blob: b4645b4e51975104f838cc01aba1c4bf25742605 [file] [log] [blame]
Scott Baker2d897982019-09-24 11:50:08 -07001/* ******************************************************************
2 Huffman coder, part of New Generation Entropy library
3 header file
4 Copyright (C) 2013-2016, Yann Collet.
5
6 BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
7
8 Redistribution and use in source and binary forms, with or without
9 modification, are permitted provided that the following conditions are
10 met:
11
12 * Redistributions of source code must retain the above copyright
13 notice, this list of conditions and the following disclaimer.
14 * Redistributions in binary form must reproduce the above
15 copyright notice, this list of conditions and the following disclaimer
16 in the documentation and/or other materials provided with the
17 distribution.
18
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30
31 You can contact the author at :
32 - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
33****************************************************************** */
34
35#if defined (__cplusplus)
36extern "C" {
37#endif
38
39#ifndef HUF_H_298734234
40#define HUF_H_298734234
41
42/* *** Dependencies *** */
43#include <stddef.h> /* size_t */
44
45
46/* *** library symbols visibility *** */
47/* Note : when linking with -fvisibility=hidden on gcc, or by default on Visual,
48 * HUF symbols remain "private" (internal symbols for library only).
49 * Set macro FSE_DLL_EXPORT to 1 if you want HUF symbols visible on DLL interface */
50#if defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) && defined(__GNUC__) && (__GNUC__ >= 4)
51# define HUF_PUBLIC_API __attribute__ ((visibility ("default")))
52#elif defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) /* Visual expected */
53# define HUF_PUBLIC_API __declspec(dllexport)
54#elif defined(FSE_DLL_IMPORT) && (FSE_DLL_IMPORT==1)
55# define HUF_PUBLIC_API __declspec(dllimport) /* not required, just to generate faster code (saves a function pointer load from IAT and an indirect jump) */
56#else
57# define HUF_PUBLIC_API
58#endif
59
60
61/* ========================== */
62/* *** simple functions *** */
63/* ========================== */
64
65/** HUF_compress() :
66 * Compress content from buffer 'src', of size 'srcSize', into buffer 'dst'.
67 * 'dst' buffer must be already allocated.
68 * Compression runs faster if `dstCapacity` >= HUF_compressBound(srcSize).
69 * `srcSize` must be <= `HUF_BLOCKSIZE_MAX` == 128 KB.
70 * @return : size of compressed data (<= `dstCapacity`).
71 * Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!!
72 * if HUF_isError(return), compression failed (more details using HUF_getErrorName())
73 */
74HUF_PUBLIC_API size_t HUF_compress(void* dst, size_t dstCapacity,
75 const void* src, size_t srcSize);
76
77/** HUF_decompress() :
78 * Decompress HUF data from buffer 'cSrc', of size 'cSrcSize',
79 * into already allocated buffer 'dst', of minimum size 'dstSize'.
80 * `originalSize` : **must** be the ***exact*** size of original (uncompressed) data.
81 * Note : in contrast with FSE, HUF_decompress can regenerate
82 * RLE (cSrcSize==1) and uncompressed (cSrcSize==dstSize) data,
83 * because it knows size to regenerate (originalSize).
84 * @return : size of regenerated data (== originalSize),
85 * or an error code, which can be tested using HUF_isError()
86 */
87HUF_PUBLIC_API size_t HUF_decompress(void* dst, size_t originalSize,
88 const void* cSrc, size_t cSrcSize);
89
90
91/* *** Tool functions *** */
92#define HUF_BLOCKSIZE_MAX (128 * 1024) /**< maximum input size for a single block compressed with HUF_compress */
93HUF_PUBLIC_API size_t HUF_compressBound(size_t size); /**< maximum compressed size (worst case) */
94
95/* Error Management */
96HUF_PUBLIC_API unsigned HUF_isError(size_t code); /**< tells if a return value is an error code */
97HUF_PUBLIC_API const char* HUF_getErrorName(size_t code); /**< provides error code string (useful for debugging) */
98
99
100/* *** Advanced function *** */
101
102/** HUF_compress2() :
103 * Same as HUF_compress(), but offers control over `maxSymbolValue` and `tableLog`.
104 * `maxSymbolValue` must be <= HUF_SYMBOLVALUE_MAX .
105 * `tableLog` must be `<= HUF_TABLELOG_MAX` . */
106HUF_PUBLIC_API size_t HUF_compress2 (void* dst, size_t dstCapacity,
107 const void* src, size_t srcSize,
108 unsigned maxSymbolValue, unsigned tableLog);
109
110/** HUF_compress4X_wksp() :
111 * Same as HUF_compress2(), but uses externally allocated `workSpace`.
112 * `workspace` must have minimum alignment of 4, and be at least as large as HUF_WORKSPACE_SIZE */
113#define HUF_WORKSPACE_SIZE (6 << 10)
114#define HUF_WORKSPACE_SIZE_U32 (HUF_WORKSPACE_SIZE / sizeof(U32))
115HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
116 const void* src, size_t srcSize,
117 unsigned maxSymbolValue, unsigned tableLog,
118 void* workSpace, size_t wkspSize);
119
120#endif /* HUF_H_298734234 */
121
122/* ******************************************************************
123 * WARNING !!
124 * The following section contains advanced and experimental definitions
125 * which shall never be used in the context of a dynamic library,
126 * because they are not guaranteed to remain stable in the future.
127 * Only consider them in association with static linking.
128 * *****************************************************************/
129#if defined(HUF_STATIC_LINKING_ONLY) && !defined(HUF_H_HUF_STATIC_LINKING_ONLY)
130#define HUF_H_HUF_STATIC_LINKING_ONLY
131
132/* *** Dependencies *** */
133#include "mem.h" /* U32 */
134
135
136/* *** Constants *** */
137#define HUF_TABLELOG_MAX 12 /* max runtime value of tableLog (due to static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */
138#define HUF_TABLELOG_DEFAULT 11 /* default tableLog value when none specified */
139#define HUF_SYMBOLVALUE_MAX 255
140
141#define HUF_TABLELOG_ABSOLUTEMAX 15 /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */
142#if (HUF_TABLELOG_MAX > HUF_TABLELOG_ABSOLUTEMAX)
143# error "HUF_TABLELOG_MAX is too large !"
144#endif
145
146
147/* ****************************************
148* Static allocation
149******************************************/
150/* HUF buffer bounds */
151#define HUF_CTABLEBOUND 129
152#define HUF_BLOCKBOUND(size) (size + (size>>8) + 8) /* only true when incompressible is pre-filtered with fast heuristic */
153#define HUF_COMPRESSBOUND(size) (HUF_CTABLEBOUND + HUF_BLOCKBOUND(size)) /* Macro version, useful for static allocation */
154
155/* static allocation of HUF's Compression Table */
156#define HUF_CTABLE_SIZE_U32(maxSymbolValue) ((maxSymbolValue)+1) /* Use tables of U32, for proper alignment */
157#define HUF_CTABLE_SIZE(maxSymbolValue) (HUF_CTABLE_SIZE_U32(maxSymbolValue) * sizeof(U32))
158#define HUF_CREATE_STATIC_CTABLE(name, maxSymbolValue) \
159 U32 name##hb[HUF_CTABLE_SIZE_U32(maxSymbolValue)]; \
160 void* name##hv = &(name##hb); \
161 HUF_CElt* name = (HUF_CElt*)(name##hv) /* no final ; */
162
163/* static allocation of HUF's DTable */
164typedef U32 HUF_DTable;
165#define HUF_DTABLE_SIZE(maxTableLog) (1 + (1<<(maxTableLog)))
166#define HUF_CREATE_STATIC_DTABLEX2(DTable, maxTableLog) \
167 HUF_DTable DTable[HUF_DTABLE_SIZE((maxTableLog)-1)] = { ((U32)((maxTableLog)-1) * 0x01000001) }
168#define HUF_CREATE_STATIC_DTABLEX4(DTable, maxTableLog) \
169 HUF_DTable DTable[HUF_DTABLE_SIZE(maxTableLog)] = { ((U32)(maxTableLog) * 0x01000001) }
170
171
172/* ****************************************
173* Advanced decompression functions
174******************************************/
175size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */
176size_t HUF_decompress4X4 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */
177
178size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< decodes RLE and uncompressed */
179size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< considers RLE and uncompressed as errors */
180size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< considers RLE and uncompressed as errors */
181size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */
182size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< single-symbol decoder */
183size_t HUF_decompress4X4_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */
184size_t HUF_decompress4X4_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< double-symbols decoder */
185
186
187/* ****************************************
188 * HUF detailed API
189 * ****************************************/
190
191/*! HUF_compress() does the following:
192 * 1. count symbol occurrence from source[] into table count[] using FSE_count() (exposed within "fse.h")
193 * 2. (optional) refine tableLog using HUF_optimalTableLog()
194 * 3. build Huffman table from count using HUF_buildCTable()
195 * 4. save Huffman table to memory buffer using HUF_writeCTable()
196 * 5. encode the data stream using HUF_compress4X_usingCTable()
197 *
198 * The following API allows targeting specific sub-functions for advanced tasks.
199 * For example, it's possible to compress several blocks using the same 'CTable',
200 * or to save and regenerate 'CTable' using external methods.
201 */
202unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue);
203typedef struct HUF_CElt_s HUF_CElt; /* incomplete type */
204size_t HUF_buildCTable (HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits); /* @return : maxNbBits; CTable and count can overlap. In which case, CTable will overwrite count content */
205size_t HUF_writeCTable (void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog);
206size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable);
207
208typedef enum {
209 HUF_repeat_none, /**< Cannot use the previous table */
210 HUF_repeat_check, /**< Can use the previous table but it must be checked. Note : The previous table must have been constructed by HUF_compress{1, 4}X_repeat */
211 HUF_repeat_valid /**< Can use the previous table and it is asumed to be valid */
212 } HUF_repeat;
213/** HUF_compress4X_repeat() :
214 * Same as HUF_compress4X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none.
215 * If it uses hufTable it does not modify hufTable or repeat.
216 * If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used.
217 * If preferRepeat then the old table will always be used if valid. */
218size_t HUF_compress4X_repeat(void* dst, size_t dstSize,
219 const void* src, size_t srcSize,
220 unsigned maxSymbolValue, unsigned tableLog,
221 void* workSpace, size_t wkspSize, /**< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */
222 HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2);
223
224/** HUF_buildCTable_wksp() :
225 * Same as HUF_buildCTable(), but using externally allocated scratch buffer.
226 * `workSpace` must be aligned on 4-bytes boundaries, and its size must be >= HUF_CTABLE_WORKSPACE_SIZE.
227 */
228#define HUF_CTABLE_WORKSPACE_SIZE_U32 (2*HUF_SYMBOLVALUE_MAX +1 +1)
229#define HUF_CTABLE_WORKSPACE_SIZE (HUF_CTABLE_WORKSPACE_SIZE_U32 * sizeof(unsigned))
230size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize);
231
232/*! HUF_readStats() :
233 * Read compact Huffman tree, saved by HUF_writeCTable().
234 * `huffWeight` is destination buffer.
235 * @return : size read from `src` , or an error Code .
236 * Note : Needed by HUF_readCTable() and HUF_readDTableXn() . */
237size_t HUF_readStats(BYTE* huffWeight, size_t hwSize,
238 U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr,
239 const void* src, size_t srcSize);
240
241/** HUF_readCTable() :
242 * Loading a CTable saved with HUF_writeCTable() */
243size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
244
245
246/*
247 * HUF_decompress() does the following:
248 * 1. select the decompression algorithm (X2, X4) based on pre-computed heuristics
249 * 2. build Huffman table from save, using HUF_readDTableX?()
250 * 3. decode 1 or 4 segments in parallel using HUF_decompress?X?_usingDTable()
251 */
252
253/** HUF_selectDecoder() :
254 * Tells which decoder is likely to decode faster,
255 * based on a set of pre-computed metrics.
256 * @return : 0==HUF_decompress4X2, 1==HUF_decompress4X4 .
257 * Assumption : 0 < dstSize <= 128 KB */
258U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize);
259
260/**
261 * The minimum workspace size for the `workSpace` used in
262 * HUF_readDTableX2_wksp() and HUF_readDTableX4_wksp().
263 *
264 * The space used depends on HUF_TABLELOG_MAX, ranging from ~1500 bytes when
265 * HUF_TABLE_LOG_MAX=12 to ~1850 bytes when HUF_TABLE_LOG_MAX=15.
266 * Buffer overflow errors may potentially occur if code modifications result in
267 * a required workspace size greater than that specified in the following
268 * macro.
269 */
270#define HUF_DECOMPRESS_WORKSPACE_SIZE (2 << 10)
271#define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32))
272
273size_t HUF_readDTableX2 (HUF_DTable* DTable, const void* src, size_t srcSize);
274size_t HUF_readDTableX2_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize);
275size_t HUF_readDTableX4 (HUF_DTable* DTable, const void* src, size_t srcSize);
276size_t HUF_readDTableX4_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize);
277
278size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
279size_t HUF_decompress4X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
280size_t HUF_decompress4X4_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
281
282
283/* ====================== */
284/* single stream variants */
285/* ====================== */
286
287size_t HUF_compress1X (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog);
288size_t HUF_compress1X_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize); /**< `workSpace` must be a table of at least HUF_WORKSPACE_SIZE_U32 unsigned */
289size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable);
290/** HUF_compress1X_repeat() :
291 * Same as HUF_compress1X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none.
292 * If it uses hufTable it does not modify hufTable or repeat.
293 * If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used.
294 * If preferRepeat then the old table will always be used if valid. */
295size_t HUF_compress1X_repeat(void* dst, size_t dstSize,
296 const void* src, size_t srcSize,
297 unsigned maxSymbolValue, unsigned tableLog,
298 void* workSpace, size_t wkspSize, /**< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */
299 HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2);
300
301size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* single-symbol decoder */
302size_t HUF_decompress1X4 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* double-symbol decoder */
303
304size_t HUF_decompress1X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);
305size_t HUF_decompress1X_DCtx_wksp (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);
306size_t HUF_decompress1X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */
307size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< single-symbol decoder */
308size_t HUF_decompress1X4_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */
309size_t HUF_decompress1X4_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< double-symbols decoder */
310
311size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); /**< automatic selection of sing or double symbol decoder, based on DTable */
312size_t HUF_decompress1X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
313size_t HUF_decompress1X4_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
314
315/* BMI2 variants.
316 * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0.
317 */
318size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
319size_t HUF_decompress1X2_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
320size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
321size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
322
323#endif /* HUF_STATIC_LINKING_ONLY */
324
325#if defined (__cplusplus)
326}
327#endif