1 /* ****************************************************************** 2 * huff0 huffman codec, 3 * part of Finite State Entropy library 4 * Copyright (c) Meta Platforms, Inc. and affiliates. 5 * 6 * You can contact the author at : 7 * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy 8 * 9 * This source code is licensed under both the BSD-style license (found in the 10 * LICENSE file in the root directory of this source tree) and the GPLv2 (found 11 * in the COPYING file in the root directory of this source tree). 12 * You may select, at your option, one of the above-listed licenses. 13 ****************************************************************** */ 14 15 #ifndef HUF_H_298734234 16 #define HUF_H_298734234 17 18 /* *** Dependencies *** */ 19 #include "zstd_deps.h" /* size_t */ 20 #include "mem.h" /* U32 */ 21 #define FSE_STATIC_LINKING_ONLY 22 #include "fse.h" 23 24 /* *** Tool functions *** */ 25 #define HUF_BLOCKSIZE_MAX (128 * 1024) /**< maximum input size for a single block compressed with HUF_compress */ 26 size_t HUF_compressBound(size_t size); /**< maximum compressed size (worst case) */ 27 28 /* Error Management */ 29 unsigned HUF_isError(size_t code); /**< tells if a return value is an error code */ 30 const char* HUF_getErrorName(size_t code); /**< provides error code string (useful for debugging) */ 31 32 33 #define HUF_WORKSPACE_SIZE ((8 << 10) + 512 /* sorting scratch space */) 34 #define HUF_WORKSPACE_SIZE_U64 (HUF_WORKSPACE_SIZE / sizeof(U64)) 35 36 /* *** Constants *** */ 37 #define HUF_TABLELOG_MAX 12 /* max runtime value of tableLog (due to static allocation); can be modified up to HUF_TABLELOG_ABSOLUTEMAX */ 38 #define HUF_TABLELOG_DEFAULT 11 /* default tableLog value when none specified */ 39 #define HUF_SYMBOLVALUE_MAX 255 40 41 #define HUF_TABLELOG_ABSOLUTEMAX 12 /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */ 42 #if (HUF_TABLELOG_MAX > HUF_TABLELOG_ABSOLUTEMAX) 43 # error "HUF_TABLELOG_MAX is too large !" 44 #endif 45 46 47 /* **************************************** 48 * Static allocation 49 ******************************************/ 50 /* HUF buffer bounds */ 51 #define HUF_CTABLEBOUND 129 52 #define HUF_BLOCKBOUND(size) (size + (size>>8) + 8) /* only true when incompressible is pre-filtered with fast heuristic */ 53 #define HUF_COMPRESSBOUND(size) (HUF_CTABLEBOUND + HUF_BLOCKBOUND(size)) /* Macro version, useful for static allocation */ 54 55 /* static allocation of HUF's Compression Table */ 56 /* this is a private definition, just exposed for allocation and strict aliasing purpose. never EVER access its members directly */ 57 typedef size_t HUF_CElt; /* consider it an incomplete type */ 58 #define HUF_CTABLE_SIZE_ST(maxSymbolValue) ((maxSymbolValue)+2) /* Use tables of size_t, for proper alignment */ 59 #define HUF_CTABLE_SIZE(maxSymbolValue) (HUF_CTABLE_SIZE_ST(maxSymbolValue) * sizeof(size_t)) 60 #define HUF_CREATE_STATIC_CTABLE(name, maxSymbolValue) \ 61 HUF_CElt name[HUF_CTABLE_SIZE_ST(maxSymbolValue)] /* no final ; */ 62 63 /* static allocation of HUF's DTable */ 64 typedef U32 HUF_DTable; 65 #define HUF_DTABLE_SIZE(maxTableLog) (1 + (1<<(maxTableLog))) 66 #define HUF_CREATE_STATIC_DTABLEX1(DTable, maxTableLog) \ 67 HUF_DTable DTable[HUF_DTABLE_SIZE((maxTableLog)-1)] = { ((U32)((maxTableLog)-1) * 0x01000001) } 68 #define HUF_CREATE_STATIC_DTABLEX2(DTable, maxTableLog) \ 69 HUF_DTable DTable[HUF_DTABLE_SIZE(maxTableLog)] = { ((U32)(maxTableLog) * 0x01000001) } 70 71 72 /* **************************************** 73 * Advanced decompression functions 74 ******************************************/ 75 76 /** 77 * Huffman flags bitset. 78 * For all flags, 0 is the default value. 79 */ 80 typedef enum { 81 /** 82 * If compiled with DYNAMIC_BMI2: Set flag only if the CPU supports BMI2 at runtime. 83 * Otherwise: Ignored. 84 */ 85 HUF_flags_bmi2 = (1 << 0), 86 /** 87 * If set: Test possible table depths to find the one that produces the smallest header + encoded size. 88 * If unset: Use heuristic to find the table depth. 89 */ 90 HUF_flags_optimalDepth = (1 << 1), 91 /** 92 * If set: If the previous table can encode the input, always reuse the previous table. 93 * If unset: If the previous table can encode the input, reuse the previous table if it results in a smaller output. 94 */ 95 HUF_flags_preferRepeat = (1 << 2), 96 /** 97 * If set: Sample the input and check if the sample is uncompressible, if it is then don't attempt to compress. 98 * If unset: Always histogram the entire input. 99 */ 100 HUF_flags_suspectUncompressible = (1 << 3), 101 /** 102 * If set: Don't use assembly implementations 103 * If unset: Allow using assembly implementations 104 */ 105 HUF_flags_disableAsm = (1 << 4), 106 /** 107 * If set: Don't use the fast decoding loop, always use the fallback decoding loop. 108 * If unset: Use the fast decoding loop when possible. 109 */ 110 HUF_flags_disableFast = (1 << 5) 111 } HUF_flags_e; 112 113 114 /* **************************************** 115 * HUF detailed API 116 * ****************************************/ 117 #define HUF_OPTIMAL_DEPTH_THRESHOLD ZSTD_btultra 118 119 /*! HUF_compress() does the following: 120 * 1. count symbol occurrence from source[] into table count[] using FSE_count() (exposed within "fse.h") 121 * 2. (optional) refine tableLog using HUF_optimalTableLog() 122 * 3. build Huffman table from count using HUF_buildCTable() 123 * 4. save Huffman table to memory buffer using HUF_writeCTable() 124 * 5. encode the data stream using HUF_compress4X_usingCTable() 125 * 126 * The following API allows targeting specific sub-functions for advanced tasks. 127 * For example, it's possible to compress several blocks using the same 'CTable', 128 * or to save and regenerate 'CTable' using external methods. 129 */ 130 unsigned HUF_minTableLog(unsigned symbolCardinality); 131 unsigned HUF_cardinality(const unsigned* count, unsigned maxSymbolValue); 132 unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, void* workSpace, 133 size_t wkspSize, HUF_CElt* table, const unsigned* count, int flags); /* table is used as scratch space for building and testing tables, not a return value */ 134 size_t HUF_writeCTable_wksp(void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog, void* workspace, size_t workspaceSize); 135 size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int flags); 136 size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue); 137 int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue); 138 139 typedef enum { 140 HUF_repeat_none, /**< Cannot use the previous table */ 141 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 */ 142 HUF_repeat_valid /**< Can use the previous table and it is assumed to be valid */ 143 } HUF_repeat; 144 145 /** HUF_compress4X_repeat() : 146 * Same as HUF_compress4X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none. 147 * If it uses hufTable it does not modify hufTable or repeat. 148 * If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used. 149 * If preferRepeat then the old table will always be used if valid. 150 * If suspectUncompressible then some sampling checks will be run to potentially skip huffman coding */ 151 size_t HUF_compress4X_repeat(void* dst, size_t dstSize, 152 const void* src, size_t srcSize, 153 unsigned maxSymbolValue, unsigned tableLog, 154 void* workSpace, size_t wkspSize, /**< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */ 155 HUF_CElt* hufTable, HUF_repeat* repeat, int flags); 156 157 /** HUF_buildCTable_wksp() : 158 * Same as HUF_buildCTable(), but using externally allocated scratch buffer. 159 * `workSpace` must be aligned on 4-bytes boundaries, and its size must be >= HUF_CTABLE_WORKSPACE_SIZE. 160 */ 161 #define HUF_CTABLE_WORKSPACE_SIZE_U32 ((4 * (HUF_SYMBOLVALUE_MAX + 1)) + 192) 162 #define HUF_CTABLE_WORKSPACE_SIZE (HUF_CTABLE_WORKSPACE_SIZE_U32 * sizeof(unsigned)) 163 size_t HUF_buildCTable_wksp (HUF_CElt* tree, 164 const unsigned* count, U32 maxSymbolValue, U32 maxNbBits, 165 void* workSpace, size_t wkspSize); 166 167 /*! HUF_readStats() : 168 * Read compact Huffman tree, saved by HUF_writeCTable(). 169 * `huffWeight` is destination buffer. 170 * @return : size read from `src` , or an error Code . 171 * Note : Needed by HUF_readCTable() and HUF_readDTableXn() . */ 172 size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, 173 U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr, 174 const void* src, size_t srcSize); 175 176 /*! HUF_readStats_wksp() : 177 * Same as HUF_readStats() but takes an external workspace which must be 178 * 4-byte aligned and its size must be >= HUF_READ_STATS_WORKSPACE_SIZE. 179 * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0. 180 */ 181 #define HUF_READ_STATS_WORKSPACE_SIZE_U32 FSE_DECOMPRESS_WKSP_SIZE_U32(6, HUF_TABLELOG_MAX-1) 182 #define HUF_READ_STATS_WORKSPACE_SIZE (HUF_READ_STATS_WORKSPACE_SIZE_U32 * sizeof(unsigned)) 183 size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, 184 U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr, 185 const void* src, size_t srcSize, 186 void* workspace, size_t wkspSize, 187 int flags); 188 189 /** HUF_readCTable() : 190 * Loading a CTable saved with HUF_writeCTable() */ 191 size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned *hasZeroWeights); 192 193 /** HUF_getNbBitsFromCTable() : 194 * Read nbBits from CTable symbolTable, for symbol `symbolValue` presumed <= HUF_SYMBOLVALUE_MAX 195 * Note 1 : If symbolValue > HUF_readCTableHeader(symbolTable).maxSymbolValue, returns 0 196 * Note 2 : is not inlined, as HUF_CElt definition is private 197 */ 198 U32 HUF_getNbBitsFromCTable(const HUF_CElt* symbolTable, U32 symbolValue); 199 200 typedef struct { 201 BYTE tableLog; 202 BYTE maxSymbolValue; 203 BYTE unused[sizeof(size_t) - 2]; 204 } HUF_CTableHeader; 205 206 /** HUF_readCTableHeader() : 207 * @returns The header from the CTable specifying the tableLog and the maxSymbolValue. 208 */ 209 HUF_CTableHeader HUF_readCTableHeader(HUF_CElt const* ctable); 210 211 /* 212 * HUF_decompress() does the following: 213 * 1. select the decompression algorithm (X1, X2) based on pre-computed heuristics 214 * 2. build Huffman table from save, using HUF_readDTableX?() 215 * 3. decode 1 or 4 segments in parallel using HUF_decompress?X?_usingDTable() 216 */ 217 218 /** HUF_selectDecoder() : 219 * Tells which decoder is likely to decode faster, 220 * based on a set of pre-computed metrics. 221 * @return : 0==HUF_decompress4X1, 1==HUF_decompress4X2 . 222 * Assumption : 0 < dstSize <= 128 KB */ 223 U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize); 224 225 /** 226 * The minimum workspace size for the `workSpace` used in 227 * HUF_readDTableX1_wksp() and HUF_readDTableX2_wksp(). 228 * 229 * The space used depends on HUF_TABLELOG_MAX, ranging from ~1500 bytes when 230 * HUF_TABLE_LOG_MAX=12 to ~1850 bytes when HUF_TABLE_LOG_MAX=15. 231 * Buffer overflow errors may potentially occur if code modifications result in 232 * a required workspace size greater than that specified in the following 233 * macro. 234 */ 235 #define HUF_DECOMPRESS_WORKSPACE_SIZE ((2 << 10) + (1 << 9)) 236 #define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32)) 237 238 239 /* ====================== */ 240 /* single stream variants */ 241 /* ====================== */ 242 243 size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable, int flags); 244 /** HUF_compress1X_repeat() : 245 * Same as HUF_compress1X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none. 246 * If it uses hufTable it does not modify hufTable or repeat. 247 * If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used. 248 * If preferRepeat then the old table will always be used if valid. 249 * If suspectUncompressible then some sampling checks will be run to potentially skip huffman coding */ 250 size_t HUF_compress1X_repeat(void* dst, size_t dstSize, 251 const void* src, size_t srcSize, 252 unsigned maxSymbolValue, unsigned tableLog, 253 void* workSpace, size_t wkspSize, /**< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */ 254 HUF_CElt* hufTable, HUF_repeat* repeat, int flags); 255 256 size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags); 257 #ifndef HUF_FORCE_DECOMPRESS_X1 258 size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags); /**< double-symbols decoder */ 259 #endif 260 261 /* BMI2 variants. 262 * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0. 263 */ 264 size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int flags); 265 #ifndef HUF_FORCE_DECOMPRESS_X2 266 size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags); 267 #endif 268 size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int flags); 269 size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int flags); 270 #ifndef HUF_FORCE_DECOMPRESS_X2 271 size_t HUF_readDTableX1_wksp(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int flags); 272 #endif 273 #ifndef HUF_FORCE_DECOMPRESS_X1 274 size_t HUF_readDTableX2_wksp(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int flags); 275 #endif 276 277 #endif /* HUF_H_298734234 */ 278
Indexes created Fri Jun 12 00:25:51 UTC 2026