1*4882a593Smuzhiyun /* LzmaDec.h -- LZMA Decoder 2*4882a593Smuzhiyun 2009-02-07 : Igor Pavlov : Public domain */ 3*4882a593Smuzhiyun 4*4882a593Smuzhiyun #ifndef __LZMA_DEC_H 5*4882a593Smuzhiyun #define __LZMA_DEC_H 6*4882a593Smuzhiyun 7*4882a593Smuzhiyun #include "Types.h" 8*4882a593Smuzhiyun 9*4882a593Smuzhiyun /* #define _LZMA_PROB32 */ 10*4882a593Smuzhiyun /* _LZMA_PROB32 can increase the speed on some CPUs, 11*4882a593Smuzhiyun but memory usage for CLzmaDec::probs will be doubled in that case */ 12*4882a593Smuzhiyun 13*4882a593Smuzhiyun #ifdef _LZMA_PROB32 14*4882a593Smuzhiyun #define CLzmaProb UInt32 15*4882a593Smuzhiyun #else 16*4882a593Smuzhiyun #define CLzmaProb UInt16 17*4882a593Smuzhiyun #endif 18*4882a593Smuzhiyun 19*4882a593Smuzhiyun 20*4882a593Smuzhiyun /* ---------- LZMA Properties ---------- */ 21*4882a593Smuzhiyun 22*4882a593Smuzhiyun #define LZMA_PROPS_SIZE 5 23*4882a593Smuzhiyun 24*4882a593Smuzhiyun typedef struct _CLzmaProps 25*4882a593Smuzhiyun { 26*4882a593Smuzhiyun unsigned lc, lp, pb; 27*4882a593Smuzhiyun UInt32 dicSize; 28*4882a593Smuzhiyun } CLzmaProps; 29*4882a593Smuzhiyun 30*4882a593Smuzhiyun /* LzmaProps_Decode - decodes properties 31*4882a593Smuzhiyun Returns: 32*4882a593Smuzhiyun SZ_OK 33*4882a593Smuzhiyun SZ_ERROR_UNSUPPORTED - Unsupported properties 34*4882a593Smuzhiyun */ 35*4882a593Smuzhiyun 36*4882a593Smuzhiyun SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size); 37*4882a593Smuzhiyun 38*4882a593Smuzhiyun 39*4882a593Smuzhiyun /* ---------- LZMA Decoder state ---------- */ 40*4882a593Smuzhiyun 41*4882a593Smuzhiyun /* LZMA_REQUIRED_INPUT_MAX = number of required input bytes for worst case. 42*4882a593Smuzhiyun Num bits = log2((2^11 / 31) ^ 22) + 26 < 134 + 26 = 160; */ 43*4882a593Smuzhiyun 44*4882a593Smuzhiyun #define LZMA_REQUIRED_INPUT_MAX 20 45*4882a593Smuzhiyun 46*4882a593Smuzhiyun typedef struct 47*4882a593Smuzhiyun { 48*4882a593Smuzhiyun CLzmaProps prop; 49*4882a593Smuzhiyun CLzmaProb *probs; 50*4882a593Smuzhiyun Byte *dic; 51*4882a593Smuzhiyun const Byte *buf; 52*4882a593Smuzhiyun UInt32 range, code; 53*4882a593Smuzhiyun SizeT dicPos; 54*4882a593Smuzhiyun SizeT dicBufSize; 55*4882a593Smuzhiyun UInt32 processedPos; 56*4882a593Smuzhiyun UInt32 checkDicSize; 57*4882a593Smuzhiyun unsigned state; 58*4882a593Smuzhiyun UInt32 reps[4]; 59*4882a593Smuzhiyun unsigned remainLen; 60*4882a593Smuzhiyun int needFlush; 61*4882a593Smuzhiyun int needInitState; 62*4882a593Smuzhiyun UInt32 numProbs; 63*4882a593Smuzhiyun unsigned tempBufSize; 64*4882a593Smuzhiyun Byte tempBuf[LZMA_REQUIRED_INPUT_MAX]; 65*4882a593Smuzhiyun } CLzmaDec; 66*4882a593Smuzhiyun 67*4882a593Smuzhiyun #define LzmaDec_Construct(p) { (p)->dic = 0; (p)->probs = 0; } 68*4882a593Smuzhiyun 69*4882a593Smuzhiyun void LzmaDec_Init(CLzmaDec *p); 70*4882a593Smuzhiyun 71*4882a593Smuzhiyun /* There are two types of LZMA streams: 72*4882a593Smuzhiyun 0) Stream with end mark. That end mark adds about 6 bytes to compressed size. 73*4882a593Smuzhiyun 1) Stream without end mark. You must know exact uncompressed size to decompress such stream. */ 74*4882a593Smuzhiyun 75*4882a593Smuzhiyun typedef enum 76*4882a593Smuzhiyun { 77*4882a593Smuzhiyun LZMA_FINISH_ANY, /* finish at any point */ 78*4882a593Smuzhiyun LZMA_FINISH_END /* block must be finished at the end */ 79*4882a593Smuzhiyun } ELzmaFinishMode; 80*4882a593Smuzhiyun 81*4882a593Smuzhiyun /* ELzmaFinishMode has meaning only if the decoding reaches output limit !!! 82*4882a593Smuzhiyun 83*4882a593Smuzhiyun You must use LZMA_FINISH_END, when you know that current output buffer 84*4882a593Smuzhiyun covers last bytes of block. In other cases you must use LZMA_FINISH_ANY. 85*4882a593Smuzhiyun 86*4882a593Smuzhiyun If LZMA decoder sees end marker before reaching output limit, it returns SZ_OK, 87*4882a593Smuzhiyun and output value of destLen will be less than output buffer size limit. 88*4882a593Smuzhiyun You can check status result also. 89*4882a593Smuzhiyun 90*4882a593Smuzhiyun You can use multiple checks to test data integrity after full decompression: 91*4882a593Smuzhiyun 1) Check Result and "status" variable. 92*4882a593Smuzhiyun 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. 93*4882a593Smuzhiyun 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. 94*4882a593Smuzhiyun You must use correct finish mode in that case. */ 95*4882a593Smuzhiyun 96*4882a593Smuzhiyun typedef enum 97*4882a593Smuzhiyun { 98*4882a593Smuzhiyun LZMA_STATUS_NOT_SPECIFIED, /* use main error code instead */ 99*4882a593Smuzhiyun LZMA_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ 100*4882a593Smuzhiyun LZMA_STATUS_NOT_FINISHED, /* stream was not finished */ 101*4882a593Smuzhiyun LZMA_STATUS_NEEDS_MORE_INPUT, /* you must provide more input bytes */ 102*4882a593Smuzhiyun LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK /* there is probability that stream was finished without end mark */ 103*4882a593Smuzhiyun } ELzmaStatus; 104*4882a593Smuzhiyun 105*4882a593Smuzhiyun /* ELzmaStatus is used only as output value for function call */ 106*4882a593Smuzhiyun 107*4882a593Smuzhiyun 108*4882a593Smuzhiyun /* ---------- Interfaces ---------- */ 109*4882a593Smuzhiyun 110*4882a593Smuzhiyun /* There are 3 levels of interfaces: 111*4882a593Smuzhiyun 1) Dictionary Interface 112*4882a593Smuzhiyun 2) Buffer Interface 113*4882a593Smuzhiyun 3) One Call Interface 114*4882a593Smuzhiyun You can select any of these interfaces, but don't mix functions from different 115*4882a593Smuzhiyun groups for same object. */ 116*4882a593Smuzhiyun 117*4882a593Smuzhiyun 118*4882a593Smuzhiyun /* There are two variants to allocate state for Dictionary Interface: 119*4882a593Smuzhiyun 1) LzmaDec_Allocate / LzmaDec_Free 120*4882a593Smuzhiyun 2) LzmaDec_AllocateProbs / LzmaDec_FreeProbs 121*4882a593Smuzhiyun You can use variant 2, if you set dictionary buffer manually. 122*4882a593Smuzhiyun For Buffer Interface you must always use variant 1. 123*4882a593Smuzhiyun 124*4882a593Smuzhiyun LzmaDec_Allocate* can return: 125*4882a593Smuzhiyun SZ_OK 126*4882a593Smuzhiyun SZ_ERROR_MEM - Memory allocation error 127*4882a593Smuzhiyun SZ_ERROR_UNSUPPORTED - Unsupported properties 128*4882a593Smuzhiyun */ 129*4882a593Smuzhiyun 130*4882a593Smuzhiyun SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAlloc *alloc); 131*4882a593Smuzhiyun void LzmaDec_FreeProbs(CLzmaDec *p, ISzAlloc *alloc); 132*4882a593Smuzhiyun 133*4882a593Smuzhiyun SRes LzmaDec_Allocate(CLzmaDec *state, const Byte *prop, unsigned propsSize, ISzAlloc *alloc); 134*4882a593Smuzhiyun void LzmaDec_Free(CLzmaDec *state, ISzAlloc *alloc); 135*4882a593Smuzhiyun 136*4882a593Smuzhiyun /* ---------- Dictionary Interface ---------- */ 137*4882a593Smuzhiyun 138*4882a593Smuzhiyun /* You can use it, if you want to eliminate the overhead for data copying from 139*4882a593Smuzhiyun dictionary to some other external buffer. 140*4882a593Smuzhiyun You must work with CLzmaDec variables directly in this interface. 141*4882a593Smuzhiyun 142*4882a593Smuzhiyun STEPS: 143*4882a593Smuzhiyun LzmaDec_Constr() 144*4882a593Smuzhiyun LzmaDec_Allocate() 145*4882a593Smuzhiyun for (each new stream) 146*4882a593Smuzhiyun { 147*4882a593Smuzhiyun LzmaDec_Init() 148*4882a593Smuzhiyun while (it needs more decompression) 149*4882a593Smuzhiyun { 150*4882a593Smuzhiyun LzmaDec_DecodeToDic() 151*4882a593Smuzhiyun use data from CLzmaDec::dic and update CLzmaDec::dicPos 152*4882a593Smuzhiyun } 153*4882a593Smuzhiyun } 154*4882a593Smuzhiyun LzmaDec_Free() 155*4882a593Smuzhiyun */ 156*4882a593Smuzhiyun 157*4882a593Smuzhiyun /* LzmaDec_DecodeToDic 158*4882a593Smuzhiyun 159*4882a593Smuzhiyun The decoding to internal dictionary buffer (CLzmaDec::dic). 160*4882a593Smuzhiyun You must manually update CLzmaDec::dicPos, if it reaches CLzmaDec::dicBufSize !!! 161*4882a593Smuzhiyun 162*4882a593Smuzhiyun finishMode: 163*4882a593Smuzhiyun It has meaning only if the decoding reaches output limit (dicLimit). 164*4882a593Smuzhiyun LZMA_FINISH_ANY - Decode just dicLimit bytes. 165*4882a593Smuzhiyun LZMA_FINISH_END - Stream must be finished after dicLimit. 166*4882a593Smuzhiyun 167*4882a593Smuzhiyun Returns: 168*4882a593Smuzhiyun SZ_OK 169*4882a593Smuzhiyun status: 170*4882a593Smuzhiyun LZMA_STATUS_FINISHED_WITH_MARK 171*4882a593Smuzhiyun LZMA_STATUS_NOT_FINISHED 172*4882a593Smuzhiyun LZMA_STATUS_NEEDS_MORE_INPUT 173*4882a593Smuzhiyun LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK 174*4882a593Smuzhiyun SZ_ERROR_DATA - Data error 175*4882a593Smuzhiyun */ 176*4882a593Smuzhiyun 177*4882a593Smuzhiyun SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, 178*4882a593Smuzhiyun const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); 179*4882a593Smuzhiyun 180*4882a593Smuzhiyun 181*4882a593Smuzhiyun /* ---------- Buffer Interface ---------- */ 182*4882a593Smuzhiyun 183*4882a593Smuzhiyun /* It's zlib-like interface. 184*4882a593Smuzhiyun See LzmaDec_DecodeToDic description for information about STEPS and return results, 185*4882a593Smuzhiyun but you must use LzmaDec_DecodeToBuf instead of LzmaDec_DecodeToDic and you don't need 186*4882a593Smuzhiyun to work with CLzmaDec variables manually. 187*4882a593Smuzhiyun 188*4882a593Smuzhiyun finishMode: 189*4882a593Smuzhiyun It has meaning only if the decoding reaches output limit (*destLen). 190*4882a593Smuzhiyun LZMA_FINISH_ANY - Decode just destLen bytes. 191*4882a593Smuzhiyun LZMA_FINISH_END - Stream must be finished after (*destLen). 192*4882a593Smuzhiyun */ 193*4882a593Smuzhiyun 194*4882a593Smuzhiyun SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, 195*4882a593Smuzhiyun const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); 196*4882a593Smuzhiyun 197*4882a593Smuzhiyun 198*4882a593Smuzhiyun /* ---------- One Call Interface ---------- */ 199*4882a593Smuzhiyun 200*4882a593Smuzhiyun /* LzmaDecode 201*4882a593Smuzhiyun 202*4882a593Smuzhiyun finishMode: 203*4882a593Smuzhiyun It has meaning only if the decoding reaches output limit (*destLen). 204*4882a593Smuzhiyun LZMA_FINISH_ANY - Decode just destLen bytes. 205*4882a593Smuzhiyun LZMA_FINISH_END - Stream must be finished after (*destLen). 206*4882a593Smuzhiyun 207*4882a593Smuzhiyun Returns: 208*4882a593Smuzhiyun SZ_OK 209*4882a593Smuzhiyun status: 210*4882a593Smuzhiyun LZMA_STATUS_FINISHED_WITH_MARK 211*4882a593Smuzhiyun LZMA_STATUS_NOT_FINISHED 212*4882a593Smuzhiyun LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK 213*4882a593Smuzhiyun SZ_ERROR_DATA - Data error 214*4882a593Smuzhiyun SZ_ERROR_MEM - Memory allocation error 215*4882a593Smuzhiyun SZ_ERROR_UNSUPPORTED - Unsupported properties 216*4882a593Smuzhiyun SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). 217*4882a593Smuzhiyun */ 218*4882a593Smuzhiyun 219*4882a593Smuzhiyun SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, 220*4882a593Smuzhiyun const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, 221*4882a593Smuzhiyun ELzmaStatus *status, ISzAlloc *alloc); 222*4882a593Smuzhiyun 223*4882a593Smuzhiyun #endif 224