1008 | * If @return == 0, the dictID could not be decoded. |
1009 | * This could for one of the following reasons : |
1010 | * - The frame does not require a dictionary to be decoded (most common case). |
1011 | * - The frame was built with dictID intentionally removed. Whatever dictionary is necessary is a hidden piece of information. |
1012 | * Note : this use case also happens when using a non-conformant dictionary. |
1013 | * - `srcSize` is too small, and as a result, the frame header could not be decoded (only possible if `srcSize < ZSTD_FRAMEHEADERSIZE_MAX`). |
1014 | * - This is not a Zstandard frame. |
1015 | * When identifying the exact failure cause, it's possible to use ZSTD_getFrameHeader(), which will provide a more precise error code. */ |
1016 | ZSTDLIB_API unsigned ZSTD_getDictID_fromFrame(const void* src, size_t srcSize); |
1017 | |
1018 | |
1019 | /******************************************************************************* |
1020 | * Advanced dictionary and prefix API (Requires v1.4.0+) |
1021 | * |
1022 | * This API allows dictionaries to be used with ZSTD_compress2(), |
1023 | * ZSTD_compressStream2(), and ZSTD_decompressDCtx(). |
1024 | * Dictionaries are sticky, they remain valid when same context is re-used, |
1025 | * they only reset when the context is reset |
1026 | * with ZSTD_reset_parameters or ZSTD_reset_session_and_parameters. |
1027 | * In contrast, Prefixes are single-use. |
1028 | ******************************************************************************/ |
1029 | |
1030 | |
1031 | /*! ZSTD_CCtx_loadDictionary() : Requires v1.4.0+ |
1032 | * Create an internal CDict from `dict` buffer. |
1033 | * Decompression will have to use same dictionary. |
1034 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
1035 | * Special: Loading a NULL (or 0-size) dictionary invalidates previous dictionary, |
1036 | * meaning "return to no-dictionary mode". |
1037 | * Note 1 : Dictionary is sticky, it will be used for all future compressed frames, |
1038 | * until parameters are reset, a new dictionary is loaded, or the dictionary |
1039 | * is explicitly invalidated by loading a NULL dictionary. |
1040 | * Note 2 : Loading a dictionary involves building tables. |
1041 | * It's also a CPU consuming operation, with non-negligible impact on latency. |
1042 | * Tables are dependent on compression parameters, and for this reason, |
1043 | * compression parameters can no longer be changed after loading a dictionary. |
1044 | * Note 3 :`dict` content will be copied internally. |
1045 | * Use experimental ZSTD_CCtx_loadDictionary_byReference() to reference content instead. |
1046 | * In such a case, dictionary buffer must outlive its users. |
1047 | * Note 4 : Use ZSTD_CCtx_loadDictionary_advanced() |
1048 | * to precisely select how dictionary content must be interpreted. |
1049 | * Note 5 : This method does not benefit from LDM (long distance mode). |
1050 | * If you want to employ LDM on some large dictionary content, |
1051 | * prefer employing ZSTD_CCtx_refPrefix() described below. |
1052 | */ |
1053 | ZSTDLIB_API size_t ZSTD_CCtx_loadDictionary(ZSTD_CCtx* cctx, const void* dict, size_t dictSize); |
1054 | |
1055 | /*! ZSTD_CCtx_refCDict() : Requires v1.4.0+ |
1056 | * Reference a prepared dictionary, to be used for all future compressed frames. |
1057 | * Note that compression parameters are enforced from within CDict, |
1058 | * and supersede any compression parameter previously set within CCtx. |
1059 | * The parameters ignored are labelled as "superseded-by-cdict" in the ZSTD_cParameter enum docs. |
1060 | * The ignored parameters will be used again if the CCtx is returned to no-dictionary mode. |
1061 | * The dictionary will remain valid for future compressed frames using same CCtx. |
1062 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
1063 | * Special : Referencing a NULL CDict means "return to no-dictionary mode". |
1064 | * Note 1 : Currently, only one dictionary can be managed. |
1065 | * Referencing a new dictionary effectively "discards" any previous one. |
1066 | * Note 2 : CDict is just referenced, its lifetime must outlive its usage within CCtx. */ |
1067 | ZSTDLIB_API size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict); |
1068 | |
1069 | /*! ZSTD_CCtx_refPrefix() : Requires v1.4.0+ |
1070 | * Reference a prefix (single-usage dictionary) for next compressed frame. |
1071 | * A prefix is **only used once**. Tables are discarded at end of frame (ZSTD_e_end). |
1072 | * Decompression will need same prefix to properly regenerate data. |
1073 | * Compressing with a prefix is similar in outcome as performing a diff and compressing it, |
1074 | * but performs much faster, especially during decompression (compression speed is tunable with compression level). |
1075 | * This method is compatible with LDM (long distance mode). |
1076 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
1077 | * Special: Adding any prefix (including NULL) invalidates any previous prefix or dictionary |
1078 | * Note 1 : Prefix buffer is referenced. It **must** outlive compression. |
1079 | * Its content must remain unmodified during compression. |
1080 | * Note 2 : If the intention is to diff some large src data blob with some prior version of itself, |
1081 | * ensure that the window size is large enough to contain the entire source. |
1082 | * See ZSTD_c_windowLog. |
1083 | * Note 3 : Referencing a prefix involves building tables, which are dependent on compression parameters. |
1084 | * It's a CPU consuming operation, with non-negligible impact on latency. |
1085 | * If there is a need to use the same prefix multiple times, consider loadDictionary instead. |
1086 | * Note 4 : By default, the prefix is interpreted as raw content (ZSTD_dct_rawContent). |
1087 | * Use experimental ZSTD_CCtx_refPrefix_advanced() to alter dictionary interpretation. */ |
1088 | ZSTDLIB_API size_t ZSTD_CCtx_refPrefix(ZSTD_CCtx* cctx, |
1089 | const void* prefix, size_t prefixSize); |
1090 | |
1091 | /*! ZSTD_DCtx_loadDictionary() : Requires v1.4.0+ |
1092 | * Create an internal DDict from dict buffer, to be used to decompress all future frames. |
1093 | * The dictionary remains valid for all future frames, until explicitly invalidated, or |
1094 | * a new dictionary is loaded. |
1095 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
1096 | * Special : Adding a NULL (or 0-size) dictionary invalidates any previous dictionary, |
1097 | * meaning "return to no-dictionary mode". |
1098 | * Note 1 : Loading a dictionary involves building tables, |
1099 | * which has a non-negligible impact on CPU usage and latency. |
1100 | * It's recommended to "load once, use many times", to amortize the cost |
1101 | * Note 2 :`dict` content will be copied internally, so `dict` can be released after loading. |
1102 | * Use ZSTD_DCtx_loadDictionary_byReference() to reference dictionary content instead. |
1103 | * Note 3 : Use ZSTD_DCtx_loadDictionary_advanced() to take control of |
1104 | * how dictionary content is loaded and interpreted. |
1105 | */ |
1106 | ZSTDLIB_API size_t ZSTD_DCtx_loadDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); |
1107 | |
1108 | /*! ZSTD_DCtx_refDDict() : Requires v1.4.0+ |
1109 | * Reference a prepared dictionary, to be used to decompress next frames. |
1110 | * The dictionary remains active for decompression of future frames using same DCtx. |
1111 | * |
1112 | * If called with ZSTD_d_refMultipleDDicts enabled, repeated calls of this function |
1113 | * will store the DDict references in a table, and the DDict used for decompression |
1114 | * will be determined at decompression time, as per the dict ID in the frame. |
1115 | * The memory for the table is allocated on the first call to refDDict, and can be |
1116 | * freed with ZSTD_freeDCtx(). |
1117 | * |
1118 | * If called with ZSTD_d_refMultipleDDicts disabled (the default), only one dictionary |
1119 | * will be managed, and referencing a dictionary effectively "discards" any previous one. |
1120 | * |
1121 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
1122 | * Special: referencing a NULL DDict means "return to no-dictionary mode". |
1123 | * Note 2 : DDict is just referenced, its lifetime must outlive its usage from DCtx. |
1124 | */ |
1125 | ZSTDLIB_API size_t ZSTD_DCtx_refDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); |
1126 | |
1127 | /*! ZSTD_DCtx_refPrefix() : Requires v1.4.0+ |
1128 | * Reference a prefix (single-usage dictionary) to decompress next frame. |
1129 | * This is the reverse operation of ZSTD_CCtx_refPrefix(), |
1130 | * and must use the same prefix as the one used during compression. |
1131 | * Prefix is **only used once**. Reference is discarded at end of frame. |
1132 | * End of frame is reached when ZSTD_decompressStream() returns 0. |
1133 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
1134 | * Note 1 : Adding any prefix (including NULL) invalidates any previously set prefix or dictionary |
1135 | * Note 2 : Prefix buffer is referenced. It **must** outlive decompression. |
1136 | * Prefix buffer must remain unmodified up to the end of frame, |
1137 | * reached when ZSTD_decompressStream() returns 0. |
1138 | * Note 3 : By default, the prefix is treated as raw content (ZSTD_dct_rawContent). |
1139 | * Use ZSTD_CCtx_refPrefix_advanced() to alter dictMode (Experimental section) |
1140 | * Note 4 : Referencing a raw content prefix has almost no cpu nor memory cost. |
1141 | * A full dictionary is more costly, as it requires building tables. |
1142 | */ |
1143 | ZSTDLIB_API size_t ZSTD_DCtx_refPrefix(ZSTD_DCtx* dctx, |
1144 | const void* prefix, size_t prefixSize); |
1145 | |
1146 | /* === Memory management === */ |
1147 | |
1148 | /*! ZSTD_sizeof_*() : Requires v1.4.0+ |
1149 | * These functions give the _current_ memory usage of selected object. |
1150 | * Note that object memory usage can evolve (increase or decrease) over time. */ |
1151 | ZSTDLIB_API size_t ZSTD_sizeof_CCtx(const ZSTD_CCtx* cctx); |
1152 | ZSTDLIB_API size_t ZSTD_sizeof_DCtx(const ZSTD_DCtx* dctx); |
1153 | ZSTDLIB_API size_t ZSTD_sizeof_CStream(const ZSTD_CStream* zcs); |
1154 | ZSTDLIB_API size_t ZSTD_sizeof_DStream(const ZSTD_DStream* zds); |
1155 | ZSTDLIB_API size_t ZSTD_sizeof_CDict(const ZSTD_CDict* cdict); |
1156 | ZSTDLIB_API size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict); |
1157 | |
1158 | #endif /* ZSTD_H_235446 */ |
1159 | |
1160 | |
1161 | /* ************************************************************************************** |
1162 | * ADVANCED AND EXPERIMENTAL FUNCTIONS |
1163 | **************************************************************************************** |
1164 | * The definitions in the following section are considered experimental. |
1165 | * They are provided for advanced scenarios. |
1166 | * They should never be used with a dynamic library, as prototypes may change in the future. |
1167 | * Use them only in association with static linking. |
1168 | * ***************************************************************************************/ |
1169 | |
1170 | #if defined(ZSTD_STATIC_LINKING_ONLY) && !defined(ZSTD_H_ZSTD_STATIC_LINKING_ONLY) |
1171 | #define ZSTD_H_ZSTD_STATIC_LINKING_ONLY |
1172 | |
1173 | /* This can be overridden externally to hide static symbols. */ |
1174 | #ifndef ZSTDLIB_STATIC_API |
1175 | # if defined(ZSTD_DLL_EXPORT) && (ZSTD_DLL_EXPORT==1) |
1176 | # define ZSTDLIB_STATIC_API __declspec(dllexport) ZSTDLIB_VISIBLE |
1177 | # elif defined(ZSTD_DLL_IMPORT) && (ZSTD_DLL_IMPORT==1) |
1178 | # define ZSTDLIB_STATIC_API __declspec(dllimport) ZSTDLIB_VISIBLE |
1179 | # else |
1180 | # define ZSTDLIB_STATIC_API ZSTDLIB_VISIBLE |
1181 | # endif |
1182 | #endif |
1183 | |
1184 | /**************************************************************************************** |
1185 | * experimental API (static linking only) |
1186 | **************************************************************************************** |
1187 | * The following symbols and constants |
1188 | * are not planned to join "stable API" status in the near future. |
1189 | * They can still change in future versions. |
1190 | * Some of them are planned to remain in the static_only section indefinitely. |
1191 | * Some of them might be removed in the future (especially when redundant with existing stable functions) |
1192 | * ***************************************************************************************/ |
1193 | |
1194 | #define (format) ((format) == ZSTD_f_zstd1 ? 5 : 1) /* minimum input size required to query frame header size */ |
1195 | #define (format) ((format) == ZSTD_f_zstd1 ? 6 : 2) |
1196 | #define 18 /* can be useful for static allocation */ |
1197 | #define 8 |
1198 | |
1199 | /* compression parameter bounds */ |
1200 | #define ZSTD_WINDOWLOG_MAX_32 30 |
1201 | #define ZSTD_WINDOWLOG_MAX_64 31 |
1202 | #define ZSTD_WINDOWLOG_MAX ((int)(sizeof(size_t) == 4 ? ZSTD_WINDOWLOG_MAX_32 : ZSTD_WINDOWLOG_MAX_64)) |
1203 | #define ZSTD_WINDOWLOG_MIN 10 |
1204 | #define ZSTD_HASHLOG_MAX ((ZSTD_WINDOWLOG_MAX < 30) ? ZSTD_WINDOWLOG_MAX : 30) |
1205 | #define ZSTD_HASHLOG_MIN 6 |
1206 | #define ZSTD_CHAINLOG_MAX_32 29 |
1207 | #define ZSTD_CHAINLOG_MAX_64 30 |
1208 | #define ZSTD_CHAINLOG_MAX ((int)(sizeof(size_t) == 4 ? ZSTD_CHAINLOG_MAX_32 : ZSTD_CHAINLOG_MAX_64)) |
1209 | #define ZSTD_CHAINLOG_MIN ZSTD_HASHLOG_MIN |
1210 | #define ZSTD_SEARCHLOG_MAX (ZSTD_WINDOWLOG_MAX-1) |
1211 | #define ZSTD_SEARCHLOG_MIN 1 |
1212 | #define ZSTD_MINMATCH_MAX 7 /* only for ZSTD_fast, other strategies are limited to 6 */ |
1213 | #define ZSTD_MINMATCH_MIN 3 /* only for ZSTD_btopt+, faster strategies are limited to 4 */ |
1214 | #define ZSTD_TARGETLENGTH_MAX ZSTD_BLOCKSIZE_MAX |
1215 | #define ZSTD_TARGETLENGTH_MIN 0 /* note : comparing this constant to an unsigned results in a tautological test */ |
1216 | #define ZSTD_STRATEGY_MIN ZSTD_fast |
1217 | #define ZSTD_STRATEGY_MAX ZSTD_btultra2 |
1218 | #define ZSTD_BLOCKSIZE_MAX_MIN (1 << 10) /* The minimum valid max blocksize. Maximum blocksizes smaller than this make compressBound() inaccurate. */ |
1219 | |
1220 | |
1221 | #define ZSTD_OVERLAPLOG_MIN 0 |
1222 | #define ZSTD_OVERLAPLOG_MAX 9 |
1223 | |
1224 | #define ZSTD_WINDOWLOG_LIMIT_DEFAULT 27 /* by default, the streaming decoder will refuse any frame |
1225 | * requiring larger than (1<<ZSTD_WINDOWLOG_LIMIT_DEFAULT) window size, |
1226 | * to preserve host's memory from unreasonable requirements. |
1227 | * This limit can be overridden using ZSTD_DCtx_setParameter(,ZSTD_d_windowLogMax,). |
1228 | * The limit does not apply for one-pass decoders (such as ZSTD_decompress()), since no additional memory is allocated */ |
1229 | |
1230 | |
1231 | /* LDM parameter bounds */ |
1232 | #define ZSTD_LDM_HASHLOG_MIN ZSTD_HASHLOG_MIN |
1233 | #define ZSTD_LDM_HASHLOG_MAX ZSTD_HASHLOG_MAX |
1234 | #define ZSTD_LDM_MINMATCH_MIN 4 |
1235 | #define ZSTD_LDM_MINMATCH_MAX 4096 |
1236 | #define ZSTD_LDM_BUCKETSIZELOG_MIN 1 |
1237 | #define ZSTD_LDM_BUCKETSIZELOG_MAX 8 |
1238 | #define ZSTD_LDM_HASHRATELOG_MIN 0 |
1239 | #define ZSTD_LDM_HASHRATELOG_MAX (ZSTD_WINDOWLOG_MAX - ZSTD_HASHLOG_MIN) |
1240 | |
1241 | /* Advanced parameter bounds */ |
1242 | #define ZSTD_TARGETCBLOCKSIZE_MIN 64 |
1243 | #define ZSTD_TARGETCBLOCKSIZE_MAX ZSTD_BLOCKSIZE_MAX |
1244 | #define ZSTD_SRCSIZEHINT_MIN 0 |
1245 | #define ZSTD_SRCSIZEHINT_MAX INT_MAX |
1246 | |
1247 | |
1248 | /* --- Advanced types --- */ |
1249 | |
1250 | typedef struct ZSTD_CCtx_params_s ZSTD_CCtx_params; |
1251 | |
1252 | typedef struct { |
1253 | unsigned int offset; /* The offset of the match. (NOT the same as the offset code) |
1254 | * If offset == 0 and matchLength == 0, this sequence represents the last |
1255 | * literals in the block of litLength size. |
1256 | */ |
1257 | |
1258 | unsigned int litLength; /* Literal length of the sequence. */ |
1259 | unsigned int matchLength; /* Match length of the sequence. */ |
1260 | |
1261 | /* Note: Users of this API may provide a sequence with matchLength == litLength == offset == 0. |
1262 | * In this case, we will treat the sequence as a marker for a block boundary. |
1263 | */ |
1264 | |
1265 | unsigned int rep; /* Represents which repeat offset is represented by the field 'offset'. |
1266 | * Ranges from [0, 3]. |
1267 | * |
1268 | * Repeat offsets are essentially previous offsets from previous sequences sorted in |
1269 | * recency order. For more detail, see doc/zstd_compression_format.md |
1270 | * |
1271 | * If rep == 0, then 'offset' does not contain a repeat offset. |
1272 | * If rep > 0: |
1273 | * If litLength != 0: |
1274 | * rep == 1 --> offset == repeat_offset_1 |
1275 | * rep == 2 --> offset == repeat_offset_2 |
1276 | * rep == 3 --> offset == repeat_offset_3 |
1277 | * If litLength == 0: |
1278 | * rep == 1 --> offset == repeat_offset_2 |
1279 | * rep == 2 --> offset == repeat_offset_3 |
1280 | * rep == 3 --> offset == repeat_offset_1 - 1 |
1281 | * |
1282 | * Note: This field is optional. ZSTD_generateSequences() will calculate the value of |
1283 | * 'rep', but repeat offsets do not necessarily need to be calculated from an external |
1284 | * sequence provider's perspective. For example, ZSTD_compressSequences() does not |
1285 | * use this 'rep' field at all (as of now). |
1286 | */ |
1287 | } ZSTD_Sequence; |
1288 | |
1289 | typedef struct { |
1290 | unsigned windowLog; /**< largest match distance : larger == more compression, more memory needed during decompression */ |
1291 | unsigned chainLog; /**< fully searched segment : larger == more compression, slower, more memory (useless for fast) */ |
1292 | unsigned hashLog; /**< dispatch table : larger == faster, more memory */ |
1293 | unsigned searchLog; /**< nb of searches : larger == more compression, slower */ |
1294 | unsigned minMatch; /**< match length searched : larger == faster decompression, sometimes less compression */ |
1295 | unsigned targetLength; /**< acceptable match size for optimal parser (only) : larger == more compression, slower */ |
1296 | ZSTD_strategy strategy; /**< see ZSTD_strategy definition above */ |
1297 | } ZSTD_compressionParameters; |
1298 | |
1299 | typedef struct { |
1300 | int contentSizeFlag; /**< 1: content size will be in frame header (when known) */ |
1301 | int checksumFlag; /**< 1: generate a 32-bits checksum using XXH64 algorithm at end of frame, for error detection */ |
1302 | int noDictIDFlag; /**< 1: no dictID will be saved into frame header (dictID is only useful for dictionary compression) */ |
1303 | } ZSTD_frameParameters; |
1304 | |
1305 | typedef struct { |
1306 | ZSTD_compressionParameters cParams; |
1307 | ZSTD_frameParameters fParams; |
1308 | } ZSTD_parameters; |
1309 | |
1310 | typedef enum { |
1311 | ZSTD_dct_auto = 0, /* dictionary is "full" when starting with ZSTD_MAGIC_DICTIONARY, otherwise it is "rawContent" */ |
1312 | ZSTD_dct_rawContent = 1, /* ensures dictionary is always loaded as rawContent, even if it starts with ZSTD_MAGIC_DICTIONARY */ |
1313 | ZSTD_dct_fullDict = 2 /* refuses to load a dictionary if it does not respect Zstandard's specification, starting with ZSTD_MAGIC_DICTIONARY */ |
1314 | } ZSTD_dictContentType_e; |
1315 | |
1316 | typedef enum { |
1317 | ZSTD_dlm_byCopy = 0, /**< Copy dictionary content internally */ |
1318 | ZSTD_dlm_byRef = 1 /**< Reference dictionary content -- the dictionary buffer must outlive its users. */ |
1319 | } ZSTD_dictLoadMethod_e; |
1320 | |
1321 | typedef enum { |
1322 | ZSTD_f_zstd1 = 0, /* zstd frame format, specified in zstd_compression_format.md (default) */ |
1323 | ZSTD_f_zstd1_magicless = 1 /* Variant of zstd frame format, without initial 4-bytes magic number. |
1324 | * Useful to save 4 bytes per generated frame. |
1325 | * Decoder cannot recognise automatically this format, requiring this instruction. */ |
1326 | } ZSTD_format_e; |
1327 | |
1328 | typedef enum { |
1329 | /* Note: this enum controls ZSTD_d_forceIgnoreChecksum */ |
1330 | ZSTD_d_validateChecksum = 0, |
1331 | ZSTD_d_ignoreChecksum = 1 |
1332 | } ZSTD_forceIgnoreChecksum_e; |
1333 | |
1334 | typedef enum { |
1335 | /* Note: this enum controls ZSTD_d_refMultipleDDicts */ |
1336 | ZSTD_rmd_refSingleDDict = 0, |
1337 | ZSTD_rmd_refMultipleDDicts = 1 |
1338 | } ZSTD_refMultipleDDicts_e; |
1339 | |
1340 | typedef enum { |
1341 | /* Note: this enum and the behavior it controls are effectively internal |
1342 | * implementation details of the compressor. They are expected to continue |
1343 | * to evolve and should be considered only in the context of extremely |
1344 | * advanced performance tuning. |
1345 | * |
1346 | * Zstd currently supports the use of a CDict in three ways: |
1347 | * |
1348 | * - The contents of the CDict can be copied into the working context. This |
1349 | * means that the compression can search both the dictionary and input |
1350 | * while operating on a single set of internal tables. This makes |
1351 | * the compression faster per-byte of input. However, the initial copy of |
1352 | * the CDict's tables incurs a fixed cost at the beginning of the |
1353 | * compression. For small compressions (< 8 KB), that copy can dominate |
1354 | * the cost of the compression. |
1355 | * |
1356 | * - The CDict's tables can be used in-place. In this model, compression is |
1357 | * slower per input byte, because the compressor has to search two sets of |
1358 | * tables. However, this model incurs no start-up cost (as long as the |
1359 | * working context's tables can be reused). For small inputs, this can be |
1360 | * faster than copying the CDict's tables. |
1361 | * |
1362 | * - The CDict's tables are not used at all, and instead we use the working |
1363 | * context alone to reload the dictionary and use params based on the source |
1364 | * size. See ZSTD_compress_insertDictionary() and ZSTD_compress_usingDict(). |
1365 | * This method is effective when the dictionary sizes are very small relative |
1366 | * to the input size, and the input size is fairly large to begin with. |
1367 | * |
1368 | * Zstd has a simple internal heuristic that selects which strategy to use |
1369 | * at the beginning of a compression. However, if experimentation shows that |
1370 | * Zstd is making poor choices, it is possible to override that choice with |
1371 | * this enum. |
1372 | */ |
1373 | ZSTD_dictDefaultAttach = 0, /* Use the default heuristic. */ |
1374 | ZSTD_dictForceAttach = 1, /* Never copy the dictionary. */ |
1375 | ZSTD_dictForceCopy = 2, /* Always copy the dictionary. */ |
1376 | ZSTD_dictForceLoad = 3 /* Always reload the dictionary */ |
1377 | } ZSTD_dictAttachPref_e; |
1378 | |
1379 | typedef enum { |
1380 | ZSTD_lcm_auto = 0, /**< Automatically determine the compression mode based on the compression level. |
1381 | * Negative compression levels will be uncompressed, and positive compression |
1382 | * levels will be compressed. */ |
1383 | ZSTD_lcm_huffman = 1, /**< Always attempt Huffman compression. Uncompressed literals will still be |
1384 | * emitted if Huffman compression is not profitable. */ |
1385 | ZSTD_lcm_uncompressed = 2 /**< Always emit uncompressed literals. */ |
1386 | } ZSTD_literalCompressionMode_e; |
1387 | |
1388 | typedef enum { |
1389 | /* Note: This enum controls features which are conditionally beneficial. Zstd typically will make a final |
1390 | * decision on whether or not to enable the feature (ZSTD_ps_auto), but setting the switch to ZSTD_ps_enable |
1391 | * or ZSTD_ps_disable allow for a force enable/disable the feature. |
1392 | */ |
1393 | ZSTD_ps_auto = 0, /* Let the library automatically determine whether the feature shall be enabled */ |
1394 | ZSTD_ps_enable = 1, /* Force-enable the feature */ |
1395 | ZSTD_ps_disable = 2 /* Do not use the feature */ |
1396 | } ZSTD_paramSwitch_e; |
1397 | |
1398 | /*************************************** |
1399 | * Frame header and size functions |
1400 | ***************************************/ |
1401 | |
1402 | /*! ZSTD_findDecompressedSize() : |
1403 | * `src` should point to the start of a series of ZSTD encoded and/or skippable frames |
1404 | * `srcSize` must be the _exact_ size of this series |
1405 | * (i.e. there should be a frame boundary at `src + srcSize`) |
1406 | * @return : - decompressed size of all data in all successive frames |
1407 | * - if the decompressed size cannot be determined: ZSTD_CONTENTSIZE_UNKNOWN |
1408 | * - if an error occurred: ZSTD_CONTENTSIZE_ERROR |
1409 | * |
1410 | * note 1 : decompressed size is an optional field, that may not be present, especially in streaming mode. |
1411 | * When `return==ZSTD_CONTENTSIZE_UNKNOWN`, data to decompress could be any size. |
1412 | * In which case, it's necessary to use streaming mode to decompress data. |
1413 | * note 2 : decompressed size is always present when compression is done with ZSTD_compress() |
1414 | * note 3 : decompressed size can be very large (64-bits value), |
1415 | * potentially larger than what local system can handle as a single memory segment. |
1416 | * In which case, it's necessary to use streaming mode to decompress data. |
1417 | * note 4 : If source is untrusted, decompressed size could be wrong or intentionally modified. |
1418 | * Always ensure result fits within application's authorized limits. |
1419 | * Each application can set its own limits. |
1420 | * note 5 : ZSTD_findDecompressedSize handles multiple frames, and so it must traverse the input to |
1421 | * read each contained frame header. This is fast as most of the data is skipped, |
1422 | * however it does mean that all frame data must be present and valid. */ |
1423 | ZSTDLIB_STATIC_API unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize); |
1424 | |
1425 | /*! ZSTD_decompressBound() : |
1426 | * `src` should point to the start of a series of ZSTD encoded and/or skippable frames |
1427 | * `srcSize` must be the _exact_ size of this series |
1428 | * (i.e. there should be a frame boundary at `src + srcSize`) |
1429 | * @return : - upper-bound for the decompressed size of all data in all successive frames |
1430 | * - if an error occurred: ZSTD_CONTENTSIZE_ERROR |
1431 | * |
1432 | * note 1 : an error can occur if `src` contains an invalid or incorrectly formatted frame. |
1433 | * note 2 : the upper-bound is exact when the decompressed size field is available in every ZSTD encoded frame of `src`. |
1434 | * in this case, `ZSTD_findDecompressedSize` and `ZSTD_decompressBound` return the same value. |
1435 | * note 3 : when the decompressed size field isn't available, the upper-bound for that frame is calculated by: |
1436 | * upper-bound = # blocks * min(128 KB, Window_Size) |
1437 | */ |
1438 | ZSTDLIB_STATIC_API unsigned long long ZSTD_decompressBound(const void* src, size_t srcSize); |
1439 | |
1440 | /*! ZSTD_frameHeaderSize() : |
1441 | * srcSize must be >= ZSTD_FRAMEHEADERSIZE_PREFIX. |
1442 | * @return : size of the Frame Header, |
1443 | * or an error code (if srcSize is too small) */ |
1444 | ZSTDLIB_STATIC_API size_t (const void* src, size_t srcSize); |
1445 | |
1446 | typedef enum { ZSTD_frame, ZSTD_skippableFrame } ZSTD_frameType_e; |
1447 | typedef struct { |
1448 | unsigned long long frameContentSize; /* if == ZSTD_CONTENTSIZE_UNKNOWN, it means this field is not available. 0 means "empty" */ |
1449 | unsigned long long windowSize; /* can be very large, up to <= frameContentSize */ |
1450 | unsigned blockSizeMax; |
1451 | ZSTD_frameType_e frameType; /* if == ZSTD_skippableFrame, frameContentSize is the size of skippable content */ |
1452 | unsigned ; |
1453 | unsigned dictID; |
1454 | unsigned checksumFlag; |
1455 | unsigned _reserved1; |
1456 | unsigned _reserved2; |
1457 | } ; |
1458 | |
1459 | /*! ZSTD_getFrameHeader() : |
1460 | * decode Frame Header, or requires larger `srcSize`. |
1461 | * @return : 0, `zfhPtr` is correctly filled, |
1462 | * >0, `srcSize` is too small, value is wanted `srcSize` amount, |
1463 | * or an error code, which can be tested using ZSTD_isError() */ |
1464 | ZSTDLIB_STATIC_API size_t (ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize); /**< doesn't consume input */ |
1465 | /*! ZSTD_getFrameHeader_advanced() : |
1466 | * same as ZSTD_getFrameHeader(), |
1467 | * with added capability to select a format (like ZSTD_f_zstd1_magicless) */ |
1468 | ZSTDLIB_STATIC_API size_t (ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format); |
1469 | |
1470 | /*! ZSTD_decompressionMargin() : |
1471 | * Zstd supports in-place decompression, where the input and output buffers overlap. |
1472 | * In this case, the output buffer must be at least (Margin + Output_Size) bytes large, |
1473 | * and the input buffer must be at the end of the output buffer. |
1474 | * |
1475 | * _______________________ Output Buffer ________________________ |
1476 | * | | |
1477 | * | ____ Input Buffer ____| |
1478 | * | | | |
1479 | * v v v |
1480 | * |---------------------------------------|-----------|----------| |
1481 | * ^ ^ ^ |
1482 | * |___________________ Output_Size ___________________|_ Margin _| |
1483 | * |
1484 | * NOTE: See also ZSTD_DECOMPRESSION_MARGIN(). |
1485 | * NOTE: This applies only to single-pass decompression through ZSTD_decompress() or |
1486 | * ZSTD_decompressDCtx(). |
1487 | * NOTE: This function supports multi-frame input. |
1488 | * |
1489 | * @param src The compressed frame(s) |
1490 | * @param srcSize The size of the compressed frame(s) |
1491 | * @returns The decompression margin or an error that can be checked with ZSTD_isError(). |
1492 | */ |
1493 | ZSTDLIB_STATIC_API size_t ZSTD_decompressionMargin(const void* src, size_t srcSize); |
1494 | |
1495 | /*! ZSTD_DECOMPRESS_MARGIN() : |
1496 | * Similar to ZSTD_decompressionMargin(), but instead of computing the margin from |
1497 | * the compressed frame, compute it from the original size and the blockSizeLog. |
1498 | * See ZSTD_decompressionMargin() for details. |
1499 | * |
1500 | * WARNING: This macro does not support multi-frame input, the input must be a single |
1501 | * zstd frame. If you need that support use the function, or implement it yourself. |
1502 | * |
1503 | * @param originalSize The original uncompressed size of the data. |
1504 | * @param blockSize The block size == MIN(windowSize, ZSTD_BLOCKSIZE_MAX). |
1505 | * Unless you explicitly set the windowLog smaller than |
1506 | * ZSTD_BLOCKSIZELOG_MAX you can just use ZSTD_BLOCKSIZE_MAX. |
1507 | */ |
1508 | #define ZSTD_DECOMPRESSION_MARGIN(originalSize, blockSize) ((size_t)( \ |
1509 | ZSTD_FRAMEHEADERSIZE_MAX /* Frame header */ + \ |
1510 | 4 /* checksum */ + \ |
1511 | ((originalSize) == 0 ? 0 : 3 * (((originalSize) + (blockSize) - 1) / blockSize)) /* 3 bytes per block */ + \ |
1512 | (blockSize) /* One block of margin */ \ |
1513 | )) |
1514 | |
1515 | typedef enum { |
1516 | ZSTD_sf_noBlockDelimiters = 0, /* Representation of ZSTD_Sequence has no block delimiters, sequences only */ |
1517 | ZSTD_sf_explicitBlockDelimiters = 1 /* Representation of ZSTD_Sequence contains explicit block delimiters */ |
1518 | } ZSTD_sequenceFormat_e; |
1519 | |
1520 | /*! ZSTD_sequenceBound() : |
1521 | * `srcSize` : size of the input buffer |
1522 | * @return : upper-bound for the number of sequences that can be generated |
1523 | * from a buffer of srcSize bytes |
1524 | * |
1525 | * note : returns number of sequences - to get bytes, multiply by sizeof(ZSTD_Sequence). |
1526 | */ |
1527 | ZSTDLIB_STATIC_API size_t ZSTD_sequenceBound(size_t srcSize); |
1528 | |
1529 | /*! ZSTD_generateSequences() : |
1530 | * Generate sequences using ZSTD_compress2(), given a source buffer. |
1531 | * |
1532 | * Each block will end with a dummy sequence |
1533 | * with offset == 0, matchLength == 0, and litLength == length of last literals. |
1534 | * litLength may be == 0, and if so, then the sequence of (of: 0 ml: 0 ll: 0) |
1535 | * simply acts as a block delimiter. |
1536 | * |
1537 | * @zc can be used to insert custom compression params. |
1538 | * This function invokes ZSTD_compress2(). |
1539 | * |
1540 | * The output of this function can be fed into ZSTD_compressSequences() with CCtx |
1541 | * setting of ZSTD_c_blockDelimiters as ZSTD_sf_explicitBlockDelimiters |
1542 | * @return : number of sequences generated |
1543 | */ |
1544 | |
1545 | ZSTDLIB_STATIC_API size_t |
1546 | ZSTD_generateSequences( ZSTD_CCtx* zc, |
1547 | ZSTD_Sequence* outSeqs, size_t outSeqsSize, |
1548 | const void* src, size_t srcSize); |
1549 | |
1550 | /*! ZSTD_mergeBlockDelimiters() : |
1551 | * Given an array of ZSTD_Sequence, remove all sequences that represent block delimiters/last literals |
1552 | * by merging them into the literals of the next sequence. |
1553 | * |
1554 | * As such, the final generated result has no explicit representation of block boundaries, |
1555 | * and the final last literals segment is not represented in the sequences. |
1556 | * |
1557 | * The output of this function can be fed into ZSTD_compressSequences() with CCtx |
1558 | * setting of ZSTD_c_blockDelimiters as ZSTD_sf_noBlockDelimiters |
1559 | * @return : number of sequences left after merging |
1560 | */ |
1561 | ZSTDLIB_STATIC_API size_t ZSTD_mergeBlockDelimiters(ZSTD_Sequence* sequences, size_t seqsSize); |
1562 | |
1563 | /*! ZSTD_compressSequences() : |
1564 | * Compress an array of ZSTD_Sequence, associated with @src buffer, into dst. |
1565 | * @src contains the entire input (not just the literals). |
1566 | * If @srcSize > sum(sequence.length), the remaining bytes are considered all literals |
1567 | * If a dictionary is included, then the cctx should reference the dict. (see: ZSTD_CCtx_refCDict(), ZSTD_CCtx_loadDictionary(), etc.) |
1568 | * The entire source is compressed into a single frame. |
1569 | * |
1570 | * The compression behavior changes based on cctx params. In particular: |
1571 | * If ZSTD_c_blockDelimiters == ZSTD_sf_noBlockDelimiters, the array of ZSTD_Sequence is expected to contain |
1572 | * no block delimiters (defined in ZSTD_Sequence). Block boundaries are roughly determined based on |
1573 | * the block size derived from the cctx, and sequences may be split. This is the default setting. |
1574 | * |
1575 | * If ZSTD_c_blockDelimiters == ZSTD_sf_explicitBlockDelimiters, the array of ZSTD_Sequence is expected to contain |
1576 | * block delimiters (defined in ZSTD_Sequence). Behavior is undefined if no block delimiters are provided. |
1577 | * |
1578 | * If ZSTD_c_validateSequences == 0, this function will blindly accept the sequences provided. Invalid sequences cause undefined |
1579 | * behavior. If ZSTD_c_validateSequences == 1, then if sequence is invalid (see doc/zstd_compression_format.md for |
1580 | * specifics regarding offset/matchlength requirements) then the function will bail out and return an error. |
1581 | * |
1582 | * In addition to the two adjustable experimental params, there are other important cctx params. |
1583 | * - ZSTD_c_minMatch MUST be set as less than or equal to the smallest match generated by the match finder. It has a minimum value of ZSTD_MINMATCH_MIN. |
1584 | * - ZSTD_c_compressionLevel accordingly adjusts the strength of the entropy coder, as it would in typical compression. |
1585 | * - ZSTD_c_windowLog affects offset validation: this function will return an error at higher debug levels if a provided offset |
1586 | * is larger than what the spec allows for a given window log and dictionary (if present). See: doc/zstd_compression_format.md |
1587 | * |
1588 | * Note: Repcodes are, as of now, always re-calculated within this function, so ZSTD_Sequence::rep is unused. |
1589 | * Note 2: Once we integrate ability to ingest repcodes, the explicit block delims mode must respect those repcodes exactly, |
1590 | * and cannot emit an RLE block that disagrees with the repcode history |
1591 | * @return : final compressed size, or a ZSTD error code. |
1592 | */ |
1593 | ZSTDLIB_STATIC_API size_t |
1594 | ZSTD_compressSequences( ZSTD_CCtx* cctx, void* dst, size_t dstSize, |
1595 | const ZSTD_Sequence* inSeqs, size_t inSeqsSize, |
1596 | const void* src, size_t srcSize); |
1597 | |
1598 | |
1599 | /*! ZSTD_writeSkippableFrame() : |
1600 | * Generates a zstd skippable frame containing data given by src, and writes it to dst buffer. |
1601 | * |
1602 | * Skippable frames begin with a 4-byte magic number. There are 16 possible choices of magic number, |
1603 | * ranging from ZSTD_MAGIC_SKIPPABLE_START to ZSTD_MAGIC_SKIPPABLE_START+15. |
1604 | * As such, the parameter magicVariant controls the exact skippable frame magic number variant used, so |
1605 | * the magic number used will be ZSTD_MAGIC_SKIPPABLE_START + magicVariant. |
1606 | * |
1607 | * Returns an error if destination buffer is not large enough, if the source size is not representable |
1608 | * with a 4-byte unsigned int, or if the parameter magicVariant is greater than 15 (and therefore invalid). |
1609 | * |
1610 | * @return : number of bytes written or a ZSTD error. |
1611 | */ |
1612 | ZSTDLIB_STATIC_API size_t ZSTD_writeSkippableFrame(void* dst, size_t dstCapacity, |
1613 | const void* src, size_t srcSize, unsigned magicVariant); |
1614 | |
1615 | /*! ZSTD_readSkippableFrame() : |
1616 | * Retrieves a zstd skippable frame containing data given by src, and writes it to dst buffer. |
1617 | * |
1618 | * The parameter magicVariant will receive the magicVariant that was supplied when the frame was written, |
1619 | * i.e. magicNumber - ZSTD_MAGIC_SKIPPABLE_START. This can be NULL if the caller is not interested |
1620 | * in the magicVariant. |
1621 | * |
1622 | * Returns an error if destination buffer is not large enough, or if the frame is not skippable. |
1623 | * |
1624 | * @return : number of bytes written or a ZSTD error. |
1625 | */ |
1626 | ZSTDLIB_API size_t ZSTD_readSkippableFrame(void* dst, size_t dstCapacity, unsigned* magicVariant, |
1627 | const void* src, size_t srcSize); |
1628 | |
1629 | /*! ZSTD_isSkippableFrame() : |
1630 | * Tells if the content of `buffer` starts with a valid Frame Identifier for a skippable frame. |
1631 | */ |
1632 | ZSTDLIB_API unsigned ZSTD_isSkippableFrame(const void* buffer, size_t size); |
1633 | |
1634 | |
1635 | |
1636 | /*************************************** |
1637 | * Memory management |
1638 | ***************************************/ |
1639 | |
1640 | /*! ZSTD_estimate*() : |
1641 | * These functions make it possible to estimate memory usage |
1642 | * of a future {D,C}Ctx, before its creation. |
1643 | * |
1644 | * ZSTD_estimateCCtxSize() will provide a memory budget large enough |
1645 | * for any compression level up to selected one. |
1646 | * Note : Unlike ZSTD_estimateCStreamSize*(), this estimate |
1647 | * does not include space for a window buffer. |
1648 | * Therefore, the estimation is only guaranteed for single-shot compressions, not streaming. |
1649 | * The estimate will assume the input may be arbitrarily large, |
1650 | * which is the worst case. |
1651 | * |
1652 | * When srcSize can be bound by a known and rather "small" value, |
1653 | * this fact can be used to provide a tighter estimation |
1654 | * because the CCtx compression context will need less memory. |
1655 | * This tighter estimation can be provided by more advanced functions |
1656 | * ZSTD_estimateCCtxSize_usingCParams(), which can be used in tandem with ZSTD_getCParams(), |
1657 | * and ZSTD_estimateCCtxSize_usingCCtxParams(), which can be used in tandem with ZSTD_CCtxParams_setParameter(). |
1658 | * Both can be used to estimate memory using custom compression parameters and arbitrary srcSize limits. |
1659 | * |
1660 | * Note : only single-threaded compression is supported. |
1661 | * ZSTD_estimateCCtxSize_usingCCtxParams() will return an error code if ZSTD_c_nbWorkers is >= 1. |
1662 | * |
1663 | * Note 2 : ZSTD_estimateCCtxSize* functions are not compatible with the Block-Level Sequence Producer API at this time. |
1664 | * Size estimates assume that no external sequence producer is registered. |
1665 | */ |
1666 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCCtxSize(int compressionLevel); |
1667 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCCtxSize_usingCParams(ZSTD_compressionParameters cParams); |
1668 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCCtxSize_usingCCtxParams(const ZSTD_CCtx_params* params); |
1669 | ZSTDLIB_STATIC_API size_t ZSTD_estimateDCtxSize(void); |
1670 | |
1671 | /*! ZSTD_estimateCStreamSize() : |
1672 | * ZSTD_estimateCStreamSize() will provide a budget large enough for any compression level up to selected one. |
1673 | * It will also consider src size to be arbitrarily "large", which is worst case. |
1674 | * If srcSize is known to always be small, ZSTD_estimateCStreamSize_usingCParams() can provide a tighter estimation. |
1675 | * ZSTD_estimateCStreamSize_usingCParams() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. |
1676 | * ZSTD_estimateCStreamSize_usingCCtxParams() can be used in tandem with ZSTD_CCtxParams_setParameter(). Only single-threaded compression is supported. This function will return an error code if ZSTD_c_nbWorkers is >= 1. |
1677 | * Note : CStream size estimation is only correct for single-threaded compression. |
1678 | * ZSTD_DStream memory budget depends on window Size. |
1679 | * This information can be passed manually, using ZSTD_estimateDStreamSize, |
1680 | * or deducted from a valid frame Header, using ZSTD_estimateDStreamSize_fromFrame(); |
1681 | * Note : if streaming is init with function ZSTD_init?Stream_usingDict(), |
1682 | * an internal ?Dict will be created, which additional size is not estimated here. |
1683 | * In this case, get total size by adding ZSTD_estimate?DictSize |
1684 | * Note 2 : only single-threaded compression is supported. |
1685 | * ZSTD_estimateCStreamSize_usingCCtxParams() will return an error code if ZSTD_c_nbWorkers is >= 1. |
1686 | * Note 3 : ZSTD_estimateCStreamSize* functions are not compatible with the Block-Level Sequence Producer API at this time. |
1687 | * Size estimates assume that no external sequence producer is registered. |
1688 | */ |
1689 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCStreamSize(int compressionLevel); |
1690 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCStreamSize_usingCParams(ZSTD_compressionParameters cParams); |
1691 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCStreamSize_usingCCtxParams(const ZSTD_CCtx_params* params); |
1692 | ZSTDLIB_STATIC_API size_t ZSTD_estimateDStreamSize(size_t windowSize); |
1693 | ZSTDLIB_STATIC_API size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize); |
1694 | |
1695 | /*! ZSTD_estimate?DictSize() : |
1696 | * ZSTD_estimateCDictSize() will bet that src size is relatively "small", and content is copied, like ZSTD_createCDict(). |
1697 | * ZSTD_estimateCDictSize_advanced() makes it possible to control compression parameters precisely, like ZSTD_createCDict_advanced(). |
1698 | * Note : dictionaries created by reference (`ZSTD_dlm_byRef`) are logically smaller. |
1699 | */ |
1700 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCDictSize(size_t dictSize, int compressionLevel); |
1701 | ZSTDLIB_STATIC_API size_t ZSTD_estimateCDictSize_advanced(size_t dictSize, ZSTD_compressionParameters cParams, ZSTD_dictLoadMethod_e dictLoadMethod); |
1702 | ZSTDLIB_STATIC_API size_t ZSTD_estimateDDictSize(size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod); |
1703 | |
1704 | /*! ZSTD_initStatic*() : |
1705 | * Initialize an object using a pre-allocated fixed-size buffer. |
1706 | * workspace: The memory area to emplace the object into. |
1707 | * Provided pointer *must be 8-bytes aligned*. |
1708 | * Buffer must outlive object. |
1709 | * workspaceSize: Use ZSTD_estimate*Size() to determine |
1710 | * how large workspace must be to support target scenario. |
1711 | * @return : pointer to object (same address as workspace, just different type), |
1712 | * or NULL if error (size too small, incorrect alignment, etc.) |
1713 | * Note : zstd will never resize nor malloc() when using a static buffer. |
1714 | * If the object requires more memory than available, |
1715 | * zstd will just error out (typically ZSTD_error_memory_allocation). |
1716 | * Note 2 : there is no corresponding "free" function. |
1717 | * Since workspace is allocated externally, it must be freed externally too. |
1718 | * Note 3 : cParams : use ZSTD_getCParams() to convert a compression level |
1719 | * into its associated cParams. |
1720 | * Limitation 1 : currently not compatible with internal dictionary creation, triggered by |
1721 | * ZSTD_CCtx_loadDictionary(), ZSTD_initCStream_usingDict() or ZSTD_initDStream_usingDict(). |
1722 | * Limitation 2 : static cctx currently not compatible with multi-threading. |
1723 | * Limitation 3 : static dctx is incompatible with legacy support. |
1724 | */ |
1725 | ZSTDLIB_STATIC_API ZSTD_CCtx* ZSTD_initStaticCCtx(void* workspace, size_t workspaceSize); |
1726 | ZSTDLIB_STATIC_API ZSTD_CStream* ZSTD_initStaticCStream(void* workspace, size_t workspaceSize); /**< same as ZSTD_initStaticCCtx() */ |
1727 | |
1728 | ZSTDLIB_STATIC_API ZSTD_DCtx* ZSTD_initStaticDCtx(void* workspace, size_t workspaceSize); |
1729 | ZSTDLIB_STATIC_API ZSTD_DStream* ZSTD_initStaticDStream(void* workspace, size_t workspaceSize); /**< same as ZSTD_initStaticDCtx() */ |
1730 | |
1731 | ZSTDLIB_STATIC_API const ZSTD_CDict* ZSTD_initStaticCDict( |
1732 | void* workspace, size_t workspaceSize, |
1733 | const void* dict, size_t dictSize, |
1734 | ZSTD_dictLoadMethod_e dictLoadMethod, |
1735 | ZSTD_dictContentType_e dictContentType, |
1736 | ZSTD_compressionParameters cParams); |
1737 | |
1738 | ZSTDLIB_STATIC_API const ZSTD_DDict* ZSTD_initStaticDDict( |
1739 | void* workspace, size_t workspaceSize, |
1740 | const void* dict, size_t dictSize, |
1741 | ZSTD_dictLoadMethod_e dictLoadMethod, |
1742 | ZSTD_dictContentType_e dictContentType); |
1743 | |
1744 | |
1745 | /*! Custom memory allocation : |
1746 | * These prototypes make it possible to pass your own allocation/free functions. |
1747 | * ZSTD_customMem is provided at creation time, using ZSTD_create*_advanced() variants listed below. |
1748 | * All allocation/free operations will be completed using these custom variants instead of regular <stdlib.h> ones. |
1749 | */ |
1750 | typedef void* (*ZSTD_allocFunction) (void* opaque, size_t size); |
1751 | typedef void (*ZSTD_freeFunction) (void* opaque, void* address); |
1752 | typedef struct { ZSTD_allocFunction customAlloc; ZSTD_freeFunction customFree; void* opaque; } ZSTD_customMem; |
1753 | static |
1754 | #ifdef __GNUC__ |
1755 | __attribute__((__unused__)) |
1756 | #endif |
1757 | ZSTD_customMem const ZSTD_defaultCMem = { NULL, NULL, NULL }; /**< this constant defers to stdlib's functions */ |
1758 | |
1759 | ZSTDLIB_STATIC_API ZSTD_CCtx* ZSTD_createCCtx_advanced(ZSTD_customMem customMem); |
1760 | ZSTDLIB_STATIC_API ZSTD_CStream* ZSTD_createCStream_advanced(ZSTD_customMem customMem); |
1761 | ZSTDLIB_STATIC_API ZSTD_DCtx* ZSTD_createDCtx_advanced(ZSTD_customMem customMem); |
1762 | ZSTDLIB_STATIC_API ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem); |
1763 | |
1764 | ZSTDLIB_STATIC_API ZSTD_CDict* ZSTD_createCDict_advanced(const void* dict, size_t dictSize, |
1765 | ZSTD_dictLoadMethod_e dictLoadMethod, |
1766 | ZSTD_dictContentType_e dictContentType, |
1767 | ZSTD_compressionParameters cParams, |
1768 | ZSTD_customMem customMem); |
1769 | |
1770 | /*! Thread pool : |
1771 | * These prototypes make it possible to share a thread pool among multiple compression contexts. |
1772 | * This can limit resources for applications with multiple threads where each one uses |
1773 | * a threaded compression mode (via ZSTD_c_nbWorkers parameter). |
1774 | * ZSTD_createThreadPool creates a new thread pool with a given number of threads. |
1775 | * Note that the lifetime of such pool must exist while being used. |
1776 | * ZSTD_CCtx_refThreadPool assigns a thread pool to a context (use NULL argument value |
1777 | * to use an internal thread pool). |
1778 | * ZSTD_freeThreadPool frees a thread pool, accepts NULL pointer. |
1779 | */ |
1780 | typedef struct POOL_ctx_s ZSTD_threadPool; |
1781 | ZSTDLIB_STATIC_API ZSTD_threadPool* ZSTD_createThreadPool(size_t numThreads); |
1782 | ZSTDLIB_STATIC_API void ZSTD_freeThreadPool (ZSTD_threadPool* pool); /* accept NULL pointer */ |
1783 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_refThreadPool(ZSTD_CCtx* cctx, ZSTD_threadPool* pool); |
1784 | |
1785 | |
1786 | /* |
1787 | * This API is temporary and is expected to change or disappear in the future! |
1788 | */ |
1789 | ZSTDLIB_STATIC_API ZSTD_CDict* ZSTD_createCDict_advanced2( |
1790 | const void* dict, size_t dictSize, |
1791 | ZSTD_dictLoadMethod_e dictLoadMethod, |
1792 | ZSTD_dictContentType_e dictContentType, |
1793 | const ZSTD_CCtx_params* cctxParams, |
1794 | ZSTD_customMem customMem); |
1795 | |
1796 | ZSTDLIB_STATIC_API ZSTD_DDict* ZSTD_createDDict_advanced( |
1797 | const void* dict, size_t dictSize, |
1798 | ZSTD_dictLoadMethod_e dictLoadMethod, |
1799 | ZSTD_dictContentType_e dictContentType, |
1800 | ZSTD_customMem customMem); |
1801 | |
1802 | |
1803 | /*************************************** |
1804 | * Advanced compression functions |
1805 | ***************************************/ |
1806 | |
1807 | /*! ZSTD_createCDict_byReference() : |
1808 | * Create a digested dictionary for compression |
1809 | * Dictionary content is just referenced, not duplicated. |
1810 | * As a consequence, `dictBuffer` **must** outlive CDict, |
1811 | * and its content must remain unmodified throughout the lifetime of CDict. |
1812 | * note: equivalent to ZSTD_createCDict_advanced(), with dictLoadMethod==ZSTD_dlm_byRef */ |
1813 | ZSTDLIB_STATIC_API ZSTD_CDict* ZSTD_createCDict_byReference(const void* dictBuffer, size_t dictSize, int compressionLevel); |
1814 | |
1815 | /*! ZSTD_getCParams() : |
1816 | * @return ZSTD_compressionParameters structure for a selected compression level and estimated srcSize. |
1817 | * `estimatedSrcSize` value is optional, select 0 if not known */ |
1818 | ZSTDLIB_STATIC_API ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, unsigned long long estimatedSrcSize, size_t dictSize); |
1819 | |
1820 | /*! ZSTD_getParams() : |
1821 | * same as ZSTD_getCParams(), but @return a full `ZSTD_parameters` object instead of sub-component `ZSTD_compressionParameters`. |
1822 | * All fields of `ZSTD_frameParameters` are set to default : contentSize=1, checksum=0, noDictID=0 */ |
1823 | ZSTDLIB_STATIC_API ZSTD_parameters ZSTD_getParams(int compressionLevel, unsigned long long estimatedSrcSize, size_t dictSize); |
1824 | |
1825 | /*! ZSTD_checkCParams() : |
1826 | * Ensure param values remain within authorized range. |
1827 | * @return 0 on success, or an error code (can be checked with ZSTD_isError()) */ |
1828 | ZSTDLIB_STATIC_API size_t ZSTD_checkCParams(ZSTD_compressionParameters params); |
1829 | |
1830 | /*! ZSTD_adjustCParams() : |
1831 | * optimize params for a given `srcSize` and `dictSize`. |
1832 | * `srcSize` can be unknown, in which case use ZSTD_CONTENTSIZE_UNKNOWN. |
1833 | * `dictSize` must be `0` when there is no dictionary. |
1834 | * cPar can be invalid : all parameters will be clamped within valid range in the @return struct. |
1835 | * This function never fails (wide contract) */ |
1836 | ZSTDLIB_STATIC_API ZSTD_compressionParameters ZSTD_adjustCParams(ZSTD_compressionParameters cPar, unsigned long long srcSize, size_t dictSize); |
1837 | |
1838 | /*! ZSTD_CCtx_setCParams() : |
1839 | * Set all parameters provided within @p cparams into the working @p cctx. |
1840 | * Note : if modifying parameters during compression (MT mode only), |
1841 | * note that changes to the .windowLog parameter will be ignored. |
1842 | * @return 0 on success, or an error code (can be checked with ZSTD_isError()). |
1843 | * On failure, no parameters are updated. |
1844 | */ |
1845 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setCParams(ZSTD_CCtx* cctx, ZSTD_compressionParameters cparams); |
1846 | |
1847 | /*! ZSTD_CCtx_setFParams() : |
1848 | * Set all parameters provided within @p fparams into the working @p cctx. |
1849 | * @return 0 on success, or an error code (can be checked with ZSTD_isError()). |
1850 | */ |
1851 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setFParams(ZSTD_CCtx* cctx, ZSTD_frameParameters fparams); |
1852 | |
1853 | /*! ZSTD_CCtx_setParams() : |
1854 | * Set all parameters provided within @p params into the working @p cctx. |
1855 | * @return 0 on success, or an error code (can be checked with ZSTD_isError()). |
1856 | */ |
1857 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setParams(ZSTD_CCtx* cctx, ZSTD_parameters params); |
1858 | |
1859 | /*! ZSTD_compress_advanced() : |
1860 | * Note : this function is now DEPRECATED. |
1861 | * It can be replaced by ZSTD_compress2(), in combination with ZSTD_CCtx_setParameter() and other parameter setters. |
1862 | * This prototype will generate compilation warnings. */ |
1863 | ZSTD_DEPRECATED("use ZSTD_compress2" ) |
1864 | ZSTDLIB_STATIC_API |
1865 | size_t ZSTD_compress_advanced(ZSTD_CCtx* cctx, |
1866 | void* dst, size_t dstCapacity, |
1867 | const void* src, size_t srcSize, |
1868 | const void* dict,size_t dictSize, |
1869 | ZSTD_parameters params); |
1870 | |
1871 | /*! ZSTD_compress_usingCDict_advanced() : |
1872 | * Note : this function is now DEPRECATED. |
1873 | * It can be replaced by ZSTD_compress2(), in combination with ZSTD_CCtx_loadDictionary() and other parameter setters. |
1874 | * This prototype will generate compilation warnings. */ |
1875 | ZSTD_DEPRECATED("use ZSTD_compress2 with ZSTD_CCtx_loadDictionary" ) |
1876 | ZSTDLIB_STATIC_API |
1877 | size_t ZSTD_compress_usingCDict_advanced(ZSTD_CCtx* cctx, |
1878 | void* dst, size_t dstCapacity, |
1879 | const void* src, size_t srcSize, |
1880 | const ZSTD_CDict* cdict, |
1881 | ZSTD_frameParameters fParams); |
1882 | |
1883 | |
1884 | /*! ZSTD_CCtx_loadDictionary_byReference() : |
1885 | * Same as ZSTD_CCtx_loadDictionary(), but dictionary content is referenced, instead of being copied into CCtx. |
1886 | * It saves some memory, but also requires that `dict` outlives its usage within `cctx` */ |
1887 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_loadDictionary_byReference(ZSTD_CCtx* cctx, const void* dict, size_t dictSize); |
1888 | |
1889 | /*! ZSTD_CCtx_loadDictionary_advanced() : |
1890 | * Same as ZSTD_CCtx_loadDictionary(), but gives finer control over |
1891 | * how to load the dictionary (by copy ? by reference ?) |
1892 | * and how to interpret it (automatic ? force raw mode ? full mode only ?) */ |
1893 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_loadDictionary_advanced(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType); |
1894 | |
1895 | /*! ZSTD_CCtx_refPrefix_advanced() : |
1896 | * Same as ZSTD_CCtx_refPrefix(), but gives finer control over |
1897 | * how to interpret prefix content (automatic ? force raw mode (default) ? full mode only ?) */ |
1898 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_refPrefix_advanced(ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType); |
1899 | |
1900 | /* === experimental parameters === */ |
1901 | /* these parameters can be used with ZSTD_setParameter() |
1902 | * they are not guaranteed to remain supported in the future */ |
1903 | |
1904 | /* Enables rsyncable mode, |
1905 | * which makes compressed files more rsync friendly |
1906 | * by adding periodic synchronization points to the compressed data. |
1907 | * The target average block size is ZSTD_c_jobSize / 2. |
1908 | * It's possible to modify the job size to increase or decrease |
1909 | * the granularity of the synchronization point. |
1910 | * Once the jobSize is smaller than the window size, |
1911 | * it will result in compression ratio degradation. |
1912 | * NOTE 1: rsyncable mode only works when multithreading is enabled. |
1913 | * NOTE 2: rsyncable performs poorly in combination with long range mode, |
1914 | * since it will decrease the effectiveness of synchronization points, |
1915 | * though mileage may vary. |
1916 | * NOTE 3: Rsyncable mode limits maximum compression speed to ~400 MB/s. |
1917 | * If the selected compression level is already running significantly slower, |
1918 | * the overall speed won't be significantly impacted. |
1919 | */ |
1920 | #define ZSTD_c_rsyncable ZSTD_c_experimentalParam1 |
1921 | |
1922 | /* Select a compression format. |
1923 | * The value must be of type ZSTD_format_e. |
1924 | * See ZSTD_format_e enum definition for details */ |
1925 | #define ZSTD_c_format ZSTD_c_experimentalParam2 |
1926 | |
1927 | /* Force back-reference distances to remain < windowSize, |
1928 | * even when referencing into Dictionary content (default:0) */ |
1929 | #define ZSTD_c_forceMaxWindow ZSTD_c_experimentalParam3 |
1930 | |
1931 | /* Controls whether the contents of a CDict |
1932 | * are used in place, or copied into the working context. |
1933 | * Accepts values from the ZSTD_dictAttachPref_e enum. |
1934 | * See the comments on that enum for an explanation of the feature. */ |
1935 | #define ZSTD_c_forceAttachDict ZSTD_c_experimentalParam4 |
1936 | |
1937 | /* Controlled with ZSTD_paramSwitch_e enum. |
1938 | * Default is ZSTD_ps_auto. |
1939 | * Set to ZSTD_ps_disable to never compress literals. |
1940 | * Set to ZSTD_ps_enable to always compress literals. (Note: uncompressed literals |
1941 | * may still be emitted if huffman is not beneficial to use.) |
1942 | * |
1943 | * By default, in ZSTD_ps_auto, the library will decide at runtime whether to use |
1944 | * literals compression based on the compression parameters - specifically, |
1945 | * negative compression levels do not use literal compression. |
1946 | */ |
1947 | #define ZSTD_c_literalCompressionMode ZSTD_c_experimentalParam5 |
1948 | |
1949 | /* Tries to fit compressed block size to be around targetCBlockSize. |
1950 | * No target when targetCBlockSize == 0. |
1951 | * There is no guarantee on compressed block size (default:0) */ |
1952 | #define ZSTD_c_targetCBlockSize ZSTD_c_experimentalParam6 |
1953 | |
1954 | /* User's best guess of source size. |
1955 | * Hint is not valid when srcSizeHint == 0. |
1956 | * There is no guarantee that hint is close to actual source size, |
1957 | * but compression ratio may regress significantly if guess considerably underestimates */ |
1958 | #define ZSTD_c_srcSizeHint ZSTD_c_experimentalParam7 |
1959 | |
1960 | /* Controls whether the new and experimental "dedicated dictionary search |
1961 | * structure" can be used. This feature is still rough around the edges, be |
1962 | * prepared for surprising behavior! |
1963 | * |
1964 | * How to use it: |
1965 | * |
1966 | * When using a CDict, whether to use this feature or not is controlled at |
1967 | * CDict creation, and it must be set in a CCtxParams set passed into that |
1968 | * construction (via ZSTD_createCDict_advanced2()). A compression will then |
1969 | * use the feature or not based on how the CDict was constructed; the value of |
1970 | * this param, set in the CCtx, will have no effect. |
1971 | * |
1972 | * However, when a dictionary buffer is passed into a CCtx, such as via |
1973 | * ZSTD_CCtx_loadDictionary(), this param can be set on the CCtx to control |
1974 | * whether the CDict that is created internally can use the feature or not. |
1975 | * |
1976 | * What it does: |
1977 | * |
1978 | * Normally, the internal data structures of the CDict are analogous to what |
1979 | * would be stored in a CCtx after compressing the contents of a dictionary. |
1980 | * To an approximation, a compression using a dictionary can then use those |
1981 | * data structures to simply continue what is effectively a streaming |
1982 | * compression where the simulated compression of the dictionary left off. |
1983 | * Which is to say, the search structures in the CDict are normally the same |
1984 | * format as in the CCtx. |
1985 | * |
1986 | * It is possible to do better, since the CDict is not like a CCtx: the search |
1987 | * structures are written once during CDict creation, and then are only read |
1988 | * after that, while the search structures in the CCtx are both read and |
1989 | * written as the compression goes along. This means we can choose a search |
1990 | * structure for the dictionary that is read-optimized. |
1991 | * |
1992 | * This feature enables the use of that different structure. |
1993 | * |
1994 | * Note that some of the members of the ZSTD_compressionParameters struct have |
1995 | * different semantics and constraints in the dedicated search structure. It is |
1996 | * highly recommended that you simply set a compression level in the CCtxParams |
1997 | * you pass into the CDict creation call, and avoid messing with the cParams |
1998 | * directly. |
1999 | * |
2000 | * Effects: |
2001 | * |
2002 | * This will only have any effect when the selected ZSTD_strategy |
2003 | * implementation supports this feature. Currently, that's limited to |
2004 | * ZSTD_greedy, ZSTD_lazy, and ZSTD_lazy2. |
2005 | * |
2006 | * Note that this means that the CDict tables can no longer be copied into the |
2007 | * CCtx, so the dict attachment mode ZSTD_dictForceCopy will no longer be |
2008 | * usable. The dictionary can only be attached or reloaded. |
2009 | * |
2010 | * In general, you should expect compression to be faster--sometimes very much |
2011 | * so--and CDict creation to be slightly slower. Eventually, we will probably |
2012 | * make this mode the default. |
2013 | */ |
2014 | #define ZSTD_c_enableDedicatedDictSearch ZSTD_c_experimentalParam8 |
2015 | |
2016 | /* ZSTD_c_stableInBuffer |
2017 | * Experimental parameter. |
2018 | * Default is 0 == disabled. Set to 1 to enable. |
2019 | * |
2020 | * Tells the compressor that input data presented with ZSTD_inBuffer |
2021 | * will ALWAYS be the same between calls. |
2022 | * Technically, the @src pointer must never be changed, |
2023 | * and the @pos field can only be updated by zstd. |
2024 | * However, it's possible to increase the @size field, |
2025 | * allowing scenarios where more data can be appended after compressions starts. |
2026 | * These conditions are checked by the compressor, |
2027 | * and compression will fail if they are not respected. |
2028 | * Also, data in the ZSTD_inBuffer within the range [src, src + pos) |
2029 | * MUST not be modified during compression or it will result in data corruption. |
2030 | * |
2031 | * When this flag is enabled zstd won't allocate an input window buffer, |
2032 | * because the user guarantees it can reference the ZSTD_inBuffer until |
2033 | * the frame is complete. But, it will still allocate an output buffer |
2034 | * large enough to fit a block (see ZSTD_c_stableOutBuffer). This will also |
2035 | * avoid the memcpy() from the input buffer to the input window buffer. |
2036 | * |
2037 | * NOTE: So long as the ZSTD_inBuffer always points to valid memory, using |
2038 | * this flag is ALWAYS memory safe, and will never access out-of-bounds |
2039 | * memory. However, compression WILL fail if conditions are not respected. |
2040 | * |
2041 | * WARNING: The data in the ZSTD_inBuffer in the range [src, src + pos) MUST |
2042 | * not be modified during compression or it will result in data corruption. |
2043 | * This is because zstd needs to reference data in the ZSTD_inBuffer to find |
2044 | * matches. Normally zstd maintains its own window buffer for this purpose, |
2045 | * but passing this flag tells zstd to rely on user provided buffer instead. |
2046 | */ |
2047 | #define ZSTD_c_stableInBuffer ZSTD_c_experimentalParam9 |
2048 | |
2049 | /* ZSTD_c_stableOutBuffer |
2050 | * Experimental parameter. |
2051 | * Default is 0 == disabled. Set to 1 to enable. |
2052 | * |
2053 | * Tells he compressor that the ZSTD_outBuffer will not be resized between |
2054 | * calls. Specifically: (out.size - out.pos) will never grow. This gives the |
2055 | * compressor the freedom to say: If the compressed data doesn't fit in the |
2056 | * output buffer then return ZSTD_error_dstSizeTooSmall. This allows us to |
2057 | * always decompress directly into the output buffer, instead of decompressing |
2058 | * into an internal buffer and copying to the output buffer. |
2059 | * |
2060 | * When this flag is enabled zstd won't allocate an output buffer, because |
2061 | * it can write directly to the ZSTD_outBuffer. It will still allocate the |
2062 | * input window buffer (see ZSTD_c_stableInBuffer). |
2063 | * |
2064 | * Zstd will check that (out.size - out.pos) never grows and return an error |
2065 | * if it does. While not strictly necessary, this should prevent surprises. |
2066 | */ |
2067 | #define ZSTD_c_stableOutBuffer ZSTD_c_experimentalParam10 |
2068 | |
2069 | /* ZSTD_c_blockDelimiters |
2070 | * Default is 0 == ZSTD_sf_noBlockDelimiters. |
2071 | * |
2072 | * For use with sequence compression API: ZSTD_compressSequences(). |
2073 | * |
2074 | * Designates whether or not the given array of ZSTD_Sequence contains block delimiters |
2075 | * and last literals, which are defined as sequences with offset == 0 and matchLength == 0. |
2076 | * See the definition of ZSTD_Sequence for more specifics. |
2077 | */ |
2078 | #define ZSTD_c_blockDelimiters ZSTD_c_experimentalParam11 |
2079 | |
2080 | /* ZSTD_c_validateSequences |
2081 | * Default is 0 == disabled. Set to 1 to enable sequence validation. |
2082 | * |
2083 | * For use with sequence compression API: ZSTD_compressSequences(). |
2084 | * Designates whether or not we validate sequences provided to ZSTD_compressSequences() |
2085 | * during function execution. |
2086 | * |
2087 | * Without validation, providing a sequence that does not conform to the zstd spec will cause |
2088 | * undefined behavior, and may produce a corrupted block. |
2089 | * |
2090 | * With validation enabled, if sequence is invalid (see doc/zstd_compression_format.md for |
2091 | * specifics regarding offset/matchlength requirements) then the function will bail out and |
2092 | * return an error. |
2093 | * |
2094 | */ |
2095 | #define ZSTD_c_validateSequences ZSTD_c_experimentalParam12 |
2096 | |
2097 | /* ZSTD_c_useBlockSplitter |
2098 | * Controlled with ZSTD_paramSwitch_e enum. |
2099 | * Default is ZSTD_ps_auto. |
2100 | * Set to ZSTD_ps_disable to never use block splitter. |
2101 | * Set to ZSTD_ps_enable to always use block splitter. |
2102 | * |
2103 | * By default, in ZSTD_ps_auto, the library will decide at runtime whether to use |
2104 | * block splitting based on the compression parameters. |
2105 | */ |
2106 | #define ZSTD_c_useBlockSplitter ZSTD_c_experimentalParam13 |
2107 | |
2108 | /* ZSTD_c_useRowMatchFinder |
2109 | * Controlled with ZSTD_paramSwitch_e enum. |
2110 | * Default is ZSTD_ps_auto. |
2111 | * Set to ZSTD_ps_disable to never use row-based matchfinder. |
2112 | * Set to ZSTD_ps_enable to force usage of row-based matchfinder. |
2113 | * |
2114 | * By default, in ZSTD_ps_auto, the library will decide at runtime whether to use |
2115 | * the row-based matchfinder based on support for SIMD instructions and the window log. |
2116 | * Note that this only pertains to compression strategies: greedy, lazy, and lazy2 |
2117 | */ |
2118 | #define ZSTD_c_useRowMatchFinder ZSTD_c_experimentalParam14 |
2119 | |
2120 | /* ZSTD_c_deterministicRefPrefix |
2121 | * Default is 0 == disabled. Set to 1 to enable. |
2122 | * |
2123 | * Zstd produces different results for prefix compression when the prefix is |
2124 | * directly adjacent to the data about to be compressed vs. when it isn't. |
2125 | * This is because zstd detects that the two buffers are contiguous and it can |
2126 | * use a more efficient match finding algorithm. However, this produces different |
2127 | * results than when the two buffers are non-contiguous. This flag forces zstd |
2128 | * to always load the prefix in non-contiguous mode, even if it happens to be |
2129 | * adjacent to the data, to guarantee determinism. |
2130 | * |
2131 | * If you really care about determinism when using a dictionary or prefix, |
2132 | * like when doing delta compression, you should select this option. It comes |
2133 | * at a speed penalty of about ~2.5% if the dictionary and data happened to be |
2134 | * contiguous, and is free if they weren't contiguous. We don't expect that |
2135 | * intentionally making the dictionary and data contiguous will be worth the |
2136 | * cost to memcpy() the data. |
2137 | */ |
2138 | #define ZSTD_c_deterministicRefPrefix ZSTD_c_experimentalParam15 |
2139 | |
2140 | /* ZSTD_c_prefetchCDictTables |
2141 | * Controlled with ZSTD_paramSwitch_e enum. Default is ZSTD_ps_auto. |
2142 | * |
2143 | * In some situations, zstd uses CDict tables in-place rather than copying them |
2144 | * into the working context. (See docs on ZSTD_dictAttachPref_e above for details). |
2145 | * In such situations, compression speed is seriously impacted when CDict tables are |
2146 | * "cold" (outside CPU cache). This parameter instructs zstd to prefetch CDict tables |
2147 | * when they are used in-place. |
2148 | * |
2149 | * For sufficiently small inputs, the cost of the prefetch will outweigh the benefit. |
2150 | * For sufficiently large inputs, zstd will by default memcpy() CDict tables |
2151 | * into the working context, so there is no need to prefetch. This parameter is |
2152 | * targeted at a middle range of input sizes, where a prefetch is cheap enough to be |
2153 | * useful but memcpy() is too expensive. The exact range of input sizes where this |
2154 | * makes sense is best determined by careful experimentation. |
2155 | * |
2156 | * Note: for this parameter, ZSTD_ps_auto is currently equivalent to ZSTD_ps_disable, |
2157 | * but in the future zstd may conditionally enable this feature via an auto-detection |
2158 | * heuristic for cold CDicts. |
2159 | * Use ZSTD_ps_disable to opt out of prefetching under any circumstances. |
2160 | */ |
2161 | #define ZSTD_c_prefetchCDictTables ZSTD_c_experimentalParam16 |
2162 | |
2163 | /* ZSTD_c_enableSeqProducerFallback |
2164 | * Allowed values are 0 (disable) and 1 (enable). The default setting is 0. |
2165 | * |
2166 | * Controls whether zstd will fall back to an internal sequence producer if an |
2167 | * external sequence producer is registered and returns an error code. This fallback |
2168 | * is block-by-block: the internal sequence producer will only be called for blocks |
2169 | * where the external sequence producer returns an error code. Fallback parsing will |
2170 | * follow any other cParam settings, such as compression level, the same as in a |
2171 | * normal (fully-internal) compression operation. |
2172 | * |
2173 | * The user is strongly encouraged to read the full Block-Level Sequence Producer API |
2174 | * documentation (below) before setting this parameter. */ |
2175 | #define ZSTD_c_enableSeqProducerFallback ZSTD_c_experimentalParam17 |
2176 | |
2177 | /* ZSTD_c_maxBlockSize |
2178 | * Allowed values are between 1KB and ZSTD_BLOCKSIZE_MAX (128KB). |
2179 | * The default is ZSTD_BLOCKSIZE_MAX, and setting to 0 will set to the default. |
2180 | * |
2181 | * This parameter can be used to set an upper bound on the blocksize |
2182 | * that overrides the default ZSTD_BLOCKSIZE_MAX. It cannot be used to set upper |
2183 | * bounds greater than ZSTD_BLOCKSIZE_MAX or bounds lower than 1KB (will make |
2184 | * compressBound() inaccurate). Only currently meant to be used for testing. |
2185 | * |
2186 | */ |
2187 | #define ZSTD_c_maxBlockSize ZSTD_c_experimentalParam18 |
2188 | |
2189 | /* ZSTD_c_searchForExternalRepcodes |
2190 | * This parameter affects how zstd parses external sequences, such as sequences |
2191 | * provided through the compressSequences() API or from an external block-level |
2192 | * sequence producer. |
2193 | * |
2194 | * If set to ZSTD_ps_enable, the library will check for repeated offsets in |
2195 | * external sequences, even if those repcodes are not explicitly indicated in |
2196 | * the "rep" field. Note that this is the only way to exploit repcode matches |
2197 | * while using compressSequences() or an external sequence producer, since zstd |
2198 | * currently ignores the "rep" field of external sequences. |
2199 | * |
2200 | * If set to ZSTD_ps_disable, the library will not exploit repeated offsets in |
2201 | * external sequences, regardless of whether the "rep" field has been set. This |
2202 | * reduces sequence compression overhead by about 25% while sacrificing some |
2203 | * compression ratio. |
2204 | * |
2205 | * The default value is ZSTD_ps_auto, for which the library will enable/disable |
2206 | * based on compression level. |
2207 | * |
2208 | * Note: for now, this param only has an effect if ZSTD_c_blockDelimiters is |
2209 | * set to ZSTD_sf_explicitBlockDelimiters. That may change in the future. |
2210 | */ |
2211 | #define ZSTD_c_searchForExternalRepcodes ZSTD_c_experimentalParam19 |
2212 | |
2213 | /*! ZSTD_CCtx_getParameter() : |
2214 | * Get the requested compression parameter value, selected by enum ZSTD_cParameter, |
2215 | * and store it into int* value. |
2216 | * @return : 0, or an error code (which can be tested with ZSTD_isError()). |
2217 | */ |
2218 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_getParameter(const ZSTD_CCtx* cctx, ZSTD_cParameter param, int* value); |
2219 | |
2220 | |
2221 | /*! ZSTD_CCtx_params : |
2222 | * Quick howto : |
2223 | * - ZSTD_createCCtxParams() : Create a ZSTD_CCtx_params structure |
2224 | * - ZSTD_CCtxParams_setParameter() : Push parameters one by one into |
2225 | * an existing ZSTD_CCtx_params structure. |
2226 | * This is similar to |
2227 | * ZSTD_CCtx_setParameter(). |
2228 | * - ZSTD_CCtx_setParametersUsingCCtxParams() : Apply parameters to |
2229 | * an existing CCtx. |
2230 | * These parameters will be applied to |
2231 | * all subsequent frames. |
2232 | * - ZSTD_compressStream2() : Do compression using the CCtx. |
2233 | * - ZSTD_freeCCtxParams() : Free the memory, accept NULL pointer. |
2234 | * |
2235 | * This can be used with ZSTD_estimateCCtxSize_advanced_usingCCtxParams() |
2236 | * for static allocation of CCtx for single-threaded compression. |
2237 | */ |
2238 | ZSTDLIB_STATIC_API ZSTD_CCtx_params* ZSTD_createCCtxParams(void); |
2239 | ZSTDLIB_STATIC_API size_t ZSTD_freeCCtxParams(ZSTD_CCtx_params* params); /* accept NULL pointer */ |
2240 | |
2241 | /*! ZSTD_CCtxParams_reset() : |
2242 | * Reset params to default values. |
2243 | */ |
2244 | ZSTDLIB_STATIC_API size_t ZSTD_CCtxParams_reset(ZSTD_CCtx_params* params); |
2245 | |
2246 | /*! ZSTD_CCtxParams_init() : |
2247 | * Initializes the compression parameters of cctxParams according to |
2248 | * compression level. All other parameters are reset to their default values. |
2249 | */ |
2250 | ZSTDLIB_STATIC_API size_t ZSTD_CCtxParams_init(ZSTD_CCtx_params* cctxParams, int compressionLevel); |
2251 | |
2252 | /*! ZSTD_CCtxParams_init_advanced() : |
2253 | * Initializes the compression and frame parameters of cctxParams according to |
2254 | * params. All other parameters are reset to their default values. |
2255 | */ |
2256 | ZSTDLIB_STATIC_API size_t ZSTD_CCtxParams_init_advanced(ZSTD_CCtx_params* cctxParams, ZSTD_parameters params); |
2257 | |
2258 | /*! ZSTD_CCtxParams_setParameter() : Requires v1.4.0+ |
2259 | * Similar to ZSTD_CCtx_setParameter. |
2260 | * Set one compression parameter, selected by enum ZSTD_cParameter. |
2261 | * Parameters must be applied to a ZSTD_CCtx using |
2262 | * ZSTD_CCtx_setParametersUsingCCtxParams(). |
2263 | * @result : a code representing success or failure (which can be tested with |
2264 | * ZSTD_isError()). |
2265 | */ |
2266 | ZSTDLIB_STATIC_API size_t ZSTD_CCtxParams_setParameter(ZSTD_CCtx_params* params, ZSTD_cParameter param, int value); |
2267 | |
2268 | /*! ZSTD_CCtxParams_getParameter() : |
2269 | * Similar to ZSTD_CCtx_getParameter. |
2270 | * Get the requested value of one compression parameter, selected by enum ZSTD_cParameter. |
2271 | * @result : 0, or an error code (which can be tested with ZSTD_isError()). |
2272 | */ |
2273 | ZSTDLIB_STATIC_API size_t ZSTD_CCtxParams_getParameter(const ZSTD_CCtx_params* params, ZSTD_cParameter param, int* value); |
2274 | |
2275 | /*! ZSTD_CCtx_setParametersUsingCCtxParams() : |
2276 | * Apply a set of ZSTD_CCtx_params to the compression context. |
2277 | * This can be done even after compression is started, |
2278 | * if nbWorkers==0, this will have no impact until a new compression is started. |
2279 | * if nbWorkers>=1, new parameters will be picked up at next job, |
2280 | * with a few restrictions (windowLog, pledgedSrcSize, nbWorkers, jobSize, and overlapLog are not updated). |
2281 | */ |
2282 | ZSTDLIB_STATIC_API size_t ZSTD_CCtx_setParametersUsingCCtxParams( |
2283 | ZSTD_CCtx* cctx, const ZSTD_CCtx_params* params); |
2284 | |
2285 | /*! ZSTD_compressStream2_simpleArgs() : |
2286 | * Same as ZSTD_compressStream2(), |
2287 | * but using only integral types as arguments. |
2288 | * This variant might be helpful for binders from dynamic languages |
2289 | * which have troubles handling structures containing memory pointers. |
2290 | */ |
2291 | ZSTDLIB_STATIC_API size_t ZSTD_compressStream2_simpleArgs ( |
2292 | ZSTD_CCtx* cctx, |
2293 | void* dst, size_t dstCapacity, size_t* dstPos, |
2294 | const void* src, size_t srcSize, size_t* srcPos, |
2295 | ZSTD_EndDirective endOp); |
2296 | |
2297 | |
2298 | /*************************************** |
2299 | * Advanced decompression functions |
2300 | ***************************************/ |
2301 | |
2302 | /*! ZSTD_isFrame() : |
2303 | * Tells if the content of `buffer` starts with a valid Frame Identifier. |
2304 | * Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0. |
2305 | * Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled. |
2306 | * Note 3 : Skippable Frame Identifiers are considered valid. */ |
2307 | ZSTDLIB_STATIC_API unsigned ZSTD_isFrame(const void* buffer, size_t size); |
2308 | |
2309 | /*! ZSTD_createDDict_byReference() : |
2310 | * Create a digested dictionary, ready to start decompression operation without startup delay. |
2311 | * Dictionary content is referenced, and therefore stays in dictBuffer. |
2312 | * It is important that dictBuffer outlives DDict, |
2313 | * it must remain read accessible throughout the lifetime of DDict */ |
2314 | ZSTDLIB_STATIC_API ZSTD_DDict* ZSTD_createDDict_byReference(const void* dictBuffer, size_t dictSize); |
2315 | |
2316 | /*! ZSTD_DCtx_loadDictionary_byReference() : |
2317 | * Same as ZSTD_DCtx_loadDictionary(), |
2318 | * but references `dict` content instead of copying it into `dctx`. |
2319 | * This saves memory if `dict` remains around., |
2320 | * However, it's imperative that `dict` remains accessible (and unmodified) while being used, so it must outlive decompression. */ |
2321 | ZSTDLIB_STATIC_API size_t ZSTD_DCtx_loadDictionary_byReference(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); |
2322 | |
2323 | /*! ZSTD_DCtx_loadDictionary_advanced() : |
2324 | * Same as ZSTD_DCtx_loadDictionary(), |
2325 | * but gives direct control over |
2326 | * how to load the dictionary (by copy ? by reference ?) |
2327 | * and how to interpret it (automatic ? force raw mode ? full mode only ?). */ |
2328 | ZSTDLIB_STATIC_API size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx, const void* dict, size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType); |
2329 | |
2330 | /*! ZSTD_DCtx_refPrefix_advanced() : |
2331 | * Same as ZSTD_DCtx_refPrefix(), but gives finer control over |
2332 | * how to interpret prefix content (automatic ? force raw mode (default) ? full mode only ?) */ |
2333 | ZSTDLIB_STATIC_API size_t ZSTD_DCtx_refPrefix_advanced(ZSTD_DCtx* dctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType); |
2334 | |
2335 | /*! ZSTD_DCtx_setMaxWindowSize() : |
2336 | * Refuses allocating internal buffers for frames requiring a window size larger than provided limit. |
2337 | * This protects a decoder context from reserving too much memory for itself (potential attack scenario). |
2338 | * This parameter is only useful in streaming mode, since no internal buffer is allocated in single-pass mode. |
2339 | * By default, a decompression context accepts all window sizes <= (1 << ZSTD_WINDOWLOG_LIMIT_DEFAULT) |
2340 | * @return : 0, or an error code (which can be tested using ZSTD_isError()). |
2341 | */ |
2342 | ZSTDLIB_STATIC_API size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize); |
2343 | |
2344 | /*! ZSTD_DCtx_getParameter() : |
2345 | * Get the requested decompression parameter value, selected by enum ZSTD_dParameter, |
2346 | * and store it into int* value. |
2347 | * @return : 0, or an error code (which can be tested with ZSTD_isError()). |
2348 | */ |
2349 | ZSTDLIB_STATIC_API size_t ZSTD_DCtx_getParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int* value); |
2350 | |
2351 | /* ZSTD_d_format |
2352 | * experimental parameter, |
2353 | * allowing selection between ZSTD_format_e input compression formats |
2354 | */ |
2355 | #define ZSTD_d_format ZSTD_d_experimentalParam1 |
2356 | /* ZSTD_d_stableOutBuffer |
2357 | * Experimental parameter. |
2358 | * Default is 0 == disabled. Set to 1 to enable. |
2359 | * |
2360 | * Tells the decompressor that the ZSTD_outBuffer will ALWAYS be the same |
2361 | * between calls, except for the modifications that zstd makes to pos (the |
2362 | * caller must not modify pos). This is checked by the decompressor, and |
2363 | * decompression will fail if it ever changes. Therefore the ZSTD_outBuffer |
2364 | * MUST be large enough to fit the entire decompressed frame. This will be |
2365 | * checked when the frame content size is known. The data in the ZSTD_outBuffer |
2366 | * in the range [dst, dst + pos) MUST not be modified during decompression |
2367 | * or you will get data corruption. |
2368 | * |
2369 | * When this flag is enabled zstd won't allocate an output buffer, because |
2370 | * it can write directly to the ZSTD_outBuffer, but it will still allocate |
2371 | * an input buffer large enough to fit any compressed block. This will also |
2372 | * avoid the memcpy() from the internal output buffer to the ZSTD_outBuffer. |
2373 | * If you need to avoid the input buffer allocation use the buffer-less |
2374 | * streaming API. |
2375 | * |
2376 | * NOTE: So long as the ZSTD_outBuffer always points to valid memory, using |
2377 | * this flag is ALWAYS memory safe, and will never access out-of-bounds |
2378 | * memory. However, decompression WILL fail if you violate the preconditions. |
2379 | * |
2380 | * WARNING: The data in the ZSTD_outBuffer in the range [dst, dst + pos) MUST |
2381 | * not be modified during decompression or you will get data corruption. This |
2382 | * is because zstd needs to reference data in the ZSTD_outBuffer to regenerate |
2383 | * matches. Normally zstd maintains its own buffer for this purpose, but passing |
2384 | * this flag tells zstd to use the user provided buffer. |
2385 | */ |
2386 | #define ZSTD_d_stableOutBuffer ZSTD_d_experimentalParam2 |
2387 | |
2388 | /* ZSTD_d_forceIgnoreChecksum |
2389 | * Experimental parameter. |
2390 | * Default is 0 == disabled. Set to 1 to enable |
2391 | * |
2392 | * Tells the decompressor to skip checksum validation during decompression, regardless |
2393 | * of whether checksumming was specified during compression. This offers some |
2394 | * slight performance benefits, and may be useful for debugging. |
2395 | * Param has values of type ZSTD_forceIgnoreChecksum_e |
2396 | */ |
2397 | #define ZSTD_d_forceIgnoreChecksum ZSTD_d_experimentalParam3 |
2398 | |
2399 | /* ZSTD_d_refMultipleDDicts |
2400 | * Experimental parameter. |
2401 | * Default is 0 == disabled. Set to 1 to enable |
2402 | * |
2403 | * If enabled and dctx is allocated on the heap, then additional memory will be allocated |
2404 | * to store references to multiple ZSTD_DDict. That is, multiple calls of ZSTD_refDDict() |
2405 | * using a given ZSTD_DCtx, rather than overwriting the previous DDict reference, will instead |
2406 | * store all references. At decompression time, the appropriate dictID is selected |
2407 | * from the set of DDicts based on the dictID in the frame. |
2408 | * |
2409 | * Usage is simply calling ZSTD_refDDict() on multiple dict buffers. |
2410 | * |
2411 | * Param has values of byte ZSTD_refMultipleDDicts_e |
2412 | * |
2413 | * WARNING: Enabling this parameter and calling ZSTD_DCtx_refDDict(), will trigger memory |
2414 | * allocation for the hash table. ZSTD_freeDCtx() also frees this memory. |
2415 | * Memory is allocated as per ZSTD_DCtx::customMem. |
2416 | * |
2417 | * Although this function allocates memory for the table, the user is still responsible for |
2418 | * memory management of the underlying ZSTD_DDict* themselves. |
2419 | */ |
2420 | #define ZSTD_d_refMultipleDDicts ZSTD_d_experimentalParam4 |
2421 | |
2422 | /* ZSTD_d_disableHuffmanAssembly |
2423 | * Set to 1 to disable the Huffman assembly implementation. |
2424 | * The default value is 0, which allows zstd to use the Huffman assembly |
2425 | * implementation if available. |
2426 | * |
2427 | * This parameter can be used to disable Huffman assembly at runtime. |
2428 | * If you want to disable it at compile time you can define the macro |
2429 | * ZSTD_DISABLE_ASM. |
2430 | */ |
2431 | #define ZSTD_d_disableHuffmanAssembly ZSTD_d_experimentalParam5 |
2432 | |
2433 | |
2434 | /*! ZSTD_DCtx_setFormat() : |
2435 | * This function is REDUNDANT. Prefer ZSTD_DCtx_setParameter(). |
2436 | * Instruct the decoder context about what kind of data to decode next. |
2437 | * This instruction is mandatory to decode data without a fully-formed header, |
2438 | * such ZSTD_f_zstd1_magicless for example. |
2439 | * @return : 0, or an error code (which can be tested using ZSTD_isError()). */ |
2440 | ZSTD_DEPRECATED("use ZSTD_DCtx_setParameter() instead" ) |
2441 | ZSTDLIB_STATIC_API |
2442 | size_t ZSTD_DCtx_setFormat(ZSTD_DCtx* dctx, ZSTD_format_e format); |
2443 | |
2444 | /*! ZSTD_decompressStream_simpleArgs() : |
2445 | * Same as ZSTD_decompressStream(), |
2446 | * but using only integral types as arguments. |
2447 | * This can be helpful for binders from dynamic languages |
2448 | * which have troubles handling structures containing memory pointers. |
2449 | */ |
2450 | ZSTDLIB_STATIC_API size_t ZSTD_decompressStream_simpleArgs ( |
2451 | ZSTD_DCtx* dctx, |
2452 | void* dst, size_t dstCapacity, size_t* dstPos, |
2453 | const void* src, size_t srcSize, size_t* srcPos); |
2454 | |
2455 | |
2456 | /******************************************************************** |
2457 | * Advanced streaming functions |
2458 | * Warning : most of these functions are now redundant with the Advanced API. |
2459 | * Once Advanced API reaches "stable" status, |
2460 | * redundant functions will be deprecated, and then at some point removed. |
2461 | ********************************************************************/ |
2462 | |
2463 | /*===== Advanced Streaming compression functions =====*/ |
2464 | |
2465 | /*! ZSTD_initCStream_srcSize() : |
2466 | * This function is DEPRECATED, and equivalent to: |
2467 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
2468 | * ZSTD_CCtx_refCDict(zcs, NULL); // clear the dictionary (if any) |
2469 | * ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel); |
2470 | * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); |
2471 | * |
2472 | * pledgedSrcSize must be correct. If it is not known at init time, use |
2473 | * ZSTD_CONTENTSIZE_UNKNOWN. Note that, for compatibility with older programs, |
2474 | * "0" also disables frame content size field. It may be enabled in the future. |
2475 | * This prototype will generate compilation warnings. |
2476 | */ |
2477 | ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions" ) |
2478 | ZSTDLIB_STATIC_API |
2479 | size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs, |
2480 | int compressionLevel, |
2481 | unsigned long long pledgedSrcSize); |
2482 | |
2483 | /*! ZSTD_initCStream_usingDict() : |
2484 | * This function is DEPRECATED, and is equivalent to: |
2485 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
2486 | * ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel); |
2487 | * ZSTD_CCtx_loadDictionary(zcs, dict, dictSize); |
2488 | * |
2489 | * Creates of an internal CDict (incompatible with static CCtx), except if |
2490 | * dict == NULL or dictSize < 8, in which case no dict is used. |
2491 | * Note: dict is loaded with ZSTD_dct_auto (treated as a full zstd dictionary if |
2492 | * it begins with ZSTD_MAGIC_DICTIONARY, else as raw content) and ZSTD_dlm_byCopy. |
2493 | * This prototype will generate compilation warnings. |
2494 | */ |
2495 | ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions" ) |
2496 | ZSTDLIB_STATIC_API |
2497 | size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs, |
2498 | const void* dict, size_t dictSize, |
2499 | int compressionLevel); |
2500 | |
2501 | /*! ZSTD_initCStream_advanced() : |
2502 | * This function is DEPRECATED, and is equivalent to: |
2503 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
2504 | * ZSTD_CCtx_setParams(zcs, params); |
2505 | * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); |
2506 | * ZSTD_CCtx_loadDictionary(zcs, dict, dictSize); |
2507 | * |
2508 | * dict is loaded with ZSTD_dct_auto and ZSTD_dlm_byCopy. |
2509 | * pledgedSrcSize must be correct. |
2510 | * If srcSize is not known at init time, use value ZSTD_CONTENTSIZE_UNKNOWN. |
2511 | * This prototype will generate compilation warnings. |
2512 | */ |
2513 | ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions" ) |
2514 | ZSTDLIB_STATIC_API |
2515 | size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs, |
2516 | const void* dict, size_t dictSize, |
2517 | ZSTD_parameters params, |
2518 | unsigned long long pledgedSrcSize); |
2519 | |
2520 | /*! ZSTD_initCStream_usingCDict() : |
2521 | * This function is DEPRECATED, and equivalent to: |
2522 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
2523 | * ZSTD_CCtx_refCDict(zcs, cdict); |
2524 | * |
2525 | * note : cdict will just be referenced, and must outlive compression session |
2526 | * This prototype will generate compilation warnings. |
2527 | */ |
2528 | ZSTD_DEPRECATED("use ZSTD_CCtx_reset and ZSTD_CCtx_refCDict, see zstd.h for detailed instructions" ) |
2529 | ZSTDLIB_STATIC_API |
2530 | size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict); |
2531 | |
2532 | /*! ZSTD_initCStream_usingCDict_advanced() : |
2533 | * This function is DEPRECATED, and is equivalent to: |
2534 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
2535 | * ZSTD_CCtx_setFParams(zcs, fParams); |
2536 | * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); |
2537 | * ZSTD_CCtx_refCDict(zcs, cdict); |
2538 | * |
2539 | * same as ZSTD_initCStream_usingCDict(), with control over frame parameters. |
2540 | * pledgedSrcSize must be correct. If srcSize is not known at init time, use |
2541 | * value ZSTD_CONTENTSIZE_UNKNOWN. |
2542 | * This prototype will generate compilation warnings. |
2543 | */ |
2544 | ZSTD_DEPRECATED("use ZSTD_CCtx_reset and ZSTD_CCtx_refCDict, see zstd.h for detailed instructions" ) |
2545 | ZSTDLIB_STATIC_API |
2546 | size_t ZSTD_initCStream_usingCDict_advanced(ZSTD_CStream* zcs, |
2547 | const ZSTD_CDict* cdict, |
2548 | ZSTD_frameParameters fParams, |
2549 | unsigned long long pledgedSrcSize); |
2550 | |
2551 | /*! ZSTD_resetCStream() : |
2552 | * This function is DEPRECATED, and is equivalent to: |
2553 | * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); |
2554 | * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); |
2555 | * Note: ZSTD_resetCStream() interprets pledgedSrcSize == 0 as ZSTD_CONTENTSIZE_UNKNOWN, but |
2556 | * ZSTD_CCtx_setPledgedSrcSize() does not do the same, so ZSTD_CONTENTSIZE_UNKNOWN must be |
2557 | * explicitly specified. |
2558 | * |
2559 | * start a new frame, using same parameters from previous frame. |
2560 | * This is typically useful to skip dictionary loading stage, since it will re-use it in-place. |
2561 | * Note that zcs must be init at least once before using ZSTD_resetCStream(). |
2562 | * If pledgedSrcSize is not known at reset time, use macro ZSTD_CONTENTSIZE_UNKNOWN. |
2563 | * If pledgedSrcSize > 0, its value must be correct, as it will be written in header, and controlled at the end. |
2564 | * For the time being, pledgedSrcSize==0 is interpreted as "srcSize unknown" for compatibility with older programs, |
2565 | * but it will change to mean "empty" in future version, so use macro ZSTD_CONTENTSIZE_UNKNOWN instead. |
2566 | * @return : 0, or an error code (which can be tested using ZSTD_isError()) |
2567 | * This prototype will generate compilation warnings. |
2568 | */ |
2569 | ZSTD_DEPRECATED("use ZSTD_CCtx_reset, see zstd.h for detailed instructions" ) |
2570 | ZSTDLIB_STATIC_API |
2571 | size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize); |
2572 | |
2573 | |
2574 | typedef struct { |
2575 | unsigned long long ingested; /* nb input bytes read and buffered */ |
2576 | unsigned long long consumed; /* nb input bytes actually compressed */ |
2577 | unsigned long long produced; /* nb of compressed bytes generated and buffered */ |
2578 | unsigned long long flushed; /* nb of compressed bytes flushed : not provided; can be tracked from caller side */ |
2579 | unsigned currentJobID; /* MT only : latest started job nb */ |
2580 | unsigned nbActiveWorkers; /* MT only : nb of workers actively compressing at probe time */ |
2581 | } ZSTD_frameProgression; |
2582 | |
2583 | /* ZSTD_getFrameProgression() : |
2584 | * tells how much data has been ingested (read from input) |
2585 | * consumed (input actually compressed) and produced (output) for current frame. |
2586 | * Note : (ingested - consumed) is amount of input data buffered internally, not yet compressed. |
2587 | * Aggregates progression inside active worker threads. |
2588 | */ |
2589 | ZSTDLIB_STATIC_API ZSTD_frameProgression ZSTD_getFrameProgression(const ZSTD_CCtx* cctx); |
2590 | |
2591 | /*! ZSTD_toFlushNow() : |
2592 | * Tell how many bytes are ready to be flushed immediately. |
2593 | * Useful for multithreading scenarios (nbWorkers >= 1). |
2594 | * Probe the oldest active job, defined as oldest job not yet entirely flushed, |
2595 | * and check its output buffer. |
2596 | * @return : amount of data stored in oldest job and ready to be flushed immediately. |
2597 | * if @return == 0, it means either : |
2598 | * + there is no active job (could be checked with ZSTD_frameProgression()), or |
2599 | * + oldest job is still actively compressing data, |
2600 | * but everything it has produced has also been flushed so far, |
2601 | * therefore flush speed is limited by production speed of oldest job |
2602 | * irrespective of the speed of concurrent (and newer) jobs. |
2603 | */ |
2604 | ZSTDLIB_STATIC_API size_t ZSTD_toFlushNow(ZSTD_CCtx* cctx); |
2605 | |
2606 | |
2607 | /*===== Advanced Streaming decompression functions =====*/ |
2608 | |
2609 | /*! |
2610 | * This function is deprecated, and is equivalent to: |
2611 | * |
2612 | * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); |
2613 | * ZSTD_DCtx_loadDictionary(zds, dict, dictSize); |
2614 | * |
2615 | * note: no dictionary will be used if dict == NULL or dictSize < 8 |
2616 | */ |
2617 | ZSTD_DEPRECATED("use ZSTD_DCtx_reset + ZSTD_DCtx_loadDictionary, see zstd.h for detailed instructions" ) |
2618 | ZSTDLIB_STATIC_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize); |
2619 | |
2620 | /*! |
2621 | * This function is deprecated, and is equivalent to: |
2622 | * |
2623 | * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); |
2624 | * ZSTD_DCtx_refDDict(zds, ddict); |
2625 | * |
2626 | * note : ddict is referenced, it must outlive decompression session |
2627 | */ |
2628 | ZSTD_DEPRECATED("use ZSTD_DCtx_reset + ZSTD_DCtx_refDDict, see zstd.h for detailed instructions" ) |
2629 | ZSTDLIB_STATIC_API size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict); |
2630 | |
2631 | /*! |
2632 | * This function is deprecated, and is equivalent to: |
2633 | * |
2634 | * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); |
2635 | * |
2636 | * re-use decompression parameters from previous init; saves dictionary loading |
2637 | */ |
2638 | ZSTD_DEPRECATED("use ZSTD_DCtx_reset, see zstd.h for detailed instructions" ) |
2639 | ZSTDLIB_STATIC_API size_t ZSTD_resetDStream(ZSTD_DStream* zds); |
2640 | |
2641 | |
2642 | /* ********************* BLOCK-LEVEL SEQUENCE PRODUCER API ********************* |
2643 | * |
2644 | * *** OVERVIEW *** |
2645 | * The Block-Level Sequence Producer API allows users to provide their own custom |
2646 | * sequence producer which libzstd invokes to process each block. The produced list |
2647 | * of sequences (literals and matches) is then post-processed by libzstd to produce |
2648 | * valid compressed blocks. |
2649 | * |
2650 | * This block-level offload API is a more granular complement of the existing |
2651 | * frame-level offload API compressSequences() (introduced in v1.5.1). It offers |
2652 | * an easier migration story for applications already integrated with libzstd: the |
2653 | * user application continues to invoke the same compression functions |
2654 | * ZSTD_compress2() or ZSTD_compressStream2() as usual, and transparently benefits |
2655 | * from the specific advantages of the external sequence producer. For example, |
2656 | * the sequence producer could be tuned to take advantage of known characteristics |
2657 | * of the input, to offer better speed / ratio, or could leverage hardware |
2658 | * acceleration not available within libzstd itself. |
2659 | * |
2660 | * See contrib/externalSequenceProducer for an example program employing the |
2661 | * Block-Level Sequence Producer API. |
2662 | * |
2663 | * *** USAGE *** |
2664 | * The user is responsible for implementing a function of type |
2665 | * ZSTD_sequenceProducer_F. For each block, zstd will pass the following |
2666 | * arguments to the user-provided function: |
2667 | * |
2668 | * - sequenceProducerState: a pointer to a user-managed state for the sequence |
2669 | * producer. |
2670 | * |
2671 | * - outSeqs, outSeqsCapacity: an output buffer for the sequence producer. |
2672 | * outSeqsCapacity is guaranteed >= ZSTD_sequenceBound(srcSize). The memory |
2673 | * backing outSeqs is managed by the CCtx. |
2674 | * |
2675 | * - src, srcSize: an input buffer for the sequence producer to parse. |
2676 | * srcSize is guaranteed to be <= ZSTD_BLOCKSIZE_MAX. |
2677 | * |
2678 | * - dict, dictSize: a history buffer, which may be empty, which the sequence |
2679 | * producer may reference as it parses the src buffer. Currently, zstd will |
2680 | * always pass dictSize == 0 into external sequence producers, but this will |
2681 | * change in the future. |
2682 | * |
2683 | * - compressionLevel: a signed integer representing the zstd compression level |
2684 | * set by the user for the current operation. The sequence producer may choose |
2685 | * to use this information to change its compression strategy and speed/ratio |
2686 | * tradeoff. Note: the compression level does not reflect zstd parameters set |
2687 | * through the advanced API. |
2688 | * |
2689 | * - windowSize: a size_t representing the maximum allowed offset for external |
2690 | * sequences. Note that sequence offsets are sometimes allowed to exceed the |
2691 | * windowSize if a dictionary is present, see doc/zstd_compression_format.md |
2692 | * for details. |
2693 | * |
2694 | * The user-provided function shall return a size_t representing the number of |
2695 | * sequences written to outSeqs. This return value will be treated as an error |
2696 | * code if it is greater than outSeqsCapacity. The return value must be non-zero |
2697 | * if srcSize is non-zero. The ZSTD_SEQUENCE_PRODUCER_ERROR macro is provided |
2698 | * for convenience, but any value greater than outSeqsCapacity will be treated as |
2699 | * an error code. |
2700 | * |
2701 | * If the user-provided function does not return an error code, the sequences |
2702 | * written to outSeqs must be a valid parse of the src buffer. Data corruption may |
2703 | * occur if the parse is not valid. A parse is defined to be valid if the |
2704 | * following conditions hold: |
2705 | * - The sum of matchLengths and literalLengths must equal srcSize. |
2706 | * - All sequences in the parse, except for the final sequence, must have |
2707 | * matchLength >= ZSTD_MINMATCH_MIN. The final sequence must have |
2708 | * matchLength >= ZSTD_MINMATCH_MIN or matchLength == 0. |
2709 | * - All offsets must respect the windowSize parameter as specified in |
2710 | * doc/zstd_compression_format.md. |
2711 | * - If the final sequence has matchLength == 0, it must also have offset == 0. |
2712 | * |
2713 | * zstd will only validate these conditions (and fail compression if they do not |
2714 | * hold) if the ZSTD_c_validateSequences cParam is enabled. Note that sequence |
2715 | * validation has a performance cost. |
2716 | * |
2717 | * If the user-provided function returns an error, zstd will either fall back |
2718 | * to an internal sequence producer or fail the compression operation. The user can |
2719 | * choose between the two behaviors by setting the ZSTD_c_enableSeqProducerFallback |
2720 | * cParam. Fallback compression will follow any other cParam settings, such as |
2721 | * compression level, the same as in a normal compression operation. |
2722 | * |
2723 | * The user shall instruct zstd to use a particular ZSTD_sequenceProducer_F |
2724 | * function by calling |
2725 | * ZSTD_registerSequenceProducer(cctx, |
2726 | * sequenceProducerState, |
2727 | * sequenceProducer) |
2728 | * This setting will persist until the next parameter reset of the CCtx. |
2729 | * |
2730 | * The sequenceProducerState must be initialized by the user before calling |
2731 | * ZSTD_registerSequenceProducer(). The user is responsible for destroying the |
2732 | * sequenceProducerState. |
2733 | * |
2734 | * *** LIMITATIONS *** |
2735 | * This API is compatible with all zstd compression APIs which respect advanced parameters. |
2736 | * However, there are three limitations: |
2737 | * |
2738 | * First, the ZSTD_c_enableLongDistanceMatching cParam is not currently supported. |
2739 | * COMPRESSION WILL FAIL if it is enabled and the user tries to compress with a block-level |
2740 | * external sequence producer. |
2741 | * - Note that ZSTD_c_enableLongDistanceMatching is auto-enabled by default in some |
2742 | * cases (see its documentation for details). Users must explicitly set |
2743 | * ZSTD_c_enableLongDistanceMatching to ZSTD_ps_disable in such cases if an external |
2744 | * sequence producer is registered. |
2745 | * - As of this writing, ZSTD_c_enableLongDistanceMatching is disabled by default |
2746 | * whenever ZSTD_c_windowLog < 128MB, but that's subject to change. Users should |
2747 | * check the docs on ZSTD_c_enableLongDistanceMatching whenever the Block-Level Sequence |
2748 | * Producer API is used in conjunction with advanced settings (like ZSTD_c_windowLog). |
2749 | * |
2750 | * Second, history buffers are not currently supported. Concretely, zstd will always pass |
2751 | * dictSize == 0 to the external sequence producer (for now). This has two implications: |
2752 | * - Dictionaries are not currently supported. Compression will *not* fail if the user |
2753 | * references a dictionary, but the dictionary won't have any effect. |
2754 | * - Stream history is not currently supported. All advanced compression APIs, including |
2755 | * streaming APIs, work with external sequence producers, but each block is treated as |
2756 | * an independent chunk without history from previous blocks. |
2757 | * |
2758 | * Third, multi-threading within a single compression is not currently supported. In other words, |
2759 | * COMPRESSION WILL FAIL if ZSTD_c_nbWorkers > 0 and an external sequence producer is registered. |
2760 | * Multi-threading across compressions is fine: simply create one CCtx per thread. |
2761 | * |
2762 | * Long-term, we plan to overcome all three limitations. There is no technical blocker to |
2763 | * overcoming them. It is purely a question of engineering effort. |
2764 | */ |
2765 | |
2766 | #define ZSTD_SEQUENCE_PRODUCER_ERROR ((size_t)(-1)) |
2767 | |
2768 | typedef size_t ZSTD_sequenceProducer_F ( |
2769 | void* sequenceProducerState, |
2770 | ZSTD_Sequence* outSeqs, size_t outSeqsCapacity, |
2771 | const void* src, size_t srcSize, |
2772 | const void* dict, size_t dictSize, |
2773 | int compressionLevel, |
2774 | size_t windowSize |
2775 | ); |
2776 | |
2777 | /*! ZSTD_registerSequenceProducer() : |
2778 | * Instruct zstd to use a block-level external sequence producer function. |
2779 | * |
2780 | * The sequenceProducerState must be initialized by the caller, and the caller is |
2781 | * responsible for managing its lifetime. This parameter is sticky across |
2782 | * compressions. It will remain set until the user explicitly resets compression |
2783 | * parameters. |
2784 | * |
2785 | * Sequence producer registration is considered to be an "advanced parameter", |
2786 | * part of the "advanced API". This means it will only have an effect on compression |
2787 | * APIs which respect advanced parameters, such as compress2() and compressStream2(). |
2788 | * Older compression APIs such as compressCCtx(), which predate the introduction of |
2789 | * "advanced parameters", will ignore any external sequence producer setting. |
2790 | * |
2791 | * The sequence producer can be "cleared" by registering a NULL function pointer. This |
2792 | * removes all limitations described above in the "LIMITATIONS" section of the API docs. |
2793 | * |
2794 | * The user is strongly encouraged to read the full API documentation (above) before |
2795 | * calling this function. */ |
2796 | ZSTDLIB_STATIC_API void |
2797 | ZSTD_registerSequenceProducer( |
2798 | ZSTD_CCtx* cctx, |
2799 | void* sequenceProducerState, |
2800 | ZSTD_sequenceProducer_F* sequenceProducer |
2801 | ); |
2802 | |
2803 | |
2804 | /********************************************************************* |
2805 | * Buffer-less and synchronous inner streaming functions (DEPRECATED) |
2806 | * |
2807 | * This API is deprecated, and will be removed in a future version. |
2808 | * It allows streaming (de)compression with user allocated buffers. |
2809 | * However, it is hard to use, and not as well tested as the rest of |
2810 | * our API. |
2811 | * |
2812 | * Please use the normal streaming API instead: ZSTD_compressStream2, |
2813 | * and ZSTD_decompressStream. |
2814 | * If there is functionality that you need, but it doesn't provide, |
2815 | * please open an issue on our GitHub. |
2816 | ********************************************************************* */ |
2817 | |
2818 | /** |
2819 | Buffer-less streaming compression (synchronous mode) |
2820 | |
2821 | A ZSTD_CCtx object is required to track streaming operations. |
2822 | Use ZSTD_createCCtx() / ZSTD_freeCCtx() to manage resource. |
2823 | ZSTD_CCtx object can be re-used multiple times within successive compression operations. |
2824 | |
2825 | Start by initializing a context. |
2826 | Use ZSTD_compressBegin(), or ZSTD_compressBegin_usingDict() for dictionary compression. |
2827 | |
2828 | Then, consume your input using ZSTD_compressContinue(). |
2829 | There are some important considerations to keep in mind when using this advanced function : |
2830 | - ZSTD_compressContinue() has no internal buffer. It uses externally provided buffers only. |
2831 | - Interface is synchronous : input is consumed entirely and produces 1+ compressed blocks. |
2832 | - Caller must ensure there is enough space in `dst` to store compressed data under worst case scenario. |
2833 | Worst case evaluation is provided by ZSTD_compressBound(). |
2834 | ZSTD_compressContinue() doesn't guarantee recover after a failed compression. |
2835 | - ZSTD_compressContinue() presumes prior input ***is still accessible and unmodified*** (up to maximum distance size, see WindowLog). |
2836 | It remembers all previous contiguous blocks, plus one separated memory segment (which can itself consists of multiple contiguous blocks) |
2837 | - ZSTD_compressContinue() detects that prior input has been overwritten when `src` buffer overlaps. |
2838 | In which case, it will "discard" the relevant memory section from its history. |
2839 | |
2840 | Finish a frame with ZSTD_compressEnd(), which will write the last block(s) and optional checksum. |
2841 | It's possible to use srcSize==0, in which case, it will write a final empty block to end the frame. |
2842 | Without last block mark, frames are considered unfinished (hence corrupted) by compliant decoders. |
2843 | |
2844 | `ZSTD_CCtx` object can be re-used (ZSTD_compressBegin()) to compress again. |
2845 | */ |
2846 | |
2847 | /*===== Buffer-less streaming compression functions =====*/ |
2848 | ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs." ) |
2849 | ZSTDLIB_STATIC_API size_t ZSTD_compressBegin(ZSTD_CCtx* cctx, int compressionLevel); |
2850 | ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs." ) |
2851 | ZSTDLIB_STATIC_API size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, int compressionLevel); |
2852 | ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs." ) |
2853 | ZSTDLIB_STATIC_API size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict); /**< note: fails if cdict==NULL */ |
2854 | |
2855 | ZSTD_DEPRECATED("This function will likely be removed in a future release. It is misleading and has very limited utility." ) |
2856 | ZSTDLIB_STATIC_API |
2857 | size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* preparedCCtx, unsigned long long pledgedSrcSize); /**< note: if pledgedSrcSize is not known, use ZSTD_CONTENTSIZE_UNKNOWN */ |
2858 | |
2859 | ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs." ) |
2860 | ZSTDLIB_STATIC_API size_t ZSTD_compressContinue(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); |
2861 | ZSTD_DEPRECATED("The buffer-less API is deprecated in favor of the normal streaming API. See docs." ) |
2862 | ZSTDLIB_STATIC_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); |
2863 | |
2864 | /* The ZSTD_compressBegin_advanced() and ZSTD_compressBegin_usingCDict_advanced() are now DEPRECATED and will generate a compiler warning */ |
2865 | ZSTD_DEPRECATED("use advanced API to access custom parameters" ) |
2866 | ZSTDLIB_STATIC_API |
2867 | size_t ZSTD_compressBegin_advanced(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, ZSTD_parameters params, unsigned long long pledgedSrcSize); /**< pledgedSrcSize : If srcSize is not known at init time, use ZSTD_CONTENTSIZE_UNKNOWN */ |
2868 | ZSTD_DEPRECATED("use advanced API to access custom parameters" ) |
2869 | ZSTDLIB_STATIC_API |
2870 | size_t ZSTD_compressBegin_usingCDict_advanced(ZSTD_CCtx* const cctx, const ZSTD_CDict* const cdict, ZSTD_frameParameters const fParams, unsigned long long const pledgedSrcSize); /* compression parameters are already set within cdict. pledgedSrcSize must be correct. If srcSize is not known, use macro ZSTD_CONTENTSIZE_UNKNOWN */ |
2871 | /** |
2872 | Buffer-less streaming decompression (synchronous mode) |
2873 | |
2874 | A ZSTD_DCtx object is required to track streaming operations. |
2875 | Use ZSTD_createDCtx() / ZSTD_freeDCtx() to manage it. |
2876 | A ZSTD_DCtx object can be re-used multiple times. |
2877 | |
2878 | First typical operation is to retrieve frame parameters, using ZSTD_getFrameHeader(). |
2879 | Frame header is extracted from the beginning of compressed frame, so providing only the frame's beginning is enough. |
2880 | Data fragment must be large enough to ensure successful decoding. |
2881 | `ZSTD_frameHeaderSize_max` bytes is guaranteed to always be large enough. |
2882 | result : 0 : successful decoding, the `ZSTD_frameHeader` structure is correctly filled. |
2883 | >0 : `srcSize` is too small, please provide at least result bytes on next attempt. |
2884 | errorCode, which can be tested using ZSTD_isError(). |
2885 | |
2886 | It fills a ZSTD_frameHeader structure with important information to correctly decode the frame, |
2887 | such as the dictionary ID, content size, or maximum back-reference distance (`windowSize`). |
2888 | Note that these values could be wrong, either because of data corruption, or because a 3rd party deliberately spoofs false information. |
2889 | As a consequence, check that values remain within valid application range. |
2890 | For example, do not allocate memory blindly, check that `windowSize` is within expectation. |
2891 | Each application can set its own limits, depending on local restrictions. |
2892 | For extended interoperability, it is recommended to support `windowSize` of at least 8 MB. |
2893 | |
2894 | ZSTD_decompressContinue() needs previous data blocks during decompression, up to `windowSize` bytes. |
2895 | ZSTD_decompressContinue() is very sensitive to contiguity, |
2896 | if 2 blocks don't follow each other, make sure that either the compressor breaks contiguity at the same place, |
2897 | or that previous contiguous segment is large enough to properly handle maximum back-reference distance. |
2898 | There are multiple ways to guarantee this condition. |
2899 | |
2900 | The most memory efficient way is to use a round buffer of sufficient size. |
2901 | Sufficient size is determined by invoking ZSTD_decodingBufferSize_min(), |
2902 | which can return an error code if required value is too large for current system (in 32-bits mode). |
2903 | In a round buffer methodology, ZSTD_decompressContinue() decompresses each block next to previous one, |
2904 | up to the moment there is not enough room left in the buffer to guarantee decoding another full block, |
2905 | which maximum size is provided in `ZSTD_frameHeader` structure, field `blockSizeMax`. |
2906 | At which point, decoding can resume from the beginning of the buffer. |
2907 | Note that already decoded data stored in the buffer should be flushed before being overwritten. |
2908 | |
2909 | There are alternatives possible, for example using two or more buffers of size `windowSize` each, though they consume more memory. |
2910 | |
2911 | Finally, if you control the compression process, you can also ignore all buffer size rules, |
2912 | as long as the encoder and decoder progress in "lock-step", |
2913 | aka use exactly the same buffer sizes, break contiguity at the same place, etc. |
2914 | |
2915 | Once buffers are setup, start decompression, with ZSTD_decompressBegin(). |
2916 | If decompression requires a dictionary, use ZSTD_decompressBegin_usingDict() or ZSTD_decompressBegin_usingDDict(). |
2917 | |
2918 | Then use ZSTD_nextSrcSizeToDecompress() and ZSTD_decompressContinue() alternatively. |
2919 | ZSTD_nextSrcSizeToDecompress() tells how many bytes to provide as 'srcSize' to ZSTD_decompressContinue(). |
2920 | ZSTD_decompressContinue() requires this _exact_ amount of bytes, or it will fail. |
2921 | |
2922 | result of ZSTD_decompressContinue() is the number of bytes regenerated within 'dst' (necessarily <= dstCapacity). |
2923 | It can be zero : it just means ZSTD_decompressContinue() has decoded some metadata item. |
2924 | It can also be an error code, which can be tested with ZSTD_isError(). |
2925 | |
2926 | A frame is fully decoded when ZSTD_nextSrcSizeToDecompress() returns zero. |
2927 | Context can then be reset to start a new decompression. |
2928 | |
2929 | Note : it's possible to know if next input to present is a header or a block, using ZSTD_nextInputType(). |
2930 | This information is not required to properly decode a frame. |
2931 | |
2932 | == Special case : skippable frames == |
2933 | |
2934 | Skippable frames allow integration of user-defined data into a flow of concatenated frames. |
2935 | Skippable frames will be ignored (skipped) by decompressor. |
2936 | The format of skippable frames is as follows : |
2937 | a) Skippable frame ID - 4 Bytes, Little endian format, any value from 0x184D2A50 to 0x184D2A5F |
2938 | b) Frame Size - 4 Bytes, Little endian format, unsigned 32-bits |
2939 | c) Frame Content - any content (User Data) of length equal to Frame Size |
2940 | For skippable frames ZSTD_getFrameHeader() returns zfhPtr->frameType==ZSTD_skippableFrame. |
2941 | For skippable frames ZSTD_decompressContinue() always returns 0 : it only skips the content. |
2942 | */ |
2943 | |
2944 | /*===== Buffer-less streaming decompression functions =====*/ |
2945 | |
2946 | ZSTDLIB_STATIC_API size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize); /**< when frame content size is not known, pass in frameContentSize == ZSTD_CONTENTSIZE_UNKNOWN */ |
2947 | |
2948 | ZSTDLIB_STATIC_API size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx); |
2949 | ZSTDLIB_STATIC_API size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); |
2950 | ZSTDLIB_STATIC_API size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); |
2951 | |
2952 | ZSTDLIB_STATIC_API size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx* dctx); |
2953 | ZSTDLIB_STATIC_API size_t ZSTD_decompressContinue(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); |
2954 | |
2955 | /* misc */ |
2956 | ZSTD_DEPRECATED("This function will likely be removed in the next minor release. It is misleading and has very limited utility." ) |
2957 | ZSTDLIB_STATIC_API void ZSTD_copyDCtx(ZSTD_DCtx* dctx, const ZSTD_DCtx* preparedDCtx); |
2958 | typedef enum { , , ZSTDnit_block, ZSTDnit_lastBlock, ZSTDnit_checksum, ZSTDnit_skippableFrame } ZSTD_nextInputType_e; |
2959 | ZSTDLIB_STATIC_API ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx* dctx); |
2960 | |
2961 | |
2962 | |
2963 | |
2964 | /* ========================================= */ |
2965 | /** Block level API (DEPRECATED) */ |
2966 | /* ========================================= */ |
2967 | |
2968 | /*! |
2969 | |
2970 | This API is deprecated in favor of the regular compression API. |
2971 | You can get the frame header down to 2 bytes by setting: |
2972 | - ZSTD_c_format = ZSTD_f_zstd1_magicless |
2973 | - ZSTD_c_contentSizeFlag = 0 |
2974 | - ZSTD_c_checksumFlag = 0 |
2975 | - ZSTD_c_dictIDFlag = 0 |
2976 | |
2977 | This API is not as well tested as our normal API, so we recommend not using it. |
2978 | We will be removing it in a future version. If the normal API doesn't provide |
2979 | the functionality you need, please open a GitHub issue. |
2980 | |
2981 | Block functions produce and decode raw zstd blocks, without frame metadata. |
2982 | Frame metadata cost is typically ~12 bytes, which can be non-negligible for very small blocks (< 100 bytes). |
2983 | But users will have to take in charge needed metadata to regenerate data, such as compressed and content sizes. |
2984 | |
2985 | A few rules to respect : |
2986 | - Compressing and decompressing require a context structure |
2987 | + Use ZSTD_createCCtx() and ZSTD_createDCtx() |
2988 | - It is necessary to init context before starting |
2989 | + compression : any ZSTD_compressBegin*() variant, including with dictionary |
2990 | + decompression : any ZSTD_decompressBegin*() variant, including with dictionary |
2991 | - Block size is limited, it must be <= ZSTD_getBlockSize() <= ZSTD_BLOCKSIZE_MAX == 128 KB |
2992 | + If input is larger than a block size, it's necessary to split input data into multiple blocks |
2993 | + For inputs larger than a single block, consider using regular ZSTD_compress() instead. |
2994 | Frame metadata is not that costly, and quickly becomes negligible as source size grows larger than a block. |
2995 | - When a block is considered not compressible enough, ZSTD_compressBlock() result will be 0 (zero) ! |
2996 | ===> In which case, nothing is produced into `dst` ! |
2997 | + User __must__ test for such outcome and deal directly with uncompressed data |
2998 | + A block cannot be declared incompressible if ZSTD_compressBlock() return value was != 0. |
2999 | Doing so would mess up with statistics history, leading to potential data corruption. |
3000 | + ZSTD_decompressBlock() _doesn't accept uncompressed data as input_ !! |
3001 | + In case of multiple successive blocks, should some of them be uncompressed, |
3002 | decoder must be informed of their existence in order to follow proper history. |
3003 | Use ZSTD_insertBlock() for such a case. |
3004 | */ |
3005 | |
3006 | /*===== Raw zstd block functions =====*/ |
3007 | ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs." ) |
3008 | ZSTDLIB_STATIC_API size_t ZSTD_getBlockSize (const ZSTD_CCtx* cctx); |
3009 | ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs." ) |
3010 | ZSTDLIB_STATIC_API size_t ZSTD_compressBlock (ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); |
3011 | ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs." ) |
3012 | ZSTDLIB_STATIC_API size_t ZSTD_decompressBlock(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); |
3013 | ZSTD_DEPRECATED("The block API is deprecated in favor of the normal compression API. See docs." ) |
3014 | ZSTDLIB_STATIC_API size_t ZSTD_insertBlock (ZSTD_DCtx* dctx, const void* blockStart, size_t blockSize); /**< insert uncompressed block into `dctx` history. Useful for multi-blocks decompression. */ |
3015 | |
3016 | #endif /* ZSTD_H_ZSTD_STATIC_LINKING_ONLY */ |
3017 | |
3018 | #if defined (__cplusplus) |
3019 | } |
3020 | #endif |
3021 | |