From 13e1fa055b3d778f3066e16fc94012af7da7a4fb Mon Sep 17 00:00:00 2001 From: Connor McLaughlin Date: Thu, 30 Jan 2020 15:47:11 +1000 Subject: [PATCH] dep: Add lzma 19.00 --- dep/CMakeLists.txt | 1 + dep/lzma/CMakeLists.txt | 34 + dep/lzma/LICENSE | 3 + dep/lzma/include/7zTypes.h | 375 +++++ dep/lzma/include/Alloc.h | 51 + dep/lzma/include/Bra.h | 64 + dep/lzma/include/Compiler.h | 33 + dep/lzma/include/CpuArch.h | 336 ++++ dep/lzma/include/Delta.h | 19 + dep/lzma/include/LzFind.h | 121 ++ dep/lzma/include/LzHash.h | 57 + dep/lzma/include/Lzma86.h | 111 ++ dep/lzma/include/LzmaDec.h | 234 +++ dep/lzma/include/LzmaEnc.h | 76 + dep/lzma/include/LzmaLib.h | 131 ++ dep/lzma/include/Precomp.h | 10 + dep/lzma/include/Sort.h | 18 + dep/lzma/lzma-history.txt | 446 +++++ dep/lzma/lzma.txt | 328 ++++ dep/lzma/lzma.vcxproj | 361 ++++ dep/lzma/lzma.vcxproj.filters | 17 + dep/lzma/src/Alloc.c | 455 +++++ dep/lzma/src/Bra86.c | 82 + dep/lzma/src/BraIA64.c | 53 + dep/lzma/src/CpuArch.c | 218 +++ dep/lzma/src/Delta.c | 64 + dep/lzma/src/LzFind.c | 1127 +++++++++++++ dep/lzma/src/Lzma86Dec.c | 54 + dep/lzma/src/Lzma86Enc.c | 106 ++ dep/lzma/src/LzmaDec.c | 1185 +++++++++++++ dep/lzma/src/LzmaEnc.c | 2976 +++++++++++++++++++++++++++++++++ dep/lzma/src/LzmaLib.c | 40 + dep/lzma/src/Sort.c | 141 ++ duckstation.sln | 19 + 34 files changed, 9346 insertions(+) create mode 100644 dep/lzma/CMakeLists.txt create mode 100644 dep/lzma/LICENSE create mode 100644 dep/lzma/include/7zTypes.h create mode 100644 dep/lzma/include/Alloc.h create mode 100644 dep/lzma/include/Bra.h create mode 100644 dep/lzma/include/Compiler.h create mode 100644 dep/lzma/include/CpuArch.h create mode 100644 dep/lzma/include/Delta.h create mode 100644 dep/lzma/include/LzFind.h create mode 100644 dep/lzma/include/LzHash.h create mode 100644 dep/lzma/include/Lzma86.h create mode 100644 dep/lzma/include/LzmaDec.h create mode 100644 dep/lzma/include/LzmaEnc.h create mode 100644 dep/lzma/include/LzmaLib.h create mode 100644 dep/lzma/include/Precomp.h create mode 100644 dep/lzma/include/Sort.h create mode 100644 dep/lzma/lzma-history.txt create mode 100644 dep/lzma/lzma.txt create mode 100644 dep/lzma/lzma.vcxproj create mode 100644 dep/lzma/lzma.vcxproj.filters create mode 100644 dep/lzma/src/Alloc.c create mode 100644 dep/lzma/src/Bra86.c create mode 100644 dep/lzma/src/BraIA64.c create mode 100644 dep/lzma/src/CpuArch.c create mode 100644 dep/lzma/src/Delta.c create mode 100644 dep/lzma/src/LzFind.c create mode 100644 dep/lzma/src/Lzma86Dec.c create mode 100644 dep/lzma/src/Lzma86Enc.c create mode 100644 dep/lzma/src/LzmaDec.c create mode 100644 dep/lzma/src/LzmaEnc.c create mode 100644 dep/lzma/src/LzmaLib.c create mode 100644 dep/lzma/src/Sort.c diff --git a/dep/CMakeLists.txt b/dep/CMakeLists.txt index a2d7547c0..0368ff18b 100644 --- a/dep/CMakeLists.txt +++ b/dep/CMakeLists.txt @@ -7,6 +7,7 @@ add_subdirectory(stb) add_subdirectory(tinyxml2) add_subdirectory(zlib) add_subdirectory(minizip) +add_subdirectory(lzma) if(NOT ANDROID) add_subdirectory(nativefiledialog) diff --git a/dep/lzma/CMakeLists.txt b/dep/lzma/CMakeLists.txt new file mode 100644 index 000000000..adbdc311b --- /dev/null +++ b/dep/lzma/CMakeLists.txt @@ -0,0 +1,34 @@ +add_library(lzma + include/7zTypes.h + include/Alloc.h + include/Bra.h + include/Compiler.h + include/CpuArch.h + include/Delta.h + include/LzFind.h + include/LzHash.h + include/Lzma86.h + include/LzmaDec.h + include/LzmaEnc.h + include/LzmaLib.h + include/Precomp.h + include/Sort.h + src/Alloc.c + src/Bra86.c + src/BraIA64.c + src/CpuArch.c + src/Delta.c + src/LzFind.c + src/Lzma86Dec.c + src/Lzma86Enc.c + src/LzmaDec.c + src/LzmaEnc.c + src/LzmaLib.c + src/Sort.c +) + +target_compile_definitions(lzma PRIVATE _7ZIP_ST) + +target_include_directories(lzma PRIVATE "${CMAKE_CURRENT_SOURCE_DIR}/include") +target_include_directories(lzma INTERFACE "${CMAKE_CURRENT_SOURCE_DIR}/include") + diff --git a/dep/lzma/LICENSE b/dep/lzma/LICENSE new file mode 100644 index 000000000..5f5705169 --- /dev/null +++ b/dep/lzma/LICENSE @@ -0,0 +1,3 @@ +LZMA SDK is placed in the public domain. + +Anyone is free to copy, modify, publish, use, compile, sell, or distribute the original LZMA SDK code, either in source code form or as a compiled binary, for any purpose, commercial or non-commercial, and by any means. \ No newline at end of file diff --git a/dep/lzma/include/7zTypes.h b/dep/lzma/include/7zTypes.h new file mode 100644 index 000000000..65b3af63c --- /dev/null +++ b/dep/lzma/include/7zTypes.h @@ -0,0 +1,375 @@ +/* 7zTypes.h -- Basic types +2018-08-04 : Igor Pavlov : Public domain */ + +#ifndef __7Z_TYPES_H +#define __7Z_TYPES_H + +#ifdef _WIN32 +/* #include */ +#endif + +#include + +#ifndef EXTERN_C_BEGIN +#ifdef __cplusplus +#define EXTERN_C_BEGIN extern "C" { +#define EXTERN_C_END } +#else +#define EXTERN_C_BEGIN +#define EXTERN_C_END +#endif +#endif + +EXTERN_C_BEGIN + +#define SZ_OK 0 + +#define SZ_ERROR_DATA 1 +#define SZ_ERROR_MEM 2 +#define SZ_ERROR_CRC 3 +#define SZ_ERROR_UNSUPPORTED 4 +#define SZ_ERROR_PARAM 5 +#define SZ_ERROR_INPUT_EOF 6 +#define SZ_ERROR_OUTPUT_EOF 7 +#define SZ_ERROR_READ 8 +#define SZ_ERROR_WRITE 9 +#define SZ_ERROR_PROGRESS 10 +#define SZ_ERROR_FAIL 11 +#define SZ_ERROR_THREAD 12 + +#define SZ_ERROR_ARCHIVE 16 +#define SZ_ERROR_NO_ARCHIVE 17 + +typedef int SRes; + + +#ifdef _WIN32 + +/* typedef DWORD WRes; */ +typedef unsigned WRes; +#define MY_SRes_HRESULT_FROM_WRes(x) HRESULT_FROM_WIN32(x) + +#else + +typedef int WRes; +#define MY__FACILITY_WIN32 7 +#define MY__FACILITY__WRes MY__FACILITY_WIN32 +#define MY_SRes_HRESULT_FROM_WRes(x) ((HRESULT)(x) <= 0 ? ((HRESULT)(x)) : ((HRESULT) (((x) & 0x0000FFFF) | (MY__FACILITY__WRes << 16) | 0x80000000))) + +#endif + + +#ifndef RINOK +#define RINOK(x) { int __result__ = (x); if (__result__ != 0) return __result__; } +#endif + +typedef unsigned char Byte; +typedef short Int16; +typedef unsigned short UInt16; + +#ifdef _LZMA_UINT32_IS_ULONG +typedef long Int32; +typedef unsigned long UInt32; +#else +typedef int Int32; +typedef unsigned int UInt32; +#endif + +#ifdef _SZ_NO_INT_64 + +/* define _SZ_NO_INT_64, if your compiler doesn't support 64-bit integers. + NOTES: Some code will work incorrectly in that case! */ + +typedef long Int64; +typedef unsigned long UInt64; + +#else + +#if defined(_MSC_VER) || defined(__BORLANDC__) +typedef __int64 Int64; +typedef unsigned __int64 UInt64; +#define UINT64_CONST(n) n +#else +typedef long long int Int64; +typedef unsigned long long int UInt64; +#define UINT64_CONST(n) n ## ULL +#endif + +#endif + +#ifdef _LZMA_NO_SYSTEM_SIZE_T +typedef UInt32 SizeT; +#else +typedef size_t SizeT; +#endif + +typedef int BoolInt; +/* typedef BoolInt Bool; */ +#define True 1 +#define False 0 + + +#ifdef _WIN32 +#define MY_STD_CALL __stdcall +#else +#define MY_STD_CALL +#endif + +#ifdef _MSC_VER + +#if _MSC_VER >= 1300 +#define MY_NO_INLINE __declspec(noinline) +#else +#define MY_NO_INLINE +#endif + +#define MY_FORCE_INLINE __forceinline + +#define MY_CDECL __cdecl +#define MY_FAST_CALL __fastcall + +#else + +#define MY_NO_INLINE +#define MY_FORCE_INLINE +#define MY_CDECL +#define MY_FAST_CALL + +/* inline keyword : for C++ / C99 */ + +/* GCC, clang: */ +/* +#if defined (__GNUC__) && (__GNUC__ >= 4) +#define MY_FORCE_INLINE __attribute__((always_inline)) +#define MY_NO_INLINE __attribute__((noinline)) +#endif +*/ + +#endif + + +/* The following interfaces use first parameter as pointer to structure */ + +typedef struct IByteIn IByteIn; +struct IByteIn +{ + Byte (*Read)(const IByteIn *p); /* reads one byte, returns 0 in case of EOF or error */ +}; +#define IByteIn_Read(p) (p)->Read(p) + + +typedef struct IByteOut IByteOut; +struct IByteOut +{ + void (*Write)(const IByteOut *p, Byte b); +}; +#define IByteOut_Write(p, b) (p)->Write(p, b) + + +typedef struct ISeqInStream ISeqInStream; +struct ISeqInStream +{ + SRes (*Read)(const ISeqInStream *p, void *buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) < input(*size)) is allowed */ +}; +#define ISeqInStream_Read(p, buf, size) (p)->Read(p, buf, size) + +/* it can return SZ_ERROR_INPUT_EOF */ +SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size); +SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType); +SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf); + + +typedef struct ISeqOutStream ISeqOutStream; +struct ISeqOutStream +{ + size_t (*Write)(const ISeqOutStream *p, const void *buf, size_t size); + /* Returns: result - the number of actually written bytes. + (result < size) means error */ +}; +#define ISeqOutStream_Write(p, buf, size) (p)->Write(p, buf, size) + +typedef enum +{ + SZ_SEEK_SET = 0, + SZ_SEEK_CUR = 1, + SZ_SEEK_END = 2 +} ESzSeek; + + +typedef struct ISeekInStream ISeekInStream; +struct ISeekInStream +{ + SRes (*Read)(const ISeekInStream *p, void *buf, size_t *size); /* same as ISeqInStream::Read */ + SRes (*Seek)(const ISeekInStream *p, Int64 *pos, ESzSeek origin); +}; +#define ISeekInStream_Read(p, buf, size) (p)->Read(p, buf, size) +#define ISeekInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) + + +typedef struct ILookInStream ILookInStream; +struct ILookInStream +{ + SRes (*Look)(const ILookInStream *p, const void **buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) > input(*size)) is not allowed + (output(*size) < input(*size)) is allowed */ + SRes (*Skip)(const ILookInStream *p, size_t offset); + /* offset must be <= output(*size) of Look */ + + SRes (*Read)(const ILookInStream *p, void *buf, size_t *size); + /* reads directly (without buffer). It's same as ISeqInStream::Read */ + SRes (*Seek)(const ILookInStream *p, Int64 *pos, ESzSeek origin); +}; + +#define ILookInStream_Look(p, buf, size) (p)->Look(p, buf, size) +#define ILookInStream_Skip(p, offset) (p)->Skip(p, offset) +#define ILookInStream_Read(p, buf, size) (p)->Read(p, buf, size) +#define ILookInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) + + +SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size); +SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset); + +/* reads via ILookInStream::Read */ +SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType); +SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size); + + + +typedef struct +{ + ILookInStream vt; + const ISeekInStream *realStream; + + size_t pos; + size_t size; /* it's data size */ + + /* the following variables must be set outside */ + Byte *buf; + size_t bufSize; +} CLookToRead2; + +void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead); + +#define LookToRead2_Init(p) { (p)->pos = (p)->size = 0; } + + +typedef struct +{ + ISeqInStream vt; + const ILookInStream *realStream; +} CSecToLook; + +void SecToLook_CreateVTable(CSecToLook *p); + + + +typedef struct +{ + ISeqInStream vt; + const ILookInStream *realStream; +} CSecToRead; + +void SecToRead_CreateVTable(CSecToRead *p); + + +typedef struct ICompressProgress ICompressProgress; + +struct ICompressProgress +{ + SRes (*Progress)(const ICompressProgress *p, UInt64 inSize, UInt64 outSize); + /* Returns: result. (result != SZ_OK) means break. + Value (UInt64)(Int64)-1 for size means unknown value. */ +}; +#define ICompressProgress_Progress(p, inSize, outSize) (p)->Progress(p, inSize, outSize) + + + +typedef struct ISzAlloc ISzAlloc; +typedef const ISzAlloc * ISzAllocPtr; + +struct ISzAlloc +{ + void *(*Alloc)(ISzAllocPtr p, size_t size); + void (*Free)(ISzAllocPtr p, void *address); /* address can be 0 */ +}; + +#define ISzAlloc_Alloc(p, size) (p)->Alloc(p, size) +#define ISzAlloc_Free(p, a) (p)->Free(p, a) + +/* deprecated */ +#define IAlloc_Alloc(p, size) ISzAlloc_Alloc(p, size) +#define IAlloc_Free(p, a) ISzAlloc_Free(p, a) + + + + + +#ifndef MY_offsetof + #ifdef offsetof + #define MY_offsetof(type, m) offsetof(type, m) + /* + #define MY_offsetof(type, m) FIELD_OFFSET(type, m) + */ + #else + #define MY_offsetof(type, m) ((size_t)&(((type *)0)->m)) + #endif +#endif + + + +#ifndef MY_container_of + +/* +#define MY_container_of(ptr, type, m) container_of(ptr, type, m) +#define MY_container_of(ptr, type, m) CONTAINING_RECORD(ptr, type, m) +#define MY_container_of(ptr, type, m) ((type *)((char *)(ptr) - offsetof(type, m))) +#define MY_container_of(ptr, type, m) (&((type *)0)->m == (ptr), ((type *)(((char *)(ptr)) - MY_offsetof(type, m)))) +*/ + +/* + GCC shows warning: "perhaps the 'offsetof' macro was used incorrectly" + GCC 3.4.4 : classes with constructor + GCC 4.8.1 : classes with non-public variable members" +*/ + +#define MY_container_of(ptr, type, m) ((type *)((char *)(1 ? (ptr) : &((type *)0)->m) - MY_offsetof(type, m))) + + +#endif + +#define CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) ((type *)(ptr)) + +/* +#define CONTAINER_FROM_VTBL(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) +*/ +#define CONTAINER_FROM_VTBL(ptr, type, m) MY_container_of(ptr, type, m) + +#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) +/* +#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL(ptr, type, m) +*/ + + + +#ifdef _WIN32 + +#define CHAR_PATH_SEPARATOR '\\' +#define WCHAR_PATH_SEPARATOR L'\\' +#define STRING_PATH_SEPARATOR "\\" +#define WSTRING_PATH_SEPARATOR L"\\" + +#else + +#define CHAR_PATH_SEPARATOR '/' +#define WCHAR_PATH_SEPARATOR L'/' +#define STRING_PATH_SEPARATOR "/" +#define WSTRING_PATH_SEPARATOR L"/" + +#endif + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/Alloc.h b/dep/lzma/include/Alloc.h new file mode 100644 index 000000000..648237646 --- /dev/null +++ b/dep/lzma/include/Alloc.h @@ -0,0 +1,51 @@ +/* Alloc.h -- Memory allocation functions +2018-02-19 : Igor Pavlov : Public domain */ + +#ifndef __COMMON_ALLOC_H +#define __COMMON_ALLOC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void *MyAlloc(size_t size); +void MyFree(void *address); + +#ifdef _WIN32 + +void SetLargePageSize(); + +void *MidAlloc(size_t size); +void MidFree(void *address); +void *BigAlloc(size_t size); +void BigFree(void *address); + +#else + +#define MidAlloc(size) MyAlloc(size) +#define MidFree(address) MyFree(address) +#define BigAlloc(size) MyAlloc(size) +#define BigFree(address) MyFree(address) + +#endif + +extern const ISzAlloc g_Alloc; +extern const ISzAlloc g_BigAlloc; +extern const ISzAlloc g_MidAlloc; +extern const ISzAlloc g_AlignedAlloc; + + +typedef struct +{ + ISzAlloc vt; + ISzAllocPtr baseAlloc; + unsigned numAlignBits; /* ((1 << numAlignBits) >= sizeof(void *)) */ + size_t offset; /* (offset == (k * sizeof(void *)) && offset < (1 << numAlignBits) */ +} CAlignOffsetAlloc; + +void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p); + + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/Bra.h b/dep/lzma/include/Bra.h new file mode 100644 index 000000000..855e37a6b --- /dev/null +++ b/dep/lzma/include/Bra.h @@ -0,0 +1,64 @@ +/* Bra.h -- Branch converters for executables +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __BRA_H +#define __BRA_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* +These functions convert relative addresses to absolute addresses +in CALL instructions to increase the compression ratio. + + In: + data - data buffer + size - size of data + ip - current virtual Instruction Pinter (IP) value + state - state variable for x86 converter + encoding - 0 (for decoding), 1 (for encoding) + + Out: + state - state variable for x86 converter + + Returns: + The number of processed bytes. If you call these functions with multiple calls, + you must start next call with first byte after block of processed bytes. + + Type Endian Alignment LookAhead + + x86 little 1 4 + ARMT little 2 2 + ARM little 4 0 + PPC big 4 0 + SPARC big 4 0 + IA64 little 16 0 + + size must be >= Alignment + LookAhead, if it's not last block. + If (size < Alignment + LookAhead), converter returns 0. + + Example: + + UInt32 ip = 0; + for () + { + ; size must be >= Alignment + LookAhead, if it's not last block + SizeT processed = Convert(data, size, ip, 1); + data += processed; + size -= processed; + ip += processed; + } +*/ + +#define x86_Convert_Init(state) { state = 0; } +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding); +SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/Compiler.h b/dep/lzma/include/Compiler.h new file mode 100644 index 000000000..0cc409d8a --- /dev/null +++ b/dep/lzma/include/Compiler.h @@ -0,0 +1,33 @@ +/* Compiler.h +2017-04-03 : Igor Pavlov : Public domain */ + +#ifndef __7Z_COMPILER_H +#define __7Z_COMPILER_H + +#ifdef _MSC_VER + + #ifdef UNDER_CE + #define RPC_NO_WINDOWS_H + /* #pragma warning(disable : 4115) // '_RPC_ASYNC_STATE' : named type definition in parentheses */ + #pragma warning(disable : 4201) // nonstandard extension used : nameless struct/union + #pragma warning(disable : 4214) // nonstandard extension used : bit field types other than int + #endif + + #if _MSC_VER >= 1300 + #pragma warning(disable : 4996) // This function or variable may be unsafe + #else + #pragma warning(disable : 4511) // copy constructor could not be generated + #pragma warning(disable : 4512) // assignment operator could not be generated + #pragma warning(disable : 4514) // unreferenced inline function has been removed + #pragma warning(disable : 4702) // unreachable code + #pragma warning(disable : 4710) // not inlined + #pragma warning(disable : 4714) // function marked as __forceinline not inlined + #pragma warning(disable : 4786) // identifier was truncated to '255' characters in the debug information + #endif + +#endif + +#define UNUSED_VAR(x) (void)x; +/* #define UNUSED_VAR(x) x=x; */ + +#endif diff --git a/dep/lzma/include/CpuArch.h b/dep/lzma/include/CpuArch.h new file mode 100644 index 000000000..bd4293880 --- /dev/null +++ b/dep/lzma/include/CpuArch.h @@ -0,0 +1,336 @@ +/* CpuArch.h -- CPU specific code +2018-02-18 : Igor Pavlov : Public domain */ + +#ifndef __CPU_ARCH_H +#define __CPU_ARCH_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* +MY_CPU_LE means that CPU is LITTLE ENDIAN. +MY_CPU_BE means that CPU is BIG ENDIAN. +If MY_CPU_LE and MY_CPU_BE are not defined, we don't know about ENDIANNESS of platform. + +MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned memory accesses. +*/ + +#if defined(_M_X64) \ + || defined(_M_AMD64) \ + || defined(__x86_64__) \ + || defined(__AMD64__) \ + || defined(__amd64__) + #define MY_CPU_AMD64 + #ifdef __ILP32__ + #define MY_CPU_NAME "x32" + #else + #define MY_CPU_NAME "x64" + #endif + #define MY_CPU_64BIT +#endif + + +#if defined(_M_IX86) \ + || defined(__i386__) + #define MY_CPU_X86 + #define MY_CPU_NAME "x86" + #define MY_CPU_32BIT +#endif + + +#if defined(_M_ARM64) \ + || defined(__AARCH64EL__) \ + || defined(__AARCH64EB__) \ + || defined(__aarch64__) + #define MY_CPU_ARM64 + #define MY_CPU_NAME "arm64" + #define MY_CPU_64BIT +#endif + + +#if defined(_M_ARM) \ + || defined(_M_ARM_NT) \ + || defined(_M_ARMT) \ + || defined(__arm__) \ + || defined(__thumb__) \ + || defined(__ARMEL__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEL__) \ + || defined(__THUMBEB__) + #define MY_CPU_ARM + #define MY_CPU_NAME "arm" + #define MY_CPU_32BIT +#endif + + +#if defined(_M_IA64) \ + || defined(__ia64__) + #define MY_CPU_IA64 + #define MY_CPU_NAME "ia64" + #define MY_CPU_64BIT +#endif + + +#if defined(__mips64) \ + || defined(__mips64__) \ + || (defined(__mips) && (__mips == 64 || __mips == 4 || __mips == 3)) + #define MY_CPU_NAME "mips64" + #define MY_CPU_64BIT +#elif defined(__mips__) + #define MY_CPU_NAME "mips" + /* #define MY_CPU_32BIT */ +#endif + + +#if defined(__ppc64__) \ + || defined(__powerpc64__) + #ifdef __ILP32__ + #define MY_CPU_NAME "ppc64-32" + #else + #define MY_CPU_NAME "ppc64" + #endif + #define MY_CPU_64BIT +#elif defined(__ppc__) \ + || defined(__powerpc__) + #define MY_CPU_NAME "ppc" + #define MY_CPU_32BIT +#endif + + +#if defined(__sparc64__) + #define MY_CPU_NAME "sparc64" + #define MY_CPU_64BIT +#elif defined(__sparc__) + #define MY_CPU_NAME "sparc" + /* #define MY_CPU_32BIT */ +#endif + + +#if defined(MY_CPU_X86) || defined(MY_CPU_AMD64) +#define MY_CPU_X86_OR_AMD64 +#endif + + +#ifdef _WIN32 + + #ifdef MY_CPU_ARM + #define MY_CPU_ARM_LE + #endif + + #ifdef MY_CPU_ARM64 + #define MY_CPU_ARM64_LE + #endif + + #ifdef _M_IA64 + #define MY_CPU_IA64_LE + #endif + +#endif + + +#if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM_LE) \ + || defined(MY_CPU_ARM64_LE) \ + || defined(MY_CPU_IA64_LE) \ + || defined(__LITTLE_ENDIAN__) \ + || defined(__ARMEL__) \ + || defined(__THUMBEL__) \ + || defined(__AARCH64EL__) \ + || defined(__MIPSEL__) \ + || defined(__MIPSEL) \ + || defined(_MIPSEL) \ + || defined(__BFIN__) \ + || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)) + #define MY_CPU_LE +#endif + +#if defined(__BIG_ENDIAN__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEB__) \ + || defined(__AARCH64EB__) \ + || defined(__MIPSEB__) \ + || defined(__MIPSEB) \ + || defined(_MIPSEB) \ + || defined(__m68k__) \ + || defined(__s390__) \ + || defined(__s390x__) \ + || defined(__zarch__) \ + || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)) + #define MY_CPU_BE +#endif + + +#if defined(MY_CPU_LE) && defined(MY_CPU_BE) + #error Stop_Compiling_Bad_Endian +#endif + + +#if defined(MY_CPU_32BIT) && defined(MY_CPU_64BIT) + #error Stop_Compiling_Bad_32_64_BIT +#endif + + +#ifndef MY_CPU_NAME + #ifdef MY_CPU_LE + #define MY_CPU_NAME "LE" + #elif defined(MY_CPU_BE) + #define MY_CPU_NAME "BE" + #else + /* + #define MY_CPU_NAME "" + */ + #endif +#endif + + + + + +#ifdef MY_CPU_LE + #if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM64) \ + || defined(__ARM_FEATURE_UNALIGNED) + #define MY_CPU_LE_UNALIGN + #endif +#endif + + +#ifdef MY_CPU_LE_UNALIGN + +#define GetUi16(p) (*(const UInt16 *)(const void *)(p)) +#define GetUi32(p) (*(const UInt32 *)(const void *)(p)) +#define GetUi64(p) (*(const UInt64 *)(const void *)(p)) + +#define SetUi16(p, v) { *(UInt16 *)(p) = (v); } +#define SetUi32(p, v) { *(UInt32 *)(p) = (v); } +#define SetUi64(p, v) { *(UInt64 *)(p) = (v); } + +#else + +#define GetUi16(p) ( (UInt16) ( \ + ((const Byte *)(p))[0] | \ + ((UInt16)((const Byte *)(p))[1] << 8) )) + +#define GetUi32(p) ( \ + ((const Byte *)(p))[0] | \ + ((UInt32)((const Byte *)(p))[1] << 8) | \ + ((UInt32)((const Byte *)(p))[2] << 16) | \ + ((UInt32)((const Byte *)(p))[3] << 24)) + +#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) + +#define SetUi16(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)_vvv_; \ + _ppp_[1] = (Byte)(_vvv_ >> 8); } + +#define SetUi32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)_vvv_; \ + _ppp_[1] = (Byte)(_vvv_ >> 8); \ + _ppp_[2] = (Byte)(_vvv_ >> 16); \ + _ppp_[3] = (Byte)(_vvv_ >> 24); } + +#define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \ + SetUi32(_ppp2_ , (UInt32)_vvv2_); \ + SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); } + +#endif + +#ifdef __has_builtin + #define MY__has_builtin(x) __has_builtin(x) +#else + #define MY__has_builtin(x) 0 +#endif + +#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ (_MSC_VER >= 1300) + +/* Note: we use bswap instruction, that is unsupported in 386 cpu */ + +#include + +#pragma intrinsic(_byteswap_ushort) +#pragma intrinsic(_byteswap_ulong) +#pragma intrinsic(_byteswap_uint64) + +/* #define GetBe16(p) _byteswap_ushort(*(const UInt16 *)(const Byte *)(p)) */ +#define GetBe32(p) _byteswap_ulong(*(const UInt32 *)(const Byte *)(p)) +#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const Byte *)(p)) + +#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = _byteswap_ulong(v) + +#elif defined(MY_CPU_LE_UNALIGN) && ( \ + (defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))) \ + || (defined(__clang__) && MY__has_builtin(__builtin_bswap16)) ) + +/* #define GetBe16(p) __builtin_bswap16(*(const UInt16 *)(const Byte *)(p)) */ +#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const Byte *)(p)) +#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const Byte *)(p)) + +#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = __builtin_bswap32(v) + +#else + +#define GetBe32(p) ( \ + ((UInt32)((const Byte *)(p))[0] << 24) | \ + ((UInt32)((const Byte *)(p))[1] << 16) | \ + ((UInt32)((const Byte *)(p))[2] << 8) | \ + ((const Byte *)(p))[3] ) + +#define GetBe64(p) (((UInt64)GetBe32(p) << 32) | GetBe32(((const Byte *)(p)) + 4)) + +#define SetBe32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)(_vvv_ >> 24); \ + _ppp_[1] = (Byte)(_vvv_ >> 16); \ + _ppp_[2] = (Byte)(_vvv_ >> 8); \ + _ppp_[3] = (Byte)_vvv_; } + +#endif + + +#ifndef GetBe16 + +#define GetBe16(p) ( (UInt16) ( \ + ((UInt16)((const Byte *)(p))[0] << 8) | \ + ((const Byte *)(p))[1] )) + +#endif + + + +#ifdef MY_CPU_X86_OR_AMD64 + +typedef struct +{ + UInt32 maxFunc; + UInt32 vendor[3]; + UInt32 ver; + UInt32 b; + UInt32 c; + UInt32 d; +} Cx86cpuid; + +enum +{ + CPU_FIRM_INTEL, + CPU_FIRM_AMD, + CPU_FIRM_VIA +}; + +void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d); + +BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p); +int x86cpuid_GetFirm(const Cx86cpuid *p); + +#define x86cpuid_GetFamily(ver) (((ver >> 16) & 0xFF0) | ((ver >> 8) & 0xF)) +#define x86cpuid_GetModel(ver) (((ver >> 12) & 0xF0) | ((ver >> 4) & 0xF)) +#define x86cpuid_GetStepping(ver) (ver & 0xF) + +BoolInt CPU_Is_InOrder(); +BoolInt CPU_Is_Aes_Supported(); +BoolInt CPU_IsSupported_PageGB(); + +#endif + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/Delta.h b/dep/lzma/include/Delta.h new file mode 100644 index 000000000..2fa54ad67 --- /dev/null +++ b/dep/lzma/include/Delta.h @@ -0,0 +1,19 @@ +/* Delta.h -- Delta converter +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __DELTA_H +#define __DELTA_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define DELTA_STATE_SIZE 256 + +void Delta_Init(Byte *state); +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size); +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/LzFind.h b/dep/lzma/include/LzFind.h new file mode 100644 index 000000000..42c13be15 --- /dev/null +++ b/dep/lzma/include/LzFind.h @@ -0,0 +1,121 @@ +/* LzFind.h -- Match finder for LZ algorithms +2017-06-10 : Igor Pavlov : Public domain */ + +#ifndef __LZ_FIND_H +#define __LZ_FIND_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef UInt32 CLzRef; + +typedef struct _CMatchFinder +{ + Byte *buffer; + UInt32 pos; + UInt32 posLimit; + UInt32 streamPos; + UInt32 lenLimit; + + UInt32 cyclicBufferPos; + UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ + + Byte streamEndWasReached; + Byte btMode; + Byte bigHash; + Byte directInput; + + UInt32 matchMaxLen; + CLzRef *hash; + CLzRef *son; + UInt32 hashMask; + UInt32 cutValue; + + Byte *bufferBase; + ISeqInStream *stream; + + UInt32 blockSize; + UInt32 keepSizeBefore; + UInt32 keepSizeAfter; + + UInt32 numHashBytes; + size_t directInputRem; + UInt32 historySize; + UInt32 fixedHashSize; + UInt32 hashSizeSum; + SRes result; + UInt32 crc[256]; + size_t numRefs; + + UInt64 expectedDataSize; +} CMatchFinder; + +#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((p)->buffer) + +#define Inline_MatchFinder_GetNumAvailableBytes(p) ((p)->streamPos - (p)->pos) + +#define Inline_MatchFinder_IsFinishedOK(p) \ + ((p)->streamEndWasReached \ + && (p)->streamPos == (p)->pos \ + && (!(p)->directInput || (p)->directInputRem == 0)) + +int MatchFinder_NeedMove(CMatchFinder *p); +Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); +void MatchFinder_MoveBlock(CMatchFinder *p); +void MatchFinder_ReadIfRequired(CMatchFinder *p); + +void MatchFinder_Construct(CMatchFinder *p); + +/* Conditions: + historySize <= 3 GB + keepAddBufferBefore + matchMaxLen + keepAddBufferAfter < 511MB +*/ +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAllocPtr alloc); +void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc); +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems); +void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); + +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, + UInt32 *distances, UInt32 maxLen); + +/* +Conditions: + Mf_GetNumAvailableBytes_Func must be called before each Mf_GetMatchLen_Func. + Mf_GetPointerToCurrentPos_Func's result must be used only before any other function +*/ + +typedef void (*Mf_Init_Func)(void *object); +typedef UInt32 (*Mf_GetNumAvailableBytes_Func)(void *object); +typedef const Byte * (*Mf_GetPointerToCurrentPos_Func)(void *object); +typedef UInt32 (*Mf_GetMatches_Func)(void *object, UInt32 *distances); +typedef void (*Mf_Skip_Func)(void *object, UInt32); + +typedef struct _IMatchFinder +{ + Mf_Init_Func Init; + Mf_GetNumAvailableBytes_Func GetNumAvailableBytes; + Mf_GetPointerToCurrentPos_Func GetPointerToCurrentPos; + Mf_GetMatches_Func GetMatches; + Mf_Skip_Func Skip; +} IMatchFinder; + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable); + +void MatchFinder_Init_LowHash(CMatchFinder *p); +void MatchFinder_Init_HighHash(CMatchFinder *p); +void MatchFinder_Init_3(CMatchFinder *p, int readData); +void MatchFinder_Init(CMatchFinder *p); + +UInt32 Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); +UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/LzHash.h b/dep/lzma/include/LzHash.h new file mode 100644 index 000000000..e7c942303 --- /dev/null +++ b/dep/lzma/include/LzHash.h @@ -0,0 +1,57 @@ +/* LzHash.h -- HASH functions for LZ algorithms +2015-04-12 : Igor Pavlov : Public domain */ + +#ifndef __LZ_HASH_H +#define __LZ_HASH_H + +#define kHash2Size (1 << 10) +#define kHash3Size (1 << 16) +#define kHash4Size (1 << 20) + +#define kFix3HashSize (kHash2Size) +#define kFix4HashSize (kHash2Size + kHash3Size) +#define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) + +#define HASH2_CALC hv = cur[0] | ((UInt32)cur[1] << 8); + +#define HASH3_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + hv = (temp ^ ((UInt32)cur[2] << 8)) & p->hashMask; } + +#define HASH4_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + hv = (temp ^ (p->crc[cur[3]] << 5)) & p->hashMask; } + +#define HASH5_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + temp ^= (p->crc[cur[3]] << 5); \ + h4 = temp & (kHash4Size - 1); \ + hv = (temp ^ (p->crc[cur[4]] << 3)) & p->hashMask; } + +/* #define HASH_ZIP_CALC hv = ((cur[0] | ((UInt32)cur[1] << 8)) ^ p->crc[cur[2]]) & 0xFFFF; */ +#define HASH_ZIP_CALC hv = ((cur[2] | ((UInt32)cur[0] << 8)) ^ p->crc[cur[1]]) & 0xFFFF; + + +#define MT_HASH2_CALC \ + h2 = (p->crc[cur[0]] ^ cur[1]) & (kHash2Size - 1); + +#define MT_HASH3_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } + +#define MT_HASH4_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + h4 = (temp ^ (p->crc[cur[3]] << 5)) & (kHash4Size - 1); } + +#endif diff --git a/dep/lzma/include/Lzma86.h b/dep/lzma/include/Lzma86.h new file mode 100644 index 000000000..bebed5cb7 --- /dev/null +++ b/dep/lzma/include/Lzma86.h @@ -0,0 +1,111 @@ +/* Lzma86.h -- LZMA + x86 (BCJ) Filter +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __LZMA86_H +#define __LZMA86_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define LZMA86_SIZE_OFFSET (1 + 5) +#define LZMA86_HEADER_SIZE (LZMA86_SIZE_OFFSET + 8) + +/* +It's an example for LZMA + x86 Filter use. +You can use .lzma86 extension, if you write that stream to file. +.lzma86 header adds one additional byte to standard .lzma header. +.lzma86 header (14 bytes): + Offset Size Description + 0 1 = 0 - no filter, pure LZMA + = 1 - x86 filter + LZMA + 1 1 lc, lp and pb in encoded form + 2 4 dictSize (little endian) + 6 8 uncompressed size (little endian) + + +Lzma86_Encode +------------- +level - compression level: 0 <= level <= 9, the default value for "level" is 5. + +dictSize - The dictionary size in bytes. The maximum value is + 128 MB = (1 << 27) bytes for 32-bit version + 1 GB = (1 << 30) bytes for 64-bit version + The default value is 16 MB = (1 << 24) bytes, for level = 5. + It's recommended to use the dictionary that is larger than 4 KB and + that can be calculated as (1 << N) or (3 << N) sizes. + For better compression ratio dictSize must be >= inSize. + +filterMode: + SZ_FILTER_NO - no Filter + SZ_FILTER_YES - x86 Filter + SZ_FILTER_AUTO - it tries both alternatives to select best. + Encoder will use 2 or 3 passes: + 2 passes when FILTER_NO provides better compression. + 3 passes when FILTER_YES provides better compression. + +Lzma86Encode allocates Data with MyAlloc functions. +RAM Requirements for compressing: + RamSize = dictionarySize * 11.5 + 6MB + FilterBlockSize + filterMode FilterBlockSize + SZ_FILTER_NO 0 + SZ_FILTER_YES inSize + SZ_FILTER_AUTO inSize + + +Return code: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +enum ESzFilterMode +{ + SZ_FILTER_NO, + SZ_FILTER_YES, + SZ_FILTER_AUTO +}; + +SRes Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, + int level, UInt32 dictSize, int filterMode); + + +/* +Lzma86_GetUnpackSize: + In: + src - input data + srcLen - input data size + Out: + unpackSize - size of uncompressed stream + Return code: + SZ_OK - OK + SZ_ERROR_INPUT_EOF - Error in headers +*/ + +SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize); + +/* +Lzma86_Decode: + In: + dest - output data + destLen - output data size + src - input data + srcLen - input data size + Out: + destLen - processed output size + srcLen - processed input size + Return code: + SZ_OK - OK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - unsupported file + SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer +*/ + +SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/LzmaDec.h b/dep/lzma/include/LzmaDec.h new file mode 100644 index 000000000..1f0927ab1 --- /dev/null +++ b/dep/lzma/include/LzmaDec.h @@ -0,0 +1,234 @@ +/* LzmaDec.h -- LZMA Decoder +2018-04-21 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_DEC_H +#define __LZMA_DEC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* #define _LZMA_PROB32 */ +/* _LZMA_PROB32 can increase the speed on some CPUs, + but memory usage for CLzmaDec::probs will be doubled in that case */ + +typedef +#ifdef _LZMA_PROB32 + UInt32 +#else + UInt16 +#endif + CLzmaProb; + + +/* ---------- LZMA Properties ---------- */ + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaProps +{ + Byte lc; + Byte lp; + Byte pb; + Byte _pad_; + UInt32 dicSize; +} CLzmaProps; + +/* LzmaProps_Decode - decodes properties +Returns: + SZ_OK + SZ_ERROR_UNSUPPORTED - Unsupported properties +*/ + +SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size); + + +/* ---------- LZMA Decoder state ---------- */ + +/* LZMA_REQUIRED_INPUT_MAX = number of required input bytes for worst case. + Num bits = log2((2^11 / 31) ^ 22) + 26 < 134 + 26 = 160; */ + +#define LZMA_REQUIRED_INPUT_MAX 20 + +typedef struct +{ + /* Don't change this structure. ASM code can use it. */ + CLzmaProps prop; + CLzmaProb *probs; + CLzmaProb *probs_1664; + Byte *dic; + SizeT dicBufSize; + SizeT dicPos; + const Byte *buf; + UInt32 range; + UInt32 code; + UInt32 processedPos; + UInt32 checkDicSize; + UInt32 reps[4]; + UInt32 state; + UInt32 remainLen; + + UInt32 numProbs; + unsigned tempBufSize; + Byte tempBuf[LZMA_REQUIRED_INPUT_MAX]; +} CLzmaDec; + +#define LzmaDec_Construct(p) { (p)->dic = NULL; (p)->probs = NULL; } + +void LzmaDec_Init(CLzmaDec *p); + +/* There are two types of LZMA streams: + - Stream with end mark. That end mark adds about 6 bytes to compressed size. + - Stream without end mark. You must know exact uncompressed size to decompress such stream. */ + +typedef enum +{ + LZMA_FINISH_ANY, /* finish at any point */ + LZMA_FINISH_END /* block must be finished at the end */ +} ELzmaFinishMode; + +/* ELzmaFinishMode has meaning only if the decoding reaches output limit !!! + + You must use LZMA_FINISH_END, when you know that current output buffer + covers last bytes of block. In other cases you must use LZMA_FINISH_ANY. + + If LZMA decoder sees end marker before reaching output limit, it returns SZ_OK, + and output value of destLen will be less than output buffer size limit. + You can check status result also. + + You can use multiple checks to test data integrity after full decompression: + 1) Check Result and "status" variable. + 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. + 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. + You must use correct finish mode in that case. */ + +typedef enum +{ + LZMA_STATUS_NOT_SPECIFIED, /* use main error code instead */ + LZMA_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ + LZMA_STATUS_NOT_FINISHED, /* stream was not finished */ + LZMA_STATUS_NEEDS_MORE_INPUT, /* you must provide more input bytes */ + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK /* there is probability that stream was finished without end mark */ +} ELzmaStatus; + +/* ELzmaStatus is used only as output value for function call */ + + +/* ---------- Interfaces ---------- */ + +/* There are 3 levels of interfaces: + 1) Dictionary Interface + 2) Buffer Interface + 3) One Call Interface + You can select any of these interfaces, but don't mix functions from different + groups for same object. */ + + +/* There are two variants to allocate state for Dictionary Interface: + 1) LzmaDec_Allocate / LzmaDec_Free + 2) LzmaDec_AllocateProbs / LzmaDec_FreeProbs + You can use variant 2, if you set dictionary buffer manually. + For Buffer Interface you must always use variant 1. + +LzmaDec_Allocate* can return: + SZ_OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties +*/ + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc); + +SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); +void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc); + +/* ---------- Dictionary Interface ---------- */ + +/* You can use it, if you want to eliminate the overhead for data copying from + dictionary to some other external buffer. + You must work with CLzmaDec variables directly in this interface. + + STEPS: + LzmaDec_Construct() + LzmaDec_Allocate() + for (each new stream) + { + LzmaDec_Init() + while (it needs more decompression) + { + LzmaDec_DecodeToDic() + use data from CLzmaDec::dic and update CLzmaDec::dicPos + } + } + LzmaDec_Free() +*/ + +/* LzmaDec_DecodeToDic + + The decoding to internal dictionary buffer (CLzmaDec::dic). + You must manually update CLzmaDec::dicPos, if it reaches CLzmaDec::dicBufSize !!! + +finishMode: + It has meaning only if the decoding reaches output limit (dicLimit). + LZMA_FINISH_ANY - Decode just dicLimit bytes. + LZMA_FINISH_END - Stream must be finished after dicLimit. + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_NEEDS_MORE_INPUT + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error +*/ + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- Buffer Interface ---------- */ + +/* It's zlib-like interface. + See LzmaDec_DecodeToDic description for information about STEPS and return results, + but you must use LzmaDec_DecodeToBuf instead of LzmaDec_DecodeToDic and you don't need + to work with CLzmaDec variables manually. + +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). +*/ + +SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- One Call Interface ---------- */ + +/* LzmaDecode + +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). +*/ + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/LzmaEnc.h b/dep/lzma/include/LzmaEnc.h new file mode 100644 index 000000000..9194ee576 --- /dev/null +++ b/dep/lzma/include/LzmaEnc.h @@ -0,0 +1,76 @@ +/* LzmaEnc.h -- LZMA Encoder +2017-07-27 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_ENC_H +#define __LZMA_ENC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaEncProps +{ + int level; /* 0 <= level <= 9 */ + UInt32 dictSize; /* (1 << 12) <= dictSize <= (1 << 27) for 32-bit version + (1 << 12) <= dictSize <= (3 << 29) for 64-bit version + default = (1 << 24) */ + int lc; /* 0 <= lc <= 8, default = 3 */ + int lp; /* 0 <= lp <= 4, default = 0 */ + int pb; /* 0 <= pb <= 4, default = 2 */ + int algo; /* 0 - fast, 1 - normal, default = 1 */ + int fb; /* 5 <= fb <= 273, default = 32 */ + int btMode; /* 0 - hashChain Mode, 1 - binTree mode - normal, default = 1 */ + int numHashBytes; /* 2, 3 or 4, default = 4 */ + UInt32 mc; /* 1 <= mc <= (1 << 30), default = 32 */ + unsigned writeEndMark; /* 0 - do not write EOPM, 1 - write EOPM, default = 0 */ + int numThreads; /* 1 or 2, default = 2 */ + + UInt64 reduceSize; /* estimated size of data that will be compressed. default = (UInt64)(Int64)-1. + Encoder uses this value to reduce dictionary size */ +} CLzmaEncProps; + +void LzmaEncProps_Init(CLzmaEncProps *p); +void LzmaEncProps_Normalize(CLzmaEncProps *p); +UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2); + + +/* ---------- CLzmaEncHandle Interface ---------- */ + +/* LzmaEnc* functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzmaEncHandle; + +CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc); +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig); + +SRes LzmaEnc_SetProps(CLzmaEncHandle p, const CLzmaEncProps *props); +void LzmaEnc_SetDataSize(CLzmaEncHandle p, UInt64 expectedDataSiize); +SRes LzmaEnc_WriteProperties(CLzmaEncHandle p, Byte *properties, SizeT *size); +unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle p); + +SRes LzmaEnc_Encode(CLzmaEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemEncode(CLzmaEncHandle p, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); + + +/* ---------- One Call Interface ---------- */ + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/LzmaLib.h b/dep/lzma/include/LzmaLib.h new file mode 100644 index 000000000..88fa87d35 --- /dev/null +++ b/dep/lzma/include/LzmaLib.h @@ -0,0 +1,131 @@ +/* LzmaLib.h -- LZMA library interface +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_LIB_H +#define __LZMA_LIB_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define MY_STDAPI int MY_STD_CALL + +#define LZMA_PROPS_SIZE 5 + +/* +RAM requirements for LZMA: + for compression: (dictSize * 11.5 + 6 MB) + state_size + for decompression: dictSize + state_size + state_size = (4 + (1.5 << (lc + lp))) KB + by default (lc=3, lp=0), state_size = 16 KB. + +LZMA properties (5 bytes) format + Offset Size Description + 0 1 lc, lp and pb in encoded form. + 1 4 dictSize (little endian). +*/ + +/* +LzmaCompress +------------ + +outPropsSize - + In: the pointer to the size of outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. + Out: the pointer to the size of written properties in outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. + + LZMA Encoder will use defult values for any parameter, if it is + -1 for any from: level, loc, lp, pb, fb, numThreads + 0 for dictSize + +level - compression level: 0 <= level <= 9; + + level dictSize algo fb + 0: 16 KB 0 32 + 1: 64 KB 0 32 + 2: 256 KB 0 32 + 3: 1 MB 0 32 + 4: 4 MB 0 32 + 5: 16 MB 1 32 + 6: 32 MB 1 32 + 7+: 64 MB 1 64 + + The default value for "level" is 5. + + algo = 0 means fast method + algo = 1 means normal method + +dictSize - The dictionary size in bytes. The maximum value is + 128 MB = (1 << 27) bytes for 32-bit version + 1 GB = (1 << 30) bytes for 64-bit version + The default value is 16 MB = (1 << 24) bytes. + It's recommended to use the dictionary that is larger than 4 KB and + that can be calculated as (1 << N) or (3 << N) sizes. + +lc - The number of literal context bits (high bits of previous literal). + It can be in the range from 0 to 8. The default value is 3. + Sometimes lc=4 gives the gain for big files. + +lp - The number of literal pos bits (low bits of current position for literals). + It can be in the range from 0 to 4. The default value is 0. + The lp switch is intended for periodical data when the period is equal to 2^lp. + For example, for 32-bit (4 bytes) periodical data you can use lp=2. Often it's + better to set lc=0, if you change lp switch. + +pb - The number of pos bits (low bits of current position). + It can be in the range from 0 to 4. The default value is 2. + The pb switch is intended for periodical data when the period is equal 2^pb. + +fb - Word size (the number of fast bytes). + It can be in the range from 5 to 273. The default value is 32. + Usually, a big number gives a little bit better compression ratio and + slower compression process. + +numThreads - The number of thereads. 1 or 2. The default value is 2. + Fast mode (algo = 0) can use only 1 thread. + +Out: + destLen - processed output size +Returns: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, + unsigned char *outProps, size_t *outPropsSize, /* *outPropsSize must be = 5 */ + int level, /* 0 <= level <= 9, default = 5 */ + unsigned dictSize, /* default = (1 << 24) */ + int lc, /* 0 <= lc <= 8, default = 3 */ + int lp, /* 0 <= lp <= 4, default = 0 */ + int pb, /* 0 <= pb <= 4, default = 2 */ + int fb, /* 5 <= fb <= 273, default = 32 */ + int numThreads /* 1 or 2, default = 2 */ + ); + +/* +LzmaUncompress +-------------- +In: + dest - output data + destLen - output data size + src - input data + srcLen - input data size +Out: + destLen - processed output size + srcLen - processed input size +Returns: + SZ_OK - OK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation arror + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer (src) +*/ + +MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, SizeT *srcLen, + const unsigned char *props, size_t propsSize); + +EXTERN_C_END + +#endif diff --git a/dep/lzma/include/Precomp.h b/dep/lzma/include/Precomp.h new file mode 100644 index 000000000..e8ff8b40e --- /dev/null +++ b/dep/lzma/include/Precomp.h @@ -0,0 +1,10 @@ +/* Precomp.h -- StdAfx +2013-11-12 : Igor Pavlov : Public domain */ + +#ifndef __7Z_PRECOMP_H +#define __7Z_PRECOMP_H + +#include "Compiler.h" +/* #include "7zTypes.h" */ + +#endif diff --git a/dep/lzma/include/Sort.h b/dep/lzma/include/Sort.h new file mode 100644 index 000000000..2e2963a23 --- /dev/null +++ b/dep/lzma/include/Sort.h @@ -0,0 +1,18 @@ +/* Sort.h -- Sort functions +2014-04-05 : Igor Pavlov : Public domain */ + +#ifndef __7Z_SORT_H +#define __7Z_SORT_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void HeapSort(UInt32 *p, size_t size); +void HeapSort64(UInt64 *p, size_t size); + +/* void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size); */ + +EXTERN_C_END + +#endif diff --git a/dep/lzma/lzma-history.txt b/dep/lzma/lzma-history.txt new file mode 100644 index 000000000..48ee74813 --- /dev/null +++ b/dep/lzma/lzma-history.txt @@ -0,0 +1,446 @@ +HISTORY of the LZMA SDK +----------------------- + +19.00 2019-02-21 +------------------------- +- Encryption strength for 7z archives was increased: + the size of random initialization vector was increased from 64-bit to 128-bit, + and the pseudo-random number generator was improved. +- The bug in 7zIn.c code was fixed. + + +18.06 2018-12-30 +------------------------- +- The speed for LZMA/LZMA2 compressing was increased by 3-10%, + and there are minor changes in compression ratio. +- Some bugs were fixed. +- The bug in 7-Zip 18.02-18.05 was fixed: + There was memory leak in multithreading xz decoder - XzDecMt_Decode(), + if xz stream contains only one block. +- The changes for MSVS compiler makefiles: + - the makefiles now use "PLATFORM" macroname with values (x64, x86, arm64) + instead of "CPU" macroname with values (AMD64, ARM64). + - the makefiles by default now use static version of the run-time library. + + +18.05 2018-04-30 +------------------------- +- The speed for LZMA/LZMA2 compressing was increased + by 8% for fastest/fast compression levels and + by 3% for normal/maximum compression levels. +- Previous versions of 7-Zip could work incorrectly in "Large memory pages" mode in + Windows 10 because of some BUG with "Large Pages" in Windows 10. + Now 7-Zip doesn't use "Large Pages" on Windows 10 up to revision 1709 (16299). +- The BUG was fixed in Lzma2Enc.c + Lzma2Enc_Encode2() function worked incorretly, + if (inStream == NULL) and the number of block threads is more than 1. + + +18.03 beta 2018-03-04 +------------------------- +- Asm\x86\LzmaDecOpt.asm: new optimized LZMA decoder written in asm + for x64 with about 30% higher speed than main version of LZMA decoder written in C. +- The speed for single-thread LZMA/LZMA2 decoder written in C was increased by 3%. +- 7-Zip now can use multi-threading for 7z/LZMA2 decoding, + if there are multiple independent data chunks in LZMA2 stream. +- 7-Zip now can use multi-threading for xz decoding, + if there are multiple blocks in xz stream. + + +18.01 2019-01-28 +------------------------- +- The BUG in 17.01 - 18.00 beta was fixed: + XzDec.c : random block unpacking and XzUnpacker_IsBlockFinished() + didn't work correctly for xz archives without checksum (CRC). + + +18.00 beta 2019-01-10 +------------------------- +- The BUG in xz encoder was fixed: + There was memory leak of 16 KB for each file compressed with + xz compression method, if additional filter was used. + + +17.01 beta 2017-08-28 +------------------------- +- Minor speed optimization for LZMA2 (xz and 7z) multi-threading compression. + 7-Zip now uses additional memory buffers for multi-block LZMA2 compression. + CPU utilization was slightly improved. +- 7-zip now creates multi-block xz archives by default. Block size can be + specified with -ms[Size]{m|g} switch. +- xz decoder now can unpack random block from multi-block xz archives. +- 7-Zip command line: @listfile now doesn't work after -- switch. + Use -i@listfile before -- switch instead. +- The BUGs were fixed: + 7-Zip 17.00 beta crashed for commands that write anti-item to 7z archive. + + +17.00 beta 2017-04-29 +------------------------- +- NewHandler.h / NewHandler.cpp: + now it redefines operator new() only for old MSVC compilers (_MSC_VER < 1900). +- C/7zTypes.h : the names of variables in interface structures were changed (vt). +- Some bugs were fixed. 7-Zip could crash in some cases. +- Some internal changes in code. + + +16.04 2016-10-04 +------------------------- +- The bug was fixed in DllSecur.c. + + +16.03 2016-09-28 +------------------------- +- SFX modules now use some protection against DLL preloading attack. +- Some bugs in 7z code were fixed. + + +16.02 2016-05-21 +------------------------- +- The BUG in 16.00 - 16.01 was fixed: + Split Handler (SplitHandler.cpp) returned incorrect + total size value (kpidSize) for split archives. + + +16.01 2016-05-19 +------------------------- +- Some internal changes to reduce the number of compiler warnings. + + +16.00 2016-05-10 +------------------------- +- Some bugs were fixed. + + +15.12 2015-11-19 +------------------------- +- The BUG in C version of 7z decoder was fixed: + 7zDec.c : SzDecodeLzma2() + 7z decoder could mistakenly report about decoding error for some 7z archives + that use LZMA2 compression method. + The probability to get that mistaken decoding error report was about + one error per 16384 solid blocks for solid blocks larger than 16 KB (compressed size). +- The BUG (in 9.26-15.11) in C version of 7z decoder was fixed: + 7zArcIn.c : SzReadHeader2() + 7z decoder worked incorrectly for 7z archives that contain + empty solid blocks, that can be placed to 7z archive, if some file is + unavailable for reading during archive creation. + + +15.09 beta 2015-10-16 +------------------------- +- The BUG in LZMA / LZMA2 encoding code was fixed. + The BUG in LzFind.c::MatchFinder_ReadBlock() function. + If input data size is larger than (4 GiB - dictionary_size), + the following code worked incorrectly: + - LZMA : LzmaEnc_MemEncode(), LzmaEncode() : LZMA encoding functions + for compressing from memory to memory. + That BUG is not related to LZMA encoder version that works via streams. + - LZMA2 : multi-threaded version of LZMA2 encoder worked incorrectly, if + default value of chunk size (CLzma2EncProps::blockSize) is changed + to value larger than (4 GiB - dictionary_size). + + +9.38 beta 2015-01-03 +------------------------- +- The BUG in 9.31-9.37 was fixed: + IArchiveGetRawProps interface was disabled for 7z archives. +- The BUG in 9.26-9.36 was fixed: + Some code in CPP\7zip\Archive\7z\ worked correctly only under Windows. + + +9.36 beta 2014-12-26 +------------------------- +- The BUG in command line version was fixed: + 7-Zip created temporary archive in current folder during update archive + operation, if -w{Path} switch was not specified. + The fixed 7-Zip creates temporary archive in folder that contains updated archive. +- The BUG in 9.33-9.35 was fixed: + 7-Zip silently ignored file reading errors during 7z or gz archive creation, + and the created archive contained only part of file that was read before error. + The fixed 7-Zip stops archive creation and it reports about error. + + +9.35 beta 2014-12-07 +------------------------- +- 7zr.exe now support AES encryption. +- SFX mudules were added to LZMA SDK +- Some bugs were fixed. + + +9.21 beta 2011-04-11 +------------------------- +- New class FString for file names at file systems. +- Speed optimization in CRC code for big-endian CPUs. +- The BUG in Lzma2Dec.c was fixed: + Lzma2Decode function didn't work. + + +9.18 beta 2010-11-02 +------------------------- +- New small SFX module for installers (SfxSetup). + + +9.12 beta 2010-03-24 +------------------------- +- The BUG in LZMA SDK 9.* was fixed: LZMA2 codec didn't work, + if more than 10 threads were used (or more than 20 threads in some modes). + + +9.11 beta 2010-03-15 +------------------------- +- PPMd compression method support + + +9.09 2009-12-12 +------------------------- +- The bug was fixed: + Utf16_To_Utf8 funstions in UTFConvert.cpp and 7zMain.c + incorrectly converted surrogate characters (the code >= 0x10000) to UTF-8. +- Some bugs were fixed + + +9.06 2009-08-17 +------------------------- +- Some changes in ANSI-C 7z Decoder interfaces. + + +9.04 2009-05-30 +------------------------- +- LZMA2 compression method support +- xz format support + + +4.65 2009-02-03 +------------------------- +- Some minor fixes + + +4.63 2008-12-31 +------------------------- +- Some minor fixes + + +4.61 beta 2008-11-23 +------------------------- +- The bug in ANSI-C LZMA Decoder was fixed: + If encoded stream was corrupted, decoder could access memory + outside of allocated range. +- Some changes in ANSI-C 7z Decoder interfaces. +- LZMA SDK is placed in the public domain. + + +4.60 beta 2008-08-19 +------------------------- +- Some minor fixes. + + +4.59 beta 2008-08-13 +------------------------- +- The bug was fixed: + LZMA Encoder in fast compression mode could access memory outside of + allocated range in some rare cases. + + +4.58 beta 2008-05-05 +------------------------- +- ANSI-C LZMA Decoder was rewritten for speed optimizations. +- ANSI-C LZMA Encoder was included to LZMA SDK. +- C++ LZMA code now is just wrapper over ANSI-C code. + + +4.57 2007-12-12 +------------------------- +- Speed optimizations in Ñ++ LZMA Decoder. +- Small changes for more compatibility with some C/C++ compilers. + + +4.49 beta 2007-07-05 +------------------------- +- .7z ANSI-C Decoder: + - now it supports BCJ and BCJ2 filters + - now it supports files larger than 4 GB. + - now it supports "Last Write Time" field for files. +- C++ code for .7z archives compressing/decompressing from 7-zip + was included to LZMA SDK. + + +4.43 2006-06-04 +------------------------- +- Small changes for more compatibility with some C/C++ compilers. + + +4.42 2006-05-15 +------------------------- +- Small changes in .h files in ANSI-C version. + + +4.39 beta 2006-04-14 +------------------------- +- The bug in versions 4.33b:4.38b was fixed: + C++ version of LZMA encoder could not correctly compress + files larger than 2 GB with HC4 match finder (-mfhc4). + + +4.37 beta 2005-04-06 +------------------------- +- Fixes in C++ code: code could no be compiled if _NO_EXCEPTIONS was defined. + + +4.35 beta 2005-03-02 +------------------------- +- The bug was fixed in C++ version of LZMA Decoder: + If encoded stream was corrupted, decoder could access memory + outside of allocated range. + + +4.34 beta 2006-02-27 +------------------------- +- Compressing speed and memory requirements for compressing were increased +- LZMA now can use only these match finders: HC4, BT2, BT3, BT4 + + +4.32 2005-12-09 +------------------------- +- Java version of LZMA SDK was included + + +4.30 2005-11-20 +------------------------- +- Compression ratio was improved in -a2 mode +- Speed optimizations for compressing in -a2 mode +- -fb switch now supports values up to 273 +- The bug in 7z_C (7zIn.c) was fixed: + It used Alloc/Free functions from different memory pools. + So if program used two memory pools, it worked incorrectly. +- 7z_C: .7z format supporting was improved +- LZMA# SDK (C#.NET version) was included + + +4.27 (Updated) 2005-09-21 +------------------------- +- Some GUIDs/interfaces in C++ were changed. + IStream.h: + ISequentialInStream::Read now works as old ReadPart + ISequentialOutStream::Write now works as old WritePart + + +4.27 2005-08-07 +------------------------- +- The bug in LzmaDecodeSize.c was fixed: + if _LZMA_IN_CB and _LZMA_OUT_READ were defined, + decompressing worked incorrectly. + + +4.26 2005-08-05 +------------------------- +- Fixes in 7z_C code and LzmaTest.c: + previous versions could work incorrectly, + if malloc(0) returns 0 + + +4.23 2005-06-29 +------------------------- +- Small fixes in C++ code + + +4.22 2005-06-10 +------------------------- +- Small fixes + + +4.21 2005-06-08 +------------------------- +- Interfaces for ANSI-C LZMA Decoder (LzmaDecode.c) were changed +- New additional version of ANSI-C LZMA Decoder with zlib-like interface: + - LzmaStateDecode.h + - LzmaStateDecode.c + - LzmaStateTest.c +- ANSI-C LZMA Decoder now can decompress files larger than 4 GB + + +4.17 2005-04-18 +------------------------- +- New example for RAM->RAM compressing/decompressing: + LZMA + BCJ (filter for x86 code): + - LzmaRam.h + - LzmaRam.cpp + - LzmaRamDecode.h + - LzmaRamDecode.c + - -f86 switch for lzma.exe + + +4.16 2005-03-29 +------------------------- +- The bug was fixed in LzmaDecode.c (ANSI-C LZMA Decoder): + If _LZMA_OUT_READ was defined, and if encoded stream was corrupted, + decoder could access memory outside of allocated range. +- Speed optimization of ANSI-C LZMA Decoder (now it's about 20% faster). + Old version of LZMA Decoder now is in file LzmaDecodeSize.c. + LzmaDecodeSize.c can provide slightly smaller code than LzmaDecode.c +- Small speed optimization in LZMA C++ code +- filter for SPARC's code was added +- Simplified version of .7z ANSI-C Decoder was included + + +4.06 2004-09-05 +------------------------- +- The bug in v4.05 was fixed: + LZMA-Encoder didn't release output stream in some cases. + + +4.05 2004-08-25 +------------------------- +- Source code of filters for x86, IA-64, ARM, ARM-Thumb + and PowerPC code was included to SDK +- Some internal minor changes + + +4.04 2004-07-28 +------------------------- +- More compatibility with some C++ compilers + + +4.03 2004-06-18 +------------------------- +- "Benchmark" command was added. It measures compressing + and decompressing speed and shows rating values. + Also it checks hardware errors. + + +4.02 2004-06-10 +------------------------- +- C++ LZMA Encoder/Decoder code now is more portable + and it can be compiled by GCC on Linux. + + +4.01 2004-02-15 +------------------------- +- Some detection of data corruption was enabled. + LzmaDecode.c / RangeDecoderReadByte + ..... + { + rd->ExtraBytes = 1; + return 0xFF; + } + + +4.00 2004-02-13 +------------------------- +- Original version of LZMA SDK + + + +HISTORY of the LZMA +------------------- + 2001-2008: Improvements to LZMA compressing/decompressing code, + keeping compatibility with original LZMA format + 1996-2001: Development of LZMA compression format + + Some milestones: + + 2001-08-30: LZMA compression was added to 7-Zip + 1999-01-02: First version of 7-Zip was released + + +End of document diff --git a/dep/lzma/lzma.txt b/dep/lzma/lzma.txt new file mode 100644 index 000000000..a65988fe0 --- /dev/null +++ b/dep/lzma/lzma.txt @@ -0,0 +1,328 @@ +LZMA compression +---------------- +Version: 9.35 + +This file describes LZMA encoding and decoding functions written in C language. + +LZMA is an improved version of famous LZ77 compression algorithm. +It was improved in way of maximum increasing of compression ratio, +keeping high decompression speed and low memory requirements for +decompressing. + +Note: you can read also LZMA Specification (lzma-specification.txt from LZMA SDK) + +Also you can look source code for LZMA encoding and decoding: + C/Util/Lzma/LzmaUtil.c + + +LZMA compressed file format +--------------------------- +Offset Size Description + 0 1 Special LZMA properties (lc,lp, pb in encoded form) + 1 4 Dictionary size (little endian) + 5 8 Uncompressed size (little endian). -1 means unknown size + 13 Compressed data + + + +ANSI-C LZMA Decoder +~~~~~~~~~~~~~~~~~~~ + +Please note that interfaces for ANSI-C code were changed in LZMA SDK 4.58. +If you want to use old interfaces you can download previous version of LZMA SDK +from sourceforge.net site. + +To use ANSI-C LZMA Decoder you need the following files: +1) LzmaDec.h + LzmaDec.c + 7zTypes.h + Precomp.h + Compiler.h + +Look example code: + C/Util/Lzma/LzmaUtil.c + + +Memory requirements for LZMA decoding +------------------------------------- + +Stack usage of LZMA decoding function for local variables is not +larger than 200-400 bytes. + +LZMA Decoder uses dictionary buffer and internal state structure. +Internal state structure consumes + state_size = (4 + (1.5 << (lc + lp))) KB +by default (lc=3, lp=0), state_size = 16 KB. + + +How To decompress data +---------------------- + +LZMA Decoder (ANSI-C version) now supports 2 interfaces: +1) Single-call Decompressing +2) Multi-call State Decompressing (zlib-like interface) + +You must use external allocator: +Example: +void *SzAlloc(void *p, size_t size) { p = p; return malloc(size); } +void SzFree(void *p, void *address) { p = p; free(address); } +ISzAlloc alloc = { SzAlloc, SzFree }; + +You can use p = p; operator to disable compiler warnings. + + +Single-call Decompressing +------------------------- +When to use: RAM->RAM decompressing +Compile files: LzmaDec.h + LzmaDec.c + 7zTypes.h +Compile defines: no defines +Memory Requirements: + - Input buffer: compressed size + - Output buffer: uncompressed size + - LZMA Internal Structures: state_size (16 KB for default settings) + +Interface: + int LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAlloc *alloc); + In: + dest - output data + destLen - output data size + src - input data + srcLen - input data size + propData - LZMA properties (5 bytes) + propSize - size of propData buffer (5 bytes) + finishMode - It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). + You can use LZMA_FINISH_END, when you know that + current output buffer covers last bytes of stream. + alloc - Memory allocator. + + Out: + destLen - processed output size + srcLen - processed input size + + Output: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). + + If LZMA decoder sees end_marker before reaching output limit, it returns OK result, + and output value of destLen will be less than output buffer size limit. + + You can use multiple checks to test data integrity after full decompression: + 1) Check Result and "status" variable. + 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. + 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. + You must use correct finish mode in that case. */ + + +Multi-call State Decompressing (zlib-like interface) +---------------------------------------------------- + +When to use: file->file decompressing +Compile files: LzmaDec.h + LzmaDec.c + 7zTypes.h + +Memory Requirements: + - Buffer for input stream: any size (for example, 16 KB) + - Buffer for output stream: any size (for example, 16 KB) + - LZMA Internal Structures: state_size (16 KB for default settings) + - LZMA dictionary (dictionary size is encoded in LZMA properties header) + +1) read LZMA properties (5 bytes) and uncompressed size (8 bytes, little-endian) to header: + unsigned char header[LZMA_PROPS_SIZE + 8]; + ReadFile(inFile, header, sizeof(header) + +2) Allocate CLzmaDec structures (state + dictionary) using LZMA properties + + CLzmaDec state; + LzmaDec_Constr(&state); + res = LzmaDec_Allocate(&state, header, LZMA_PROPS_SIZE, &g_Alloc); + if (res != SZ_OK) + return res; + +3) Init LzmaDec structure before any new LZMA stream. And call LzmaDec_DecodeToBuf in loop + + LzmaDec_Init(&state); + for (;;) + { + ... + int res = LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode); + ... + } + + +4) Free all allocated structures + LzmaDec_Free(&state, &g_Alloc); + +Look example code: + C/Util/Lzma/LzmaUtil.c + + +How To compress data +-------------------- + +Compile files: + 7zTypes.h + Threads.h + LzmaEnc.h + LzmaEnc.c + LzFind.h + LzFind.c + LzFindMt.h + LzFindMt.c + LzHash.h + +Memory Requirements: + - (dictSize * 11.5 + 6 MB) + state_size + +Lzma Encoder can use two memory allocators: +1) alloc - for small arrays. +2) allocBig - for big arrays. + +For example, you can use Large RAM Pages (2 MB) in allocBig allocator for +better compression speed. Note that Windows has bad implementation for +Large RAM Pages. +It's OK to use same allocator for alloc and allocBig. + + +Single-call Compression with callbacks +-------------------------------------- + +Look example code: + C/Util/Lzma/LzmaUtil.c + +When to use: file->file compressing + +1) you must implement callback structures for interfaces: +ISeqInStream +ISeqOutStream +ICompressProgress +ISzAlloc + +static void *SzAlloc(void *p, size_t size) { p = p; return MyAlloc(size); } +static void SzFree(void *p, void *address) { p = p; MyFree(address); } +static ISzAlloc g_Alloc = { SzAlloc, SzFree }; + + CFileSeqInStream inStream; + CFileSeqOutStream outStream; + + inStream.funcTable.Read = MyRead; + inStream.file = inFile; + outStream.funcTable.Write = MyWrite; + outStream.file = outFile; + + +2) Create CLzmaEncHandle object; + + CLzmaEncHandle enc; + + enc = LzmaEnc_Create(&g_Alloc); + if (enc == 0) + return SZ_ERROR_MEM; + + +3) initialize CLzmaEncProps properties; + + LzmaEncProps_Init(&props); + + Then you can change some properties in that structure. + +4) Send LZMA properties to LZMA Encoder + + res = LzmaEnc_SetProps(enc, &props); + +5) Write encoded properties to header + + Byte header[LZMA_PROPS_SIZE + 8]; + size_t headerSize = LZMA_PROPS_SIZE; + UInt64 fileSize; + int i; + + res = LzmaEnc_WriteProperties(enc, header, &headerSize); + fileSize = MyGetFileLength(inFile); + for (i = 0; i < 8; i++) + header[headerSize++] = (Byte)(fileSize >> (8 * i)); + MyWriteFileAndCheck(outFile, header, headerSize) + +6) Call encoding function: + res = LzmaEnc_Encode(enc, &outStream.funcTable, &inStream.funcTable, + NULL, &g_Alloc, &g_Alloc); + +7) Destroy LZMA Encoder Object + LzmaEnc_Destroy(enc, &g_Alloc, &g_Alloc); + + +If callback function return some error code, LzmaEnc_Encode also returns that code +or it can return the code like SZ_ERROR_READ, SZ_ERROR_WRITE or SZ_ERROR_PROGRESS. + + +Single-call RAM->RAM Compression +-------------------------------- + +Single-call RAM->RAM Compression is similar to Compression with callbacks, +but you provide pointers to buffers instead of pointers to stream callbacks: + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); + +Return code: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) + + + +Defines +------- + +_LZMA_SIZE_OPT - Enable some optimizations in LZMA Decoder to get smaller executable code. + +_LZMA_PROB32 - It can increase the speed on some 32-bit CPUs, but memory usage for + some structures will be doubled in that case. + +_LZMA_UINT32_IS_ULONG - Define it if int is 16-bit on your compiler and long is 32-bit. + +_LZMA_NO_SYSTEM_SIZE_T - Define it if you don't want to use size_t type. + + +_7ZIP_PPMD_SUPPPORT - Define it if you don't want to support PPMD method in AMSI-C .7z decoder. + + +C++ LZMA Encoder/Decoder +~~~~~~~~~~~~~~~~~~~~~~~~ +C++ LZMA code use COM-like interfaces. So if you want to use it, +you can study basics of COM/OLE. +C++ LZMA code is just wrapper over ANSI-C code. + + +C++ Notes +~~~~~~~~~~~~~~~~~~~~~~~~ +If you use some C++ code folders in 7-Zip (for example, C++ code for .7z handling), +you must check that you correctly work with "new" operator. +7-Zip can be compiled with MSVC 6.0 that doesn't throw "exception" from "new" operator. +So 7-Zip uses "CPP\Common\NewHandler.cpp" that redefines "new" operator: +operator new(size_t size) +{ + void *p = ::malloc(size); + if (p == 0) + throw CNewException(); + return p; +} +If you use MSCV that throws exception for "new" operator, you can compile without +"NewHandler.cpp". So standard exception will be used. Actually some code of +7-Zip catches any exception in internal code and converts it to HRESULT code. +So you don't need to catch CNewException, if you call COM interfaces of 7-Zip. + +--- + +http://www.7-zip.org +http://www.7-zip.org/sdk.html +http://www.7-zip.org/support.html diff --git a/dep/lzma/lzma.vcxproj b/dep/lzma/lzma.vcxproj new file mode 100644 index 000000000..e3f3540be --- /dev/null +++ b/dep/lzma/lzma.vcxproj @@ -0,0 +1,361 @@ + + + + + DebugFast + Win32 + + + DebugFast + x64 + + + Debug + Win32 + + + Debug + x64 + + + ReleaseLTCG + Win32 + + + ReleaseLTCG + x64 + + + Release + Win32 + + + Release + x64 + + + + + + + + + + + + + + + + + + {DD944834-7899-4C1C-A4C1-064B5009D239} + Win32Proj + lzma + 10.0 + + + + StaticLibrary + true + v142 + NotSet + + + StaticLibrary + true + v142 + NotSet + + + StaticLibrary + true + v142 + NotSet + + + StaticLibrary + true + v142 + NotSet + + + StaticLibrary + false + v142 + true + NotSet + + + StaticLibrary + false + v142 + true + NotSet + + + StaticLibrary + false + v142 + true + NotSet + + + StaticLibrary + false + v142 + true + NotSet + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + true + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + + + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + true + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + + + true + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + + + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + true + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + + + false + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + + + false + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + + + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + false + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + + + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + $(ProjectName)-$(Platform)-$(Configuration) + false + $(SolutionDir)build\$(ProjectName)-$(Platform)-$(Configuration)\ + + + + + + Level2 + Disabled + _7ZIP_ST;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;_DEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + true + ProgramDatabase + $(ProjectDir)include;%(AdditionalIncludeDirectories) + false + stdcpp17 + true + true + + + Windows + true + + + + + + + Level2 + Disabled + _7ZIP_ST;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;_DEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + true + ProgramDatabase + $(ProjectDir)include;%(AdditionalIncludeDirectories) + false + stdcpp17 + true + true + + + Windows + true + + + + + + + Level2 + Disabled + _7ZIP_ST;_ITERATOR_DEBUG_LEVEL=1;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;_DEBUGFAST;_DEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + true + ProgramDatabase + $(ProjectDir)include;%(AdditionalIncludeDirectories) + Default + false + stdcpp17 + false + true + OnlyExplicitInline + true + + + Windows + true + + + + + + + Level2 + Disabled + _7ZIP_ST;_ITERATOR_DEBUG_LEVEL=1;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;_DEBUGFAST;_DEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + true + ProgramDatabase + $(ProjectDir)include;%(AdditionalIncludeDirectories) + Default + false + stdcpp17 + false + true + OnlyExplicitInline + true + + + Windows + true + + + + + Level2 + + + MaxSpeed + true + _7ZIP_ST;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + $(ProjectDir)include;%(AdditionalIncludeDirectories) + false + stdcpp17 + true + true + + + Windows + true + true + true + + + + + Level2 + + + MaxSpeed + true + _7ZIP_ST;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + $(ProjectDir)include;%(AdditionalIncludeDirectories) + true + stdcpp17 + true + true + true + + + Windows + true + true + true + + + + + Level2 + + + MaxSpeed + true + _7ZIP_ST;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + $(ProjectDir)include;%(AdditionalIncludeDirectories) + false + stdcpp17 + true + true + + + Windows + true + true + true + + + + + Level2 + + + MaxSpeed + true + _7ZIP_ST;_CRT_NONSTDC_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions) + $(ProjectDir)include;%(AdditionalIncludeDirectories) + true + stdcpp17 + true + true + true + + + Windows + true + true + true + + + + + + \ No newline at end of file diff --git a/dep/lzma/lzma.vcxproj.filters b/dep/lzma/lzma.vcxproj.filters new file mode 100644 index 000000000..8725b394c --- /dev/null +++ b/dep/lzma/lzma.vcxproj.filters @@ -0,0 +1,17 @@ + + + + + + + + + + + + + + + + + \ No newline at end of file diff --git a/dep/lzma/src/Alloc.c b/dep/lzma/src/Alloc.c new file mode 100644 index 000000000..bcede4b85 --- /dev/null +++ b/dep/lzma/src/Alloc.c @@ -0,0 +1,455 @@ +/* Alloc.c -- Memory allocation functions +2018-04-27 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#ifdef _WIN32 +#include +#endif +#include + +#include "Alloc.h" + +/* #define _SZ_ALLOC_DEBUG */ + +/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ +#ifdef _SZ_ALLOC_DEBUG + +#include +int g_allocCount = 0; +int g_allocCountMid = 0; +int g_allocCountBig = 0; + + +#define CONVERT_INT_TO_STR(charType, tempSize) \ + unsigned char temp[tempSize]; unsigned i = 0; \ + while (val >= 10) { temp[i++] = (unsigned char)('0' + (unsigned)(val % 10)); val /= 10; } \ + *s++ = (charType)('0' + (unsigned)val); \ + while (i != 0) { i--; *s++ = temp[i]; } \ + *s = 0; + +static void ConvertUInt64ToString(UInt64 val, char *s) +{ + CONVERT_INT_TO_STR(char, 24); +} + +#define GET_HEX_CHAR(t) ((char)(((t < 10) ? ('0' + t) : ('A' + (t - 10))))) + +static void ConvertUInt64ToHex(UInt64 val, char *s) +{ + UInt64 v = val; + unsigned i; + for (i = 1;; i++) + { + v >>= 4; + if (v == 0) + break; + } + s[i] = 0; + do + { + unsigned t = (unsigned)(val & 0xF); + val >>= 4; + s[--i] = GET_HEX_CHAR(t); + } + while (i); +} + +#define DEBUG_OUT_STREAM stderr + +static void Print(const char *s) +{ + fputs(s, DEBUG_OUT_STREAM); +} + +static void PrintAligned(const char *s, size_t align) +{ + size_t len = strlen(s); + for(;;) + { + fputc(' ', DEBUG_OUT_STREAM); + if (len >= align) + break; + ++len; + } + Print(s); +} + +static void PrintLn() +{ + Print("\n"); +} + +static void PrintHex(UInt64 v, size_t align) +{ + char s[32]; + ConvertUInt64ToHex(v, s); + PrintAligned(s, align); +} + +static void PrintDec(UInt64 v, size_t align) +{ + char s[32]; + ConvertUInt64ToString(v, s); + PrintAligned(s, align); +} + +static void PrintAddr(void *p) +{ + PrintHex((UInt64)(size_t)(ptrdiff_t)p, 12); +} + + +#define PRINT_ALLOC(name, cnt, size, ptr) \ + Print(name " "); \ + PrintDec(cnt++, 10); \ + PrintHex(size, 10); \ + PrintAddr(ptr); \ + PrintLn(); + +#define PRINT_FREE(name, cnt, ptr) if (ptr) { \ + Print(name " "); \ + PrintDec(--cnt, 10); \ + PrintAddr(ptr); \ + PrintLn(); } + +#else + +#define PRINT_ALLOC(name, cnt, size, ptr) +#define PRINT_FREE(name, cnt, ptr) +#define Print(s) +#define PrintLn() +#define PrintHex(v, align) +#define PrintDec(v, align) +#define PrintAddr(p) + +#endif + + + +void *MyAlloc(size_t size) +{ + if (size == 0) + return NULL; + #ifdef _SZ_ALLOC_DEBUG + { + void *p = malloc(size); + PRINT_ALLOC("Alloc ", g_allocCount, size, p); + return p; + } + #else + return malloc(size); + #endif +} + +void MyFree(void *address) +{ + PRINT_FREE("Free ", g_allocCount, address); + + free(address); +} + +#ifdef _WIN32 + +void *MidAlloc(size_t size) +{ + if (size == 0) + return NULL; + + PRINT_ALLOC("Alloc-Mid", g_allocCountMid, size, NULL); + + return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); +} + +void MidFree(void *address) +{ + PRINT_FREE("Free-Mid", g_allocCountMid, address); + + if (!address) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#ifndef MEM_LARGE_PAGES +#undef _7ZIP_LARGE_PAGES +#endif + +#ifdef _7ZIP_LARGE_PAGES +SIZE_T g_LargePageSize = 0; +typedef SIZE_T (WINAPI *GetLargePageMinimumP)(); +#endif + +void SetLargePageSize() +{ + #ifdef _7ZIP_LARGE_PAGES + SIZE_T size; + GetLargePageMinimumP largePageMinimum = (GetLargePageMinimumP) + GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "GetLargePageMinimum"); + if (!largePageMinimum) + return; + size = largePageMinimum(); + if (size == 0 || (size & (size - 1)) != 0) + return; + g_LargePageSize = size; + #endif +} + + +void *BigAlloc(size_t size) +{ + if (size == 0) + return NULL; + + PRINT_ALLOC("Alloc-Big", g_allocCountBig, size, NULL); + + #ifdef _7ZIP_LARGE_PAGES + { + SIZE_T ps = g_LargePageSize; + if (ps != 0 && ps <= (1 << 30) && size > (ps / 2)) + { + size_t size2; + ps--; + size2 = (size + ps) & ~ps; + if (size2 >= size) + { + void *res = VirtualAlloc(NULL, size2, MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE); + if (res) + return res; + } + } + } + #endif + + return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); +} + +void BigFree(void *address) +{ + PRINT_FREE("Free-Big", g_allocCountBig, address); + + if (!address) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#endif + + +static void *SzAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MyAlloc(size); } +static void SzFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MyFree(address); } +const ISzAlloc g_Alloc = { SzAlloc, SzFree }; + +static void *SzMidAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MidAlloc(size); } +static void SzMidFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MidFree(address); } +const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree }; + +static void *SzBigAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); } +static void SzBigFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); BigFree(address); } +const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree }; + + +/* + uintptr_t : C99 (optional) + : unsupported in VS6 +*/ + +#ifdef _WIN32 + typedef UINT_PTR UIntPtr; +#else + /* + typedef uintptr_t UIntPtr; + */ + typedef ptrdiff_t UIntPtr; +#endif + + +#define ADJUST_ALLOC_SIZE 0 +/* +#define ADJUST_ALLOC_SIZE (sizeof(void *) - 1) +*/ +/* + Use (ADJUST_ALLOC_SIZE = (sizeof(void *) - 1)), if + MyAlloc() can return address that is NOT multiple of sizeof(void *). +*/ + + +/* +#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((char *)(p) - ((size_t)(UIntPtr)(p) & ((align) - 1)))) +*/ +#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((((UIntPtr)(p)) & ~((UIntPtr)(align) - 1)))) + +#define MY_ALIGN_PTR_UP_PLUS(p, align) MY_ALIGN_PTR_DOWN(((char *)(p) + (align) + ADJUST_ALLOC_SIZE), align) + + +#if (_POSIX_C_SOURCE >= 200112L) && !defined(_WIN32) + #define USE_posix_memalign +#endif + +/* + This posix_memalign() is for test purposes only. + We also need special Free() function instead of free(), + if this posix_memalign() is used. +*/ + +/* +static int posix_memalign(void **ptr, size_t align, size_t size) +{ + size_t newSize = size + align; + void *p; + void *pAligned; + *ptr = NULL; + if (newSize < size) + return 12; // ENOMEM + p = MyAlloc(newSize); + if (!p) + return 12; // ENOMEM + pAligned = MY_ALIGN_PTR_UP_PLUS(p, align); + ((void **)pAligned)[-1] = p; + *ptr = pAligned; + return 0; +} +*/ + +/* + ALLOC_ALIGN_SIZE >= sizeof(void *) + ALLOC_ALIGN_SIZE >= cache_line_size +*/ + +#define ALLOC_ALIGN_SIZE ((size_t)1 << 7) + +static void *SzAlignedAlloc(ISzAllocPtr pp, size_t size) +{ + #ifndef USE_posix_memalign + + void *p; + void *pAligned; + size_t newSize; + UNUSED_VAR(pp); + + /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned + block to prevent cache line sharing with another allocated blocks */ + + newSize = size + ALLOC_ALIGN_SIZE * 1 + ADJUST_ALLOC_SIZE; + if (newSize < size) + return NULL; + + p = MyAlloc(newSize); + + if (!p) + return NULL; + pAligned = MY_ALIGN_PTR_UP_PLUS(p, ALLOC_ALIGN_SIZE); + + Print(" size="); PrintHex(size, 8); + Print(" a_size="); PrintHex(newSize, 8); + Print(" ptr="); PrintAddr(p); + Print(" a_ptr="); PrintAddr(pAligned); + PrintLn(); + + ((void **)pAligned)[-1] = p; + + return pAligned; + + #else + + void *p; + UNUSED_VAR(pp); + if (posix_memalign(&p, ALLOC_ALIGN_SIZE, size)) + return NULL; + + Print(" posix_memalign="); PrintAddr(p); + PrintLn(); + + return p; + + #endif +} + + +static void SzAlignedFree(ISzAllocPtr pp, void *address) +{ + UNUSED_VAR(pp); + #ifndef USE_posix_memalign + if (address) + MyFree(((void **)address)[-1]); + #else + free(address); + #endif +} + + +const ISzAlloc g_AlignedAlloc = { SzAlignedAlloc, SzAlignedFree }; + + + +#define MY_ALIGN_PTR_DOWN_1(p) MY_ALIGN_PTR_DOWN(p, sizeof(void *)) + +/* we align ptr to support cases where CAlignOffsetAlloc::offset is not multiply of sizeof(void *) */ +#define REAL_BLOCK_PTR_VAR(p) ((void **)MY_ALIGN_PTR_DOWN_1(p))[-1] +/* +#define REAL_BLOCK_PTR_VAR(p) ((void **)(p))[-1] +*/ + +static void *AlignOffsetAlloc_Alloc(ISzAllocPtr pp, size_t size) +{ + CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); + void *adr; + void *pAligned; + size_t newSize; + size_t extra; + size_t alignSize = (size_t)1 << p->numAlignBits; + + if (alignSize < sizeof(void *)) + alignSize = sizeof(void *); + + if (p->offset >= alignSize) + return NULL; + + /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned + block to prevent cache line sharing with another allocated blocks */ + extra = p->offset & (sizeof(void *) - 1); + newSize = size + alignSize + extra + ADJUST_ALLOC_SIZE; + if (newSize < size) + return NULL; + + adr = ISzAlloc_Alloc(p->baseAlloc, newSize); + + if (!adr) + return NULL; + + pAligned = (char *)MY_ALIGN_PTR_DOWN((char *)adr + + alignSize - p->offset + extra + ADJUST_ALLOC_SIZE, alignSize) + p->offset; + + PrintLn(); + Print("- Aligned: "); + Print(" size="); PrintHex(size, 8); + Print(" a_size="); PrintHex(newSize, 8); + Print(" ptr="); PrintAddr(adr); + Print(" a_ptr="); PrintAddr(pAligned); + PrintLn(); + + REAL_BLOCK_PTR_VAR(pAligned) = adr; + + return pAligned; +} + + +static void AlignOffsetAlloc_Free(ISzAllocPtr pp, void *address) +{ + if (address) + { + CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); + PrintLn(); + Print("- Aligned Free: "); + PrintLn(); + ISzAlloc_Free(p->baseAlloc, REAL_BLOCK_PTR_VAR(address)); + } +} + + +void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p) +{ + p->vt.Alloc = AlignOffsetAlloc_Alloc; + p->vt.Free = AlignOffsetAlloc_Free; +} diff --git a/dep/lzma/src/Bra86.c b/dep/lzma/src/Bra86.c new file mode 100644 index 000000000..93ed4d762 --- /dev/null +++ b/dep/lzma/src/Bra86.c @@ -0,0 +1,82 @@ +/* Bra86.c -- Converter for x86 code (BCJ) +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Bra.h" + +#define Test86MSByte(b) ((((b) + 1) & 0xFE) == 0) + +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding) +{ + SizeT pos = 0; + UInt32 mask = *state & 7; + if (size < 5) + return 0; + size -= 4; + ip += 5; + + for (;;) + { + Byte *p = data + pos; + const Byte *limit = data + size; + for (; p < limit; p++) + if ((*p & 0xFE) == 0xE8) + break; + + { + SizeT d = (SizeT)(p - data - pos); + pos = (SizeT)(p - data); + if (p >= limit) + { + *state = (d > 2 ? 0 : mask >> (unsigned)d); + return pos; + } + if (d > 2) + mask = 0; + else + { + mask >>= (unsigned)d; + if (mask != 0 && (mask > 4 || mask == 3 || Test86MSByte(p[(size_t)(mask >> 1) + 1]))) + { + mask = (mask >> 1) | 4; + pos++; + continue; + } + } + } + + if (Test86MSByte(p[4])) + { + UInt32 v = ((UInt32)p[4] << 24) | ((UInt32)p[3] << 16) | ((UInt32)p[2] << 8) | ((UInt32)p[1]); + UInt32 cur = ip + (UInt32)pos; + pos += 5; + if (encoding) + v += cur; + else + v -= cur; + if (mask != 0) + { + unsigned sh = (mask & 6) << 2; + if (Test86MSByte((Byte)(v >> sh))) + { + v ^= (((UInt32)0x100 << sh) - 1); + if (encoding) + v += cur; + else + v -= cur; + } + mask = 0; + } + p[1] = (Byte)v; + p[2] = (Byte)(v >> 8); + p[3] = (Byte)(v >> 16); + p[4] = (Byte)(0 - ((v >> 24) & 1)); + } + else + { + mask = (mask >> 1) | 4; + pos++; + } + } +} diff --git a/dep/lzma/src/BraIA64.c b/dep/lzma/src/BraIA64.c new file mode 100644 index 000000000..d1dbc62c5 --- /dev/null +++ b/dep/lzma/src/BraIA64.c @@ -0,0 +1,53 @@ +/* BraIA64.c -- Converter for IA-64 code +2017-01-26 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "Bra.h" + +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 16) + return 0; + size -= 16; + i = 0; + do + { + unsigned m = ((UInt32)0x334B0000 >> (data[i] & 0x1E)) & 3; + if (m) + { + m++; + do + { + Byte *p = data + (i + (size_t)m * 5 - 8); + if (((p[3] >> m) & 15) == 5 + && (((p[-1] | ((UInt32)p[0] << 8)) >> m) & 0x70) == 0) + { + unsigned raw = GetUi32(p); + unsigned v = raw >> m; + v = (v & 0xFFFFF) | ((v & (1 << 23)) >> 3); + + v <<= 4; + if (encoding) + v += ip + (UInt32)i; + else + v -= ip + (UInt32)i; + v >>= 4; + + v &= 0x1FFFFF; + v += 0x700000; + v &= 0x8FFFFF; + raw &= ~((UInt32)0x8FFFFF << m); + raw |= (v << m); + SetUi32(p, raw); + } + } + while (++m <= 4); + } + i += 16; + } + while (i <= size); + return i; +} diff --git a/dep/lzma/src/CpuArch.c b/dep/lzma/src/CpuArch.c new file mode 100644 index 000000000..02e482e08 --- /dev/null +++ b/dep/lzma/src/CpuArch.c @@ -0,0 +1,218 @@ +/* CpuArch.c -- CPU specific code +2018-02-18: Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + +#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__) +#define USE_ASM +#endif + +#if !defined(USE_ASM) && _MSC_VER >= 1500 +#include +#endif + +#if defined(USE_ASM) && !defined(MY_CPU_AMD64) +static UInt32 CheckFlag(UInt32 flag) +{ + #ifdef _MSC_VER + __asm pushfd; + __asm pop EAX; + __asm mov EDX, EAX; + __asm xor EAX, flag; + __asm push EAX; + __asm popfd; + __asm pushfd; + __asm pop EAX; + __asm xor EAX, EDX; + __asm push EDX; + __asm popfd; + __asm and flag, EAX; + #else + __asm__ __volatile__ ( + "pushf\n\t" + "pop %%EAX\n\t" + "movl %%EAX,%%EDX\n\t" + "xorl %0,%%EAX\n\t" + "push %%EAX\n\t" + "popf\n\t" + "pushf\n\t" + "pop %%EAX\n\t" + "xorl %%EDX,%%EAX\n\t" + "push %%EDX\n\t" + "popf\n\t" + "andl %%EAX, %0\n\t": + "=c" (flag) : "c" (flag) : + "%eax", "%edx"); + #endif + return flag; +} +#define CHECK_CPUID_IS_SUPPORTED if (CheckFlag(1 << 18) == 0 || CheckFlag(1 << 21) == 0) return False; +#else +#define CHECK_CPUID_IS_SUPPORTED +#endif + +void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d) +{ + #ifdef USE_ASM + + #ifdef _MSC_VER + + UInt32 a2, b2, c2, d2; + __asm xor EBX, EBX; + __asm xor ECX, ECX; + __asm xor EDX, EDX; + __asm mov EAX, function; + __asm cpuid; + __asm mov a2, EAX; + __asm mov b2, EBX; + __asm mov c2, ECX; + __asm mov d2, EDX; + + *a = a2; + *b = b2; + *c = c2; + *d = d2; + + #else + + __asm__ __volatile__ ( + #if defined(MY_CPU_AMD64) && defined(__PIC__) + "mov %%rbx, %%rdi;" + "cpuid;" + "xchg %%rbx, %%rdi;" + : "=a" (*a) , + "=D" (*b) , + #elif defined(MY_CPU_X86) && defined(__PIC__) + "mov %%ebx, %%edi;" + "cpuid;" + "xchgl %%ebx, %%edi;" + : "=a" (*a) , + "=D" (*b) , + #else + "cpuid" + : "=a" (*a) , + "=b" (*b) , + #endif + "=c" (*c) , + "=d" (*d) + : "0" (function)) ; + + #endif + + #else + + int CPUInfo[4]; + __cpuid(CPUInfo, function); + *a = CPUInfo[0]; + *b = CPUInfo[1]; + *c = CPUInfo[2]; + *d = CPUInfo[3]; + + #endif +} + +BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p) +{ + CHECK_CPUID_IS_SUPPORTED + MyCPUID(0, &p->maxFunc, &p->vendor[0], &p->vendor[2], &p->vendor[1]); + MyCPUID(1, &p->ver, &p->b, &p->c, &p->d); + return True; +} + +static const UInt32 kVendors[][3] = +{ + { 0x756E6547, 0x49656E69, 0x6C65746E}, + { 0x68747541, 0x69746E65, 0x444D4163}, + { 0x746E6543, 0x48727561, 0x736C7561} +}; + +int x86cpuid_GetFirm(const Cx86cpuid *p) +{ + unsigned i; + for (i = 0; i < sizeof(kVendors) / sizeof(kVendors[i]); i++) + { + const UInt32 *v = kVendors[i]; + if (v[0] == p->vendor[0] && + v[1] == p->vendor[1] && + v[2] == p->vendor[2]) + return (int)i; + } + return -1; +} + +BoolInt CPU_Is_InOrder() +{ + Cx86cpuid p; + int firm; + UInt32 family, model; + if (!x86cpuid_CheckAndRead(&p)) + return True; + + family = x86cpuid_GetFamily(p.ver); + model = x86cpuid_GetModel(p.ver); + + firm = x86cpuid_GetFirm(&p); + + switch (firm) + { + case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && ( + /* In-Order Atom CPU */ + model == 0x1C /* 45 nm, N4xx, D4xx, N5xx, D5xx, 230, 330 */ + || model == 0x26 /* 45 nm, Z6xx */ + || model == 0x27 /* 32 nm, Z2460 */ + || model == 0x35 /* 32 nm, Z2760 */ + || model == 0x36 /* 32 nm, N2xxx, D2xxx */ + ))); + case CPU_FIRM_AMD: return (family < 5 || (family == 5 && (model < 6 || model == 0xA))); + case CPU_FIRM_VIA: return (family < 6 || (family == 6 && model < 0xF)); + } + return True; +} + +#if !defined(MY_CPU_AMD64) && defined(_WIN32) +#include +static BoolInt CPU_Sys_Is_SSE_Supported() +{ + OSVERSIONINFO vi; + vi.dwOSVersionInfoSize = sizeof(vi); + if (!GetVersionEx(&vi)) + return False; + return (vi.dwMajorVersion >= 5); +} +#define CHECK_SYS_SSE_SUPPORT if (!CPU_Sys_Is_SSE_Supported()) return False; +#else +#define CHECK_SYS_SSE_SUPPORT +#endif + +BoolInt CPU_Is_Aes_Supported() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + if (!x86cpuid_CheckAndRead(&p)) + return False; + return (p.c >> 25) & 1; +} + +BoolInt CPU_IsSupported_PageGB() +{ + Cx86cpuid cpuid; + if (!x86cpuid_CheckAndRead(&cpuid)) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(0x80000000, &d[0], &d[1], &d[2], &d[3]); + if (d[0] < 0x80000001) + return False; + } + { + UInt32 d[4] = { 0 }; + MyCPUID(0x80000001, &d[0], &d[1], &d[2], &d[3]); + return (d[3] >> 26) & 1; + } +} + +#endif diff --git a/dep/lzma/src/Delta.c b/dep/lzma/src/Delta.c new file mode 100644 index 000000000..e3edd21ed --- /dev/null +++ b/dep/lzma/src/Delta.c @@ -0,0 +1,64 @@ +/* Delta.c -- Delta converter +2009-05-26 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Delta.h" + +void Delta_Init(Byte *state) +{ + unsigned i; + for (i = 0; i < DELTA_STATE_SIZE; i++) + state[i] = 0; +} + +static void MyMemCpy(Byte *dest, const Byte *src, unsigned size) +{ + unsigned i; + for (i = 0; i < size; i++) + dest[i] = src[i]; +} + +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + Byte buf[DELTA_STATE_SIZE]; + unsigned j = 0; + MyMemCpy(buf, state, delta); + { + SizeT i; + for (i = 0; i < size;) + { + for (j = 0; j < delta && i < size; i++, j++) + { + Byte b = data[i]; + data[i] = (Byte)(b - buf[j]); + buf[j] = b; + } + } + } + if (j == delta) + j = 0; + MyMemCpy(state, buf + j, delta - j); + MyMemCpy(state + delta - j, buf, j); +} + +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + Byte buf[DELTA_STATE_SIZE]; + unsigned j = 0; + MyMemCpy(buf, state, delta); + { + SizeT i; + for (i = 0; i < size;) + { + for (j = 0; j < delta && i < size; i++, j++) + { + buf[j] = data[i] = (Byte)(buf[j] + data[i]); + } + } + } + if (j == delta) + j = 0; + MyMemCpy(state, buf + j, delta - j); + MyMemCpy(state + delta - j, buf, j); +} diff --git a/dep/lzma/src/LzFind.c b/dep/lzma/src/LzFind.c new file mode 100644 index 000000000..df55e86c1 --- /dev/null +++ b/dep/lzma/src/LzFind.c @@ -0,0 +1,1127 @@ +/* LzFind.c -- Match finder for LZ algorithms +2018-07-08 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "LzFind.h" +#include "LzHash.h" + +#define kEmptyHashValue 0 +#define kMaxValForNormalize ((UInt32)0xFFFFFFFF) +#define kNormalizeStepMin (1 << 10) /* it must be power of 2 */ +#define kNormalizeMask (~(UInt32)(kNormalizeStepMin - 1)) +#define kMaxHistorySize ((UInt32)7 << 29) + +#define kStartMaxLen 3 + +static void LzInWindow_Free(CMatchFinder *p, ISzAllocPtr alloc) +{ + if (!p->directInput) + { + ISzAlloc_Free(alloc, p->bufferBase); + p->bufferBase = NULL; + } +} + +/* keepSizeBefore + keepSizeAfter + keepSizeReserv must be < 4G) */ + +static int LzInWindow_Create(CMatchFinder *p, UInt32 keepSizeReserv, ISzAllocPtr alloc) +{ + UInt32 blockSize = p->keepSizeBefore + p->keepSizeAfter + keepSizeReserv; + if (p->directInput) + { + p->blockSize = blockSize; + return 1; + } + if (!p->bufferBase || p->blockSize != blockSize) + { + LzInWindow_Free(p, alloc); + p->blockSize = blockSize; + p->bufferBase = (Byte *)ISzAlloc_Alloc(alloc, (size_t)blockSize); + } + return (p->bufferBase != NULL); +} + +Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } + +UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return p->streamPos - p->pos; } + +void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue) +{ + p->posLimit -= subValue; + p->pos -= subValue; + p->streamPos -= subValue; +} + +static void MatchFinder_ReadBlock(CMatchFinder *p) +{ + if (p->streamEndWasReached || p->result != SZ_OK) + return; + + /* We use (p->streamPos - p->pos) value. (p->streamPos < p->pos) is allowed. */ + + if (p->directInput) + { + UInt32 curSize = 0xFFFFFFFF - (p->streamPos - p->pos); + if (curSize > p->directInputRem) + curSize = (UInt32)p->directInputRem; + p->directInputRem -= curSize; + p->streamPos += curSize; + if (p->directInputRem == 0) + p->streamEndWasReached = 1; + return; + } + + for (;;) + { + Byte *dest = p->buffer + (p->streamPos - p->pos); + size_t size = (p->bufferBase + p->blockSize - dest); + if (size == 0) + return; + + p->result = ISeqInStream_Read(p->stream, dest, &size); + if (p->result != SZ_OK) + return; + if (size == 0) + { + p->streamEndWasReached = 1; + return; + } + p->streamPos += (UInt32)size; + if (p->streamPos - p->pos > p->keepSizeAfter) + return; + } +} + +void MatchFinder_MoveBlock(CMatchFinder *p) +{ + memmove(p->bufferBase, + p->buffer - p->keepSizeBefore, + (size_t)(p->streamPos - p->pos) + p->keepSizeBefore); + p->buffer = p->bufferBase + p->keepSizeBefore; +} + +int MatchFinder_NeedMove(CMatchFinder *p) +{ + if (p->directInput) + return 0; + /* if (p->streamEndWasReached) return 0; */ + return ((size_t)(p->bufferBase + p->blockSize - p->buffer) <= p->keepSizeAfter); +} + +void MatchFinder_ReadIfRequired(CMatchFinder *p) +{ + if (p->streamEndWasReached) + return; + if (p->keepSizeAfter >= p->streamPos - p->pos) + MatchFinder_ReadBlock(p); +} + +static void MatchFinder_CheckAndMoveAndRead(CMatchFinder *p) +{ + if (MatchFinder_NeedMove(p)) + MatchFinder_MoveBlock(p); + MatchFinder_ReadBlock(p); +} + +static void MatchFinder_SetDefaultSettings(CMatchFinder *p) +{ + p->cutValue = 32; + p->btMode = 1; + p->numHashBytes = 4; + p->bigHash = 0; +} + +#define kCrcPoly 0xEDB88320 + +void MatchFinder_Construct(CMatchFinder *p) +{ + unsigned i; + p->bufferBase = NULL; + p->directInput = 0; + p->hash = NULL; + p->expectedDataSize = (UInt64)(Int64)-1; + MatchFinder_SetDefaultSettings(p); + + for (i = 0; i < 256; i++) + { + UInt32 r = (UInt32)i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); + p->crc[i] = r; + } +} + +static void MatchFinder_FreeThisClassMemory(CMatchFinder *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->hash); + p->hash = NULL; +} + +void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc) +{ + MatchFinder_FreeThisClassMemory(p, alloc); + LzInWindow_Free(p, alloc); +} + +static CLzRef* AllocRefs(size_t num, ISzAllocPtr alloc) +{ + size_t sizeInBytes = (size_t)num * sizeof(CLzRef); + if (sizeInBytes / sizeof(CLzRef) != num) + return NULL; + return (CLzRef *)ISzAlloc_Alloc(alloc, sizeInBytes); +} + +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAllocPtr alloc) +{ + UInt32 sizeReserv; + + if (historySize > kMaxHistorySize) + { + MatchFinder_Free(p, alloc); + return 0; + } + + sizeReserv = historySize >> 1; + if (historySize >= ((UInt32)3 << 30)) sizeReserv = historySize >> 3; + else if (historySize >= ((UInt32)2 << 30)) sizeReserv = historySize >> 2; + + sizeReserv += (keepAddBufferBefore + matchMaxLen + keepAddBufferAfter) / 2 + (1 << 19); + + p->keepSizeBefore = historySize + keepAddBufferBefore + 1; + p->keepSizeAfter = matchMaxLen + keepAddBufferAfter; + + /* we need one additional byte, since we use MoveBlock after pos++ and before dictionary using */ + + if (LzInWindow_Create(p, sizeReserv, alloc)) + { + UInt32 newCyclicBufferSize = historySize + 1; + UInt32 hs; + p->matchMaxLen = matchMaxLen; + { + p->fixedHashSize = 0; + if (p->numHashBytes == 2) + hs = (1 << 16) - 1; + else + { + hs = historySize; + if (hs > p->expectedDataSize) + hs = (UInt32)p->expectedDataSize; + if (hs != 0) + hs--; + hs |= (hs >> 1); + hs |= (hs >> 2); + hs |= (hs >> 4); + hs |= (hs >> 8); + hs >>= 1; + hs |= 0xFFFF; /* don't change it! It's required for Deflate */ + if (hs > (1 << 24)) + { + if (p->numHashBytes == 3) + hs = (1 << 24) - 1; + else + hs >>= 1; + /* if (bigHash) mode, GetHeads4b() in LzFindMt.c needs (hs >= ((1 << 24) - 1))) */ + } + } + p->hashMask = hs; + hs++; + if (p->numHashBytes > 2) p->fixedHashSize += kHash2Size; + if (p->numHashBytes > 3) p->fixedHashSize += kHash3Size; + if (p->numHashBytes > 4) p->fixedHashSize += kHash4Size; + hs += p->fixedHashSize; + } + + { + size_t newSize; + size_t numSons; + p->historySize = historySize; + p->hashSizeSum = hs; + p->cyclicBufferSize = newCyclicBufferSize; + + numSons = newCyclicBufferSize; + if (p->btMode) + numSons <<= 1; + newSize = hs + numSons; + + if (p->hash && p->numRefs == newSize) + return 1; + + MatchFinder_FreeThisClassMemory(p, alloc); + p->numRefs = newSize; + p->hash = AllocRefs(newSize, alloc); + + if (p->hash) + { + p->son = p->hash + p->hashSizeSum; + return 1; + } + } + } + + MatchFinder_Free(p, alloc); + return 0; +} + +static void MatchFinder_SetLimits(CMatchFinder *p) +{ + UInt32 limit = kMaxValForNormalize - p->pos; + UInt32 limit2 = p->cyclicBufferSize - p->cyclicBufferPos; + + if (limit2 < limit) + limit = limit2; + limit2 = p->streamPos - p->pos; + + if (limit2 <= p->keepSizeAfter) + { + if (limit2 > 0) + limit2 = 1; + } + else + limit2 -= p->keepSizeAfter; + + if (limit2 < limit) + limit = limit2; + + { + UInt32 lenLimit = p->streamPos - p->pos; + if (lenLimit > p->matchMaxLen) + lenLimit = p->matchMaxLen; + p->lenLimit = lenLimit; + } + p->posLimit = p->pos + limit; +} + + +void MatchFinder_Init_LowHash(CMatchFinder *p) +{ + size_t i; + CLzRef *items = p->hash; + size_t numItems = p->fixedHashSize; + for (i = 0; i < numItems; i++) + items[i] = kEmptyHashValue; +} + + +void MatchFinder_Init_HighHash(CMatchFinder *p) +{ + size_t i; + CLzRef *items = p->hash + p->fixedHashSize; + size_t numItems = (size_t)p->hashMask + 1; + for (i = 0; i < numItems; i++) + items[i] = kEmptyHashValue; +} + + +void MatchFinder_Init_3(CMatchFinder *p, int readData) +{ + p->cyclicBufferPos = 0; + p->buffer = p->bufferBase; + p->pos = + p->streamPos = p->cyclicBufferSize; + p->result = SZ_OK; + p->streamEndWasReached = 0; + + if (readData) + MatchFinder_ReadBlock(p); + + MatchFinder_SetLimits(p); +} + + +void MatchFinder_Init(CMatchFinder *p) +{ + MatchFinder_Init_HighHash(p); + MatchFinder_Init_LowHash(p); + MatchFinder_Init_3(p, True); +} + + +static UInt32 MatchFinder_GetSubValue(CMatchFinder *p) +{ + return (p->pos - p->historySize - 1) & kNormalizeMask; +} + +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems) +{ + size_t i; + for (i = 0; i < numItems; i++) + { + UInt32 value = items[i]; + if (value <= subValue) + value = kEmptyHashValue; + else + value -= subValue; + items[i] = value; + } +} + +static void MatchFinder_Normalize(CMatchFinder *p) +{ + UInt32 subValue = MatchFinder_GetSubValue(p); + MatchFinder_Normalize3(subValue, p->hash, p->numRefs); + MatchFinder_ReduceOffsets(p, subValue); +} + + +MY_NO_INLINE +static void MatchFinder_CheckLimits(CMatchFinder *p) +{ + if (p->pos == kMaxValForNormalize) + MatchFinder_Normalize(p); + if (!p->streamEndWasReached && p->keepSizeAfter == p->streamPos - p->pos) + MatchFinder_CheckAndMoveAndRead(p); + if (p->cyclicBufferPos == p->cyclicBufferSize) + p->cyclicBufferPos = 0; + MatchFinder_SetLimits(p); +} + + +/* + (lenLimit > maxLen) +*/ +MY_FORCE_INLINE +static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *distances, unsigned maxLen) +{ + /* + son[_cyclicBufferPos] = curMatch; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + return distances; + { + const Byte *pb = cur - delta; + curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; + if (pb[maxLen] == cur[maxLen] && *pb == *cur) + { + UInt32 len = 0; + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = len; + *distances++ = len; + *distances++ = delta - 1; + if (len == lenLimit) + return distances; + } + } + } + } + */ + + const Byte *lim = cur + lenLimit; + son[_cyclicBufferPos] = curMatch; + do + { + UInt32 delta = pos - curMatch; + if (delta >= _cyclicBufferSize) + break; + { + ptrdiff_t diff; + curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; + diff = (ptrdiff_t)0 - delta; + if (cur[maxLen] == cur[maxLen + diff]) + { + const Byte *c = cur; + while (*c == c[diff]) + { + if (++c == lim) + { + distances[0] = (UInt32)(lim - cur); + distances[1] = delta - 1; + return distances + 2; + } + } + { + unsigned len = (unsigned)(c - cur); + if (maxLen < len) + { + maxLen = len; + distances[0] = (UInt32)len; + distances[1] = delta - 1; + distances += 2; + } + } + } + } + } + while (--cutValue); + + return distances; +} + + +MY_FORCE_INLINE +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *distances, UInt32 maxLen) +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + unsigned len0 = 0, len1 = 0; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + { + *ptr0 = *ptr1 = kEmptyHashValue; + return distances; + } + { + CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + unsigned len = (len0 < len1 ? len0 : len1); + UInt32 pair0 = pair[0]; + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = (UInt32)len; + *distances++ = (UInt32)len; + *distances++ = delta - 1; + if (len == lenLimit) + { + *ptr1 = pair0; + *ptr0 = pair[1]; + return distances; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + ptr1 = pair + 1; + curMatch = *ptr1; + len1 = len; + } + else + { + *ptr0 = curMatch; + ptr0 = pair; + curMatch = *ptr0; + len0 = len; + } + } + } +} + +static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + unsigned len0 = 0, len1 = 0; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + { + *ptr0 = *ptr1 = kEmptyHashValue; + return; + } + { + CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + unsigned len = (len0 < len1 ? len0 : len1); + if (pb[len] == cur[len]) + { + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + { + if (len == lenLimit) + { + *ptr1 = pair[0]; + *ptr0 = pair[1]; + return; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + ptr1 = pair + 1; + curMatch = *ptr1; + len1 = len; + } + else + { + *ptr0 = curMatch; + ptr0 = pair; + curMatch = *ptr0; + len0 = len; + } + } + } +} + +#define MOVE_POS \ + ++p->cyclicBufferPos; \ + p->buffer++; \ + if (++p->pos == p->posLimit) MatchFinder_CheckLimits(p); + +#define MOVE_POS_RET MOVE_POS return (UInt32)offset; + +static void MatchFinder_MovePos(CMatchFinder *p) { MOVE_POS; } + +#define GET_MATCHES_HEADER2(minLen, ret_op) \ + unsigned lenLimit; UInt32 hv; const Byte *cur; UInt32 curMatch; \ + lenLimit = (unsigned)p->lenLimit; { if (lenLimit < minLen) { MatchFinder_MovePos(p); ret_op; }} \ + cur = p->buffer; + +#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return 0) +#define SKIP_HEADER(minLen) GET_MATCHES_HEADER2(minLen, continue) + +#define MF_PARAMS(p) p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue + +#define GET_MATCHES_FOOTER(offset, maxLen) \ + offset = (unsigned)(GetMatchesSpec1((UInt32)lenLimit, curMatch, MF_PARAMS(p), \ + distances + offset, (UInt32)maxLen) - distances); MOVE_POS_RET; + +#define SKIP_FOOTER \ + SkipMatchesSpec((UInt32)lenLimit, curMatch, MF_PARAMS(p)); MOVE_POS; + +#define UPDATE_maxLen { \ + ptrdiff_t diff = (ptrdiff_t)0 - d2; \ + const Byte *c = cur + maxLen; \ + const Byte *lim = cur + lenLimit; \ + for (; c != lim; c++) if (*(c + diff) != *c) break; \ + maxLen = (unsigned)(c - cur); } + +static UInt32 Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + unsigned offset; + GET_MATCHES_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + offset = 0; + GET_MATCHES_FOOTER(offset, 1) +} + +UInt32 Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + unsigned offset; + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + offset = 0; + GET_MATCHES_FOOTER(offset, 2) +} + +static UInt32 Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, d2, pos; + unsigned maxLen, offset; + UInt32 *hash; + GET_MATCHES_HEADER(3) + + HASH3_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash[h2]; + + curMatch = (hash + kFix3HashSize)[hv]; + + hash[h2] = pos; + (hash + kFix3HashSize)[hv] = pos; + + maxLen = 2; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + UPDATE_maxLen + distances[0] = (UInt32)maxLen; + distances[1] = d2 - 1; + offset = 2; + if (maxLen == lenLimit) + { + SkipMatchesSpec((UInt32)lenLimit, curMatch, MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + GET_MATCHES_FOOTER(offset, maxLen) +} + +static UInt32 Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, d2, d3, pos; + unsigned maxLen, offset; + UInt32 *hash; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + + curMatch = (hash + kFix4HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[hv] = pos; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + maxLen = 2; + distances[0] = 2; + distances[1] = d2 - 1; + offset = 2; + } + + if (d2 != d3 && d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + maxLen = 3; + distances[(size_t)offset + 1] = d3 - 1; + offset += 2; + d2 = d3; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[(size_t)offset - 2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec((UInt32)lenLimit, curMatch, MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + if (maxLen < 3) + maxLen = 3; + + GET_MATCHES_FOOTER(offset, maxLen) +} + +/* +static UInt32 Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, h4, d2, d3, d4, maxLen, offset, pos; + UInt32 *hash; + GET_MATCHES_HEADER(5) + + HASH5_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + d4 = pos - (hash + kFix4HashSize)[h4]; + + curMatch = (hash + kFix5HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[h4] = pos; + (hash + kFix5HashSize)[hv] = pos; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = d2 - 1; + offset = 2; + if (*(cur - d2 + 2) == cur[2]) + distances[0] = maxLen = 3; + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[2] = maxLen = 3; + distances[3] = d3 - 1; + offset = 4; + d2 = d3; + } + } + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[0] = maxLen = 3; + distances[1] = d3 - 1; + offset = 2; + d2 = d3; + } + + if (d2 != d4 && d4 < p->cyclicBufferSize + && *(cur - d4) == *cur + && *(cur - d4 + 3) == *(cur + 3)) + { + maxLen = 4; + distances[(size_t)offset + 1] = d4 - 1; + offset += 2; + d2 = d4; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[(size_t)offset - 2] = maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(lenLimit, curMatch, MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + if (maxLen < 4) + maxLen = 4; + + GET_MATCHES_FOOTER(offset, maxLen) +} +*/ + +static UInt32 Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, d2, d3, pos; + unsigned maxLen, offset; + UInt32 *hash; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + curMatch = (hash + kFix4HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[hv] = pos; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + maxLen = 2; + distances[0] = 2; + distances[1] = d2 - 1; + offset = 2; + } + + if (d2 != d3 && d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + maxLen = 3; + distances[(size_t)offset + 1] = d3 - 1; + offset += 2; + d2 = d3; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[(size_t)offset - 2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + } + + if (maxLen < 3) + maxLen = 3; + + offset = (unsigned)(Hc_GetMatchesSpec(lenLimit, curMatch, MF_PARAMS(p), + distances + offset, maxLen) - (distances)); + MOVE_POS_RET +} + +/* +static UInt32 Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, h4, d2, d3, d4, maxLen, offset, pos + UInt32 *hash; + GET_MATCHES_HEADER(5) + + HASH5_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + d4 = pos - (hash + kFix4HashSize)[h4]; + + curMatch = (hash + kFix5HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[h4] = pos; + (hash + kFix5HashSize)[hv] = pos; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = d2 - 1; + offset = 2; + if (*(cur - d2 + 2) == cur[2]) + distances[0] = maxLen = 3; + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[2] = maxLen = 3; + distances[3] = d3 - 1; + offset = 4; + d2 = d3; + } + } + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[0] = maxLen = 3; + distances[1] = d3 - 1; + offset = 2; + d2 = d3; + } + + if (d2 != d4 && d4 < p->cyclicBufferSize + && *(cur - d4) == *cur + && *(cur - d4 + 3) == *(cur + 3)) + { + maxLen = 4; + distances[(size_t)offset + 1] = d4 - 1; + offset += 2; + d2 = d4; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[(size_t)offset - 2] = maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + } + + if (maxLen < 4) + maxLen = 4; + + offset = (UInt32)(Hc_GetMatchesSpec(lenLimit, curMatch, MF_PARAMS(p), + distances + offset, maxLen) - (distances)); + MOVE_POS_RET +} +*/ + +UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + unsigned offset; + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + offset = (unsigned)(Hc_GetMatchesSpec(lenLimit, curMatch, MF_PARAMS(p), + distances, 2) - (distances)); + MOVE_POS_RET +} + +static void Bt2_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + SKIP_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + SKIP_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2; + UInt32 *hash; + SKIP_HEADER(3) + HASH3_CALC; + hash = p->hash; + curMatch = (hash + kFix3HashSize)[hv]; + hash[h2] = + (hash + kFix3HashSize)[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3; + UInt32 *hash; + SKIP_HEADER(4) + HASH4_CALC; + hash = p->hash; + curMatch = (hash + kFix4HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +/* +static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3, h4; + UInt32 *hash; + SKIP_HEADER(5) + HASH5_CALC; + hash = p->hash; + curMatch = (hash + kFix5HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[h4] = + (hash + kFix5HashSize)[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} +*/ + +static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3; + UInt32 *hash; + SKIP_HEADER(4) + HASH4_CALC; + hash = p->hash; + curMatch = (hash + kFix4HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[hv] = p->pos; + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS + } + while (--num != 0); +} + +/* +static void Hc5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3, h4; + UInt32 *hash; + SKIP_HEADER(5) + HASH5_CALC; + hash = p->hash; + curMatch = hash + kFix5HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[h4] = + (hash + kFix5HashSize)[hv] = p->pos; + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS + } + while (--num != 0); +} +*/ + +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + SKIP_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS + } + while (--num != 0); +} + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable) +{ + vTable->Init = (Mf_Init_Func)MatchFinder_Init; + vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinder_GetNumAvailableBytes; + vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinder_GetPointerToCurrentPos; + if (!p->btMode) + { + /* if (p->numHashBytes <= 4) */ + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc4_MatchFinder_Skip; + } + /* + else + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc5_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc5_MatchFinder_Skip; + } + */ + } + else if (p->numHashBytes == 2) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt2_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt2_MatchFinder_Skip; + } + else if (p->numHashBytes == 3) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt3_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt3_MatchFinder_Skip; + } + else /* if (p->numHashBytes == 4) */ + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt4_MatchFinder_Skip; + } + /* + else + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt5_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt5_MatchFinder_Skip; + } + */ +} diff --git a/dep/lzma/src/Lzma86Dec.c b/dep/lzma/src/Lzma86Dec.c new file mode 100644 index 000000000..21031745c --- /dev/null +++ b/dep/lzma/src/Lzma86Dec.c @@ -0,0 +1,54 @@ +/* Lzma86Dec.c -- LZMA + x86 (BCJ) Filter Decoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Lzma86.h" + +#include "Alloc.h" +#include "Bra.h" +#include "LzmaDec.h" + +SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize) +{ + unsigned i; + if (srcLen < LZMA86_HEADER_SIZE) + return SZ_ERROR_INPUT_EOF; + *unpackSize = 0; + for (i = 0; i < sizeof(UInt64); i++) + *unpackSize += ((UInt64)src[LZMA86_SIZE_OFFSET + i]) << (8 * i); + return SZ_OK; +} + +SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen) +{ + SRes res; + int useFilter; + SizeT inSizePure; + ELzmaStatus status; + + if (*srcLen < LZMA86_HEADER_SIZE) + return SZ_ERROR_INPUT_EOF; + + useFilter = src[0]; + + if (useFilter > 1) + { + *destLen = 0; + return SZ_ERROR_UNSUPPORTED; + } + + inSizePure = *srcLen - LZMA86_HEADER_SIZE; + res = LzmaDecode(dest, destLen, src + LZMA86_HEADER_SIZE, &inSizePure, + src + 1, LZMA_PROPS_SIZE, LZMA_FINISH_ANY, &status, &g_Alloc); + *srcLen = inSizePure + LZMA86_HEADER_SIZE; + if (res != SZ_OK) + return res; + if (useFilter == 1) + { + UInt32 x86State; + x86_Convert_Init(x86State); + x86_Convert(dest, *destLen, 0, &x86State, 0); + } + return SZ_OK; +} diff --git a/dep/lzma/src/Lzma86Enc.c b/dep/lzma/src/Lzma86Enc.c new file mode 100644 index 000000000..2617bab8e --- /dev/null +++ b/dep/lzma/src/Lzma86Enc.c @@ -0,0 +1,106 @@ +/* Lzma86Enc.c -- LZMA + x86 (BCJ) Filter Encoder +2018-07-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "Lzma86.h" + +#include "Alloc.h" +#include "Bra.h" +#include "LzmaEnc.h" + +#define SZE_OUT_OVERFLOW SZE_DATA_ERROR + +int Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, + int level, UInt32 dictSize, int filterMode) +{ + size_t outSize2 = *destLen; + Byte *filteredStream; + BoolInt useFilter; + int mainResult = SZ_ERROR_OUTPUT_EOF; + CLzmaEncProps props; + LzmaEncProps_Init(&props); + props.level = level; + props.dictSize = dictSize; + + *destLen = 0; + if (outSize2 < LZMA86_HEADER_SIZE) + return SZ_ERROR_OUTPUT_EOF; + + { + int i; + UInt64 t = srcLen; + for (i = 0; i < 8; i++, t >>= 8) + dest[LZMA86_SIZE_OFFSET + i] = (Byte)t; + } + + filteredStream = 0; + useFilter = (filterMode != SZ_FILTER_NO); + if (useFilter) + { + if (srcLen != 0) + { + filteredStream = (Byte *)MyAlloc(srcLen); + if (filteredStream == 0) + return SZ_ERROR_MEM; + memcpy(filteredStream, src, srcLen); + } + { + UInt32 x86State; + x86_Convert_Init(x86State); + x86_Convert(filteredStream, srcLen, 0, &x86State, 1); + } + } + + { + size_t minSize = 0; + BoolInt bestIsFiltered = False; + + /* passes for SZ_FILTER_AUTO: + 0 - BCJ + LZMA + 1 - LZMA + 2 - BCJ + LZMA agaian, if pass 0 (BCJ + LZMA) is better. + */ + int numPasses = (filterMode == SZ_FILTER_AUTO) ? 3 : 1; + + int i; + for (i = 0; i < numPasses; i++) + { + size_t outSizeProcessed = outSize2 - LZMA86_HEADER_SIZE; + size_t outPropsSize = 5; + SRes curRes; + BoolInt curModeIsFiltered = (numPasses > 1 && i == numPasses - 1); + if (curModeIsFiltered && !bestIsFiltered) + break; + if (useFilter && i == 0) + curModeIsFiltered = True; + + curRes = LzmaEncode(dest + LZMA86_HEADER_SIZE, &outSizeProcessed, + curModeIsFiltered ? filteredStream : src, srcLen, + &props, dest + 1, &outPropsSize, 0, + NULL, &g_Alloc, &g_Alloc); + + if (curRes != SZ_ERROR_OUTPUT_EOF) + { + if (curRes != SZ_OK) + { + mainResult = curRes; + break; + } + if (outSizeProcessed <= minSize || mainResult != SZ_OK) + { + minSize = outSizeProcessed; + bestIsFiltered = curModeIsFiltered; + mainResult = SZ_OK; + } + } + } + dest[0] = (Byte)(bestIsFiltered ? 1 : 0); + *destLen = LZMA86_HEADER_SIZE + minSize; + } + if (useFilter) + MyFree(filteredStream); + return mainResult; +} diff --git a/dep/lzma/src/LzmaDec.c b/dep/lzma/src/LzmaDec.c new file mode 100644 index 000000000..ba3e1dd50 --- /dev/null +++ b/dep/lzma/src/LzmaDec.c @@ -0,0 +1,1185 @@ +/* LzmaDec.c -- LZMA Decoder +2018-07-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #include "CpuArch.h" */ +#include "LzmaDec.h" + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) +#define kNumMoveBits 5 + +#define RC_INIT_SIZE 5 + +#define NORMALIZE if (range < kTopValue) { range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0(p) ttt = *(p); NORMALIZE; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) +#define UPDATE_0(p) range = bound; *(p) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); +#define UPDATE_1(p) range -= bound; code -= bound; *(p) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); +#define GET_BIT2(p, i, A0, A1) IF_BIT_0(p) \ + { UPDATE_0(p); i = (i + i); A0; } else \ + { UPDATE_1(p); i = (i + i) + 1; A1; } + +#define TREE_GET_BIT(probs, i) { GET_BIT2(probs + i, i, ;, ;); } + +#define REV_BIT(p, i, A0, A1) IF_BIT_0(p + i) \ + { UPDATE_0(p + i); A0; } else \ + { UPDATE_1(p + i); A1; } +#define REV_BIT_VAR( p, i, m) REV_BIT(p, i, i += m; m += m, m += m; i += m; ) +#define REV_BIT_CONST(p, i, m) REV_BIT(p, i, i += m; , i += m * 2; ) +#define REV_BIT_LAST( p, i, m) REV_BIT(p, i, i -= m , ; ) + +#define TREE_DECODE(probs, limit, i) \ + { i = 1; do { TREE_GET_BIT(probs, i); } while (i < limit); i -= limit; } + +/* #define _LZMA_SIZE_OPT */ + +#ifdef _LZMA_SIZE_OPT +#define TREE_6_DECODE(probs, i) TREE_DECODE(probs, (1 << 6), i) +#else +#define TREE_6_DECODE(probs, i) \ + { i = 1; \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + i -= 0x40; } +#endif + +#define NORMAL_LITER_DEC TREE_GET_BIT(prob, symbol) +#define MATCHED_LITER_DEC \ + matchByte += matchByte; \ + bit = offs; \ + offs &= matchByte; \ + probLit = prob + (offs + bit + symbol); \ + GET_BIT2(probLit, symbol, offs ^= bit; , ;) + + + +#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_ERROR; range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0_CHECK(p) ttt = *(p); NORMALIZE_CHECK; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) +#define UPDATE_0_CHECK range = bound; +#define UPDATE_1_CHECK range -= bound; code -= bound; +#define GET_BIT2_CHECK(p, i, A0, A1) IF_BIT_0_CHECK(p) \ + { UPDATE_0_CHECK; i = (i + i); A0; } else \ + { UPDATE_1_CHECK; i = (i + i) + 1; A1; } +#define GET_BIT_CHECK(p, i) GET_BIT2_CHECK(p, i, ; , ;) +#define TREE_DECODE_CHECK(probs, limit, i) \ + { i = 1; do { GET_BIT_CHECK(probs + i, i) } while (i < limit); i -= limit; } + + +#define REV_BIT_CHECK(p, i, m) IF_BIT_0_CHECK(p + i) \ + { UPDATE_0_CHECK; i += m; m += m; } else \ + { UPDATE_1_CHECK; m += m; i += m; } + + +#define kNumPosBitsMax 4 +#define kNumPosStatesMax (1 << kNumPosBitsMax) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) + +#define LenLow 0 +#define LenHigh (LenLow + 2 * (kNumPosStatesMax << kLenNumLowBits)) +#define kNumLenProbs (LenHigh + kLenNumHighSymbols) + +#define LenChoice LenLow +#define LenChoice2 (LenLow + (1 << kLenNumLowBits)) + +#define kNumStates 12 +#define kNumStates2 16 +#define kNumLitStates 7 + +#define kStartPosModelIndex 4 +#define kEndPosModelIndex 14 +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +#define kNumPosSlotBits 6 +#define kNumLenToPosStates 4 + +#define kNumAlignBits 4 +#define kAlignTableSize (1 << kNumAlignBits) + +#define kMatchMinLen 2 +#define kMatchSpecLenStart (kMatchMinLen + kLenNumLowSymbols * 2 + kLenNumHighSymbols) + +/* External ASM code needs same CLzmaProb array layout. So don't change it. */ + +/* (probs_1664) is faster and better for code size at some platforms */ +/* +#ifdef MY_CPU_X86_OR_AMD64 +*/ +#define kStartOffset 1664 +#define GET_PROBS p->probs_1664 +/* +#define GET_PROBS p->probs + kStartOffset +#else +#define kStartOffset 0 +#define GET_PROBS p->probs +#endif +*/ + +#define SpecPos (-kStartOffset) +#define IsRep0Long (SpecPos + kNumFullDistances) +#define RepLenCoder (IsRep0Long + (kNumStates2 << kNumPosBitsMax)) +#define LenCoder (RepLenCoder + kNumLenProbs) +#define IsMatch (LenCoder + kNumLenProbs) +#define Align (IsMatch + (kNumStates2 << kNumPosBitsMax)) +#define IsRep (Align + kAlignTableSize) +#define IsRepG0 (IsRep + kNumStates) +#define IsRepG1 (IsRepG0 + kNumStates) +#define IsRepG2 (IsRepG1 + kNumStates) +#define PosSlot (IsRepG2 + kNumStates) +#define Literal (PosSlot + (kNumLenToPosStates << kNumPosSlotBits)) +#define NUM_BASE_PROBS (Literal + kStartOffset) + +#if Align != 0 && kStartOffset != 0 + #error Stop_Compiling_Bad_LZMA_kAlign +#endif + +#if NUM_BASE_PROBS != 1984 + #error Stop_Compiling_Bad_LZMA_PROBS +#endif + + +#define LZMA_LIT_SIZE 0x300 + +#define LzmaProps_GetNumProbs(p) (NUM_BASE_PROBS + ((UInt32)LZMA_LIT_SIZE << ((p)->lc + (p)->lp))) + + +#define CALC_POS_STATE(processedPos, pbMask) (((processedPos) & (pbMask)) << 4) +#define COMBINED_PS_STATE (posState + state) +#define GET_LEN_STATE (posState) + +#define LZMA_DIC_MIN (1 << 12) + +/* +p->remainLen : shows status of LZMA decoder: + < kMatchSpecLenStart : normal remain + = kMatchSpecLenStart : finished + = kMatchSpecLenStart + 1 : need init range coder + = kMatchSpecLenStart + 2 : need init range coder and state +*/ + +/* ---------- LZMA_DECODE_REAL ---------- */ +/* +LzmaDec_DecodeReal_3() can be implemented in external ASM file. +3 - is the code compatibility version of that function for check at link time. +*/ + +#define LZMA_DECODE_REAL LzmaDec_DecodeReal_3 + +/* +LZMA_DECODE_REAL() +In: + RangeCoder is normalized + if (p->dicPos == limit) + { + LzmaDec_TryDummy() was called before to exclude LITERAL and MATCH-REP cases. + So first symbol can be only MATCH-NON-REP. And if that MATCH-NON-REP symbol + is not END_OF_PAYALOAD_MARKER, then function returns error code. + } + +Processing: + first LZMA symbol will be decoded in any case + All checks for limits are at the end of main loop, + It will decode new LZMA-symbols while (p->buf < bufLimit && dicPos < limit), + RangeCoder is still without last normalization when (p->buf < bufLimit) is being checked. + +Out: + RangeCoder is normalized + Result: + SZ_OK - OK + SZ_ERROR_DATA - Error + p->remainLen: + < kMatchSpecLenStart : normal remain + = kMatchSpecLenStart : finished +*/ + + +#ifdef _LZMA_DEC_OPT + +int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit); + +#else + +static +int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + CLzmaProb *probs = GET_PROBS; + unsigned state = (unsigned)p->state; + UInt32 rep0 = p->reps[0], rep1 = p->reps[1], rep2 = p->reps[2], rep3 = p->reps[3]; + unsigned pbMask = ((unsigned)1 << (p->prop.pb)) - 1; + unsigned lc = p->prop.lc; + unsigned lpMask = ((unsigned)0x100 << p->prop.lp) - ((unsigned)0x100 >> lc); + + Byte *dic = p->dic; + SizeT dicBufSize = p->dicBufSize; + SizeT dicPos = p->dicPos; + + UInt32 processedPos = p->processedPos; + UInt32 checkDicSize = p->checkDicSize; + unsigned len = 0; + + const Byte *buf = p->buf; + UInt32 range = p->range; + UInt32 code = p->code; + + do + { + CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = CALC_POS_STATE(processedPos, pbMask); + + prob = probs + IsMatch + COMBINED_PS_STATE; + IF_BIT_0(prob) + { + unsigned symbol; + UPDATE_0(prob); + prob = probs + Literal; + if (processedPos != 0 || checkDicSize != 0) + prob += (UInt32)3 * ((((processedPos << 8) + dic[(dicPos == 0 ? dicBufSize : dicPos) - 1]) & lpMask) << lc); + processedPos++; + + if (state < kNumLitStates) + { + state -= (state < 4) ? state : 3; + symbol = 1; + #ifdef _LZMA_SIZE_OPT + do { NORMAL_LITER_DEC } while (symbol < 0x100); + #else + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + #endif + } + else + { + unsigned matchByte = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + unsigned offs = 0x100; + state -= (state < 10) ? 3 : 6; + symbol = 1; + #ifdef _LZMA_SIZE_OPT + do + { + unsigned bit; + CLzmaProb *probLit; + MATCHED_LITER_DEC + } + while (symbol < 0x100); + #else + { + unsigned bit; + CLzmaProb *probLit; + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + } + #endif + } + + dic[dicPos++] = (Byte)symbol; + continue; + } + + { + UPDATE_1(prob); + prob = probs + IsRep + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + state += kNumStates; + prob = probs + LenCoder; + } + else + { + UPDATE_1(prob); + /* + // that case was checked before with kBadRepCode + if (checkDicSize == 0 && processedPos == 0) + return SZ_ERROR_DATA; + */ + prob = probs + IsRepG0 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + prob = probs + IsRep0Long + COMBINED_PS_STATE; + IF_BIT_0(prob) + { + UPDATE_0(prob); + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + processedPos++; + state = state < kNumLitStates ? 9 : 11; + continue; + } + UPDATE_1(prob); + } + else + { + UInt32 distance; + UPDATE_1(prob); + prob = probs + IsRepG1 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + distance = rep1; + } + else + { + UPDATE_1(prob); + prob = probs + IsRepG2 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + distance = rep2; + } + else + { + UPDATE_1(prob); + distance = rep3; + rep3 = rep2; + } + rep2 = rep1; + } + rep1 = rep0; + rep0 = distance; + } + state = state < kNumLitStates ? 8 : 11; + prob = probs + RepLenCoder; + } + + #ifdef _LZMA_SIZE_OPT + { + unsigned lim, offset; + CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE; + offset = 0; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + offset = kLenNumLowSymbols; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + offset = kLenNumLowSymbols * 2; + lim = (1 << kLenNumHighBits); + } + } + TREE_DECODE(probLen, lim, len); + len += offset; + } + #else + { + CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE; + len = 1; + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + len -= 8; + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + len = 1; + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + TREE_DECODE(probLen, (1 << kLenNumHighBits), len); + len += kLenNumLowSymbols * 2; + } + } + } + #endif + + if (state >= kNumStates) + { + UInt32 distance; + prob = probs + PosSlot + + ((len < kNumLenToPosStates ? len : kNumLenToPosStates - 1) << kNumPosSlotBits); + TREE_6_DECODE(prob, distance); + if (distance >= kStartPosModelIndex) + { + unsigned posSlot = (unsigned)distance; + unsigned numDirectBits = (unsigned)(((distance >> 1) - 1)); + distance = (2 | (distance & 1)); + if (posSlot < kEndPosModelIndex) + { + distance <<= numDirectBits; + prob = probs + SpecPos; + { + UInt32 m = 1; + distance++; + do + { + REV_BIT_VAR(prob, distance, m); + } + while (--numDirectBits); + distance -= m; + } + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE + range >>= 1; + + { + UInt32 t; + code -= range; + t = (0 - ((UInt32)code >> 31)); /* (UInt32)((Int32)code >> 31) */ + distance = (distance << 1) + (t + 1); + code += range & t; + } + /* + distance <<= 1; + if (code >= range) + { + code -= range; + distance |= 1; + } + */ + } + while (--numDirectBits); + prob = probs + Align; + distance <<= kNumAlignBits; + { + unsigned i = 1; + REV_BIT_CONST(prob, i, 1); + REV_BIT_CONST(prob, i, 2); + REV_BIT_CONST(prob, i, 4); + REV_BIT_LAST (prob, i, 8); + distance |= i; + } + if (distance == (UInt32)0xFFFFFFFF) + { + len = kMatchSpecLenStart; + state -= kNumStates; + break; + } + } + } + + rep3 = rep2; + rep2 = rep1; + rep1 = rep0; + rep0 = distance + 1; + state = (state < kNumStates + kNumLitStates) ? kNumLitStates : kNumLitStates + 3; + if (distance >= (checkDicSize == 0 ? processedPos: checkDicSize)) + { + p->dicPos = dicPos; + return SZ_ERROR_DATA; + } + } + + len += kMatchMinLen; + + { + SizeT rem; + unsigned curLen; + SizeT pos; + + if ((rem = limit - dicPos) == 0) + { + p->dicPos = dicPos; + return SZ_ERROR_DATA; + } + + curLen = ((rem < len) ? (unsigned)rem : len); + pos = dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0); + + processedPos += (UInt32)curLen; + + len -= curLen; + if (curLen <= dicBufSize - pos) + { + Byte *dest = dic + dicPos; + ptrdiff_t src = (ptrdiff_t)pos - (ptrdiff_t)dicPos; + const Byte *lim = dest + curLen; + dicPos += (SizeT)curLen; + do + *(dest) = (Byte)*(dest + src); + while (++dest != lim); + } + else + { + do + { + dic[dicPos++] = dic[pos]; + if (++pos == dicBufSize) + pos = 0; + } + while (--curLen != 0); + } + } + } + } + while (dicPos < limit && buf < bufLimit); + + NORMALIZE; + + p->buf = buf; + p->range = range; + p->code = code; + p->remainLen = (UInt32)len; + p->dicPos = dicPos; + p->processedPos = processedPos; + p->reps[0] = rep0; + p->reps[1] = rep1; + p->reps[2] = rep2; + p->reps[3] = rep3; + p->state = (UInt32)state; + + return SZ_OK; +} +#endif + +static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit) +{ + if (p->remainLen != 0 && p->remainLen < kMatchSpecLenStart) + { + Byte *dic = p->dic; + SizeT dicPos = p->dicPos; + SizeT dicBufSize = p->dicBufSize; + unsigned len = (unsigned)p->remainLen; + SizeT rep0 = p->reps[0]; /* we use SizeT to avoid the BUG of VC14 for AMD64 */ + SizeT rem = limit - dicPos; + if (rem < len) + len = (unsigned)(rem); + + if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= len) + p->checkDicSize = p->prop.dicSize; + + p->processedPos += (UInt32)len; + p->remainLen -= (UInt32)len; + while (len != 0) + { + len--; + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + } + p->dicPos = dicPos; + } +} + + +#define kRange0 0xFFFFFFFF +#define kBound0 ((kRange0 >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1)) +#define kBadRepCode (kBound0 + (((kRange0 - kBound0) >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1))) +#if kBadRepCode != (0xC0000000 - 0x400) + #error Stop_Compiling_Bad_LZMA_Check +#endif + +static int MY_FAST_CALL LzmaDec_DecodeReal2(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + do + { + SizeT limit2 = limit; + if (p->checkDicSize == 0) + { + UInt32 rem = p->prop.dicSize - p->processedPos; + if (limit - p->dicPos > rem) + limit2 = p->dicPos + rem; + + if (p->processedPos == 0) + if (p->code >= kBadRepCode) + return SZ_ERROR_DATA; + } + + RINOK(LZMA_DECODE_REAL(p, limit2, bufLimit)); + + if (p->checkDicSize == 0 && p->processedPos >= p->prop.dicSize) + p->checkDicSize = p->prop.dicSize; + + LzmaDec_WriteRem(p, limit); + } + while (p->dicPos < limit && p->buf < bufLimit && p->remainLen < kMatchSpecLenStart); + + return 0; +} + +typedef enum +{ + DUMMY_ERROR, /* unexpected end of input stream */ + DUMMY_LIT, + DUMMY_MATCH, + DUMMY_REP +} ELzmaDummy; + +static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, SizeT inSize) +{ + UInt32 range = p->range; + UInt32 code = p->code; + const Byte *bufLimit = buf + inSize; + const CLzmaProb *probs = GET_PROBS; + unsigned state = (unsigned)p->state; + ELzmaDummy res; + + { + const CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = CALC_POS_STATE(p->processedPos, (1 << p->prop.pb) - 1); + + prob = probs + IsMatch + COMBINED_PS_STATE; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK + + /* if (bufLimit - buf >= 7) return DUMMY_LIT; */ + + prob = probs + Literal; + if (p->checkDicSize != 0 || p->processedPos != 0) + prob += ((UInt32)LZMA_LIT_SIZE * + ((((p->processedPos) & ((1 << (p->prop.lp)) - 1)) << p->prop.lc) + + (p->dic[(p->dicPos == 0 ? p->dicBufSize : p->dicPos) - 1] >> (8 - p->prop.lc)))); + + if (state < kNumLitStates) + { + unsigned symbol = 1; + do { GET_BIT_CHECK(prob + symbol, symbol) } while (symbol < 0x100); + } + else + { + unsigned matchByte = p->dic[p->dicPos - p->reps[0] + + (p->dicPos < p->reps[0] ? p->dicBufSize : 0)]; + unsigned offs = 0x100; + unsigned symbol = 1; + do + { + unsigned bit; + const CLzmaProb *probLit; + matchByte += matchByte; + bit = offs; + offs &= matchByte; + probLit = prob + (offs + bit + symbol); + GET_BIT2_CHECK(probLit, symbol, offs ^= bit; , ; ) + } + while (symbol < 0x100); + } + res = DUMMY_LIT; + } + else + { + unsigned len; + UPDATE_1_CHECK; + + prob = probs + IsRep + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + state = 0; + prob = probs + LenCoder; + res = DUMMY_MATCH; + } + else + { + UPDATE_1_CHECK; + res = DUMMY_REP; + prob = probs + IsRepG0 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + prob = probs + IsRep0Long + COMBINED_PS_STATE; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + NORMALIZE_CHECK; + return DUMMY_REP; + } + else + { + UPDATE_1_CHECK; + } + } + else + { + UPDATE_1_CHECK; + prob = probs + IsRepG1 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + } + else + { + UPDATE_1_CHECK; + prob = probs + IsRepG2 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + } + else + { + UPDATE_1_CHECK; + } + } + } + state = kNumStates; + prob = probs + RepLenCoder; + } + { + unsigned limit, offset; + const CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenLow + GET_LEN_STATE; + offset = 0; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenChoice2; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + offset = kLenNumLowSymbols; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenHigh; + offset = kLenNumLowSymbols * 2; + limit = 1 << kLenNumHighBits; + } + } + TREE_DECODE_CHECK(probLen, limit, len); + len += offset; + } + + if (state < 4) + { + unsigned posSlot; + prob = probs + PosSlot + + ((len < kNumLenToPosStates - 1 ? len : kNumLenToPosStates - 1) << + kNumPosSlotBits); + TREE_DECODE_CHECK(prob, 1 << kNumPosSlotBits, posSlot); + if (posSlot >= kStartPosModelIndex) + { + unsigned numDirectBits = ((posSlot >> 1) - 1); + + /* if (bufLimit - buf >= 8) return DUMMY_MATCH; */ + + if (posSlot < kEndPosModelIndex) + { + prob = probs + SpecPos + ((2 | (posSlot & 1)) << numDirectBits); + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE_CHECK + range >>= 1; + code -= range & (((code - range) >> 31) - 1); + /* if (code >= range) code -= range; */ + } + while (--numDirectBits); + prob = probs + Align; + numDirectBits = kNumAlignBits; + } + { + unsigned i = 1; + unsigned m = 1; + do + { + REV_BIT_CHECK(prob, i, m); + } + while (--numDirectBits); + } + } + } + } + } + NORMALIZE_CHECK; + return res; +} + + +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState) +{ + p->remainLen = kMatchSpecLenStart + 1; + p->tempBufSize = 0; + + if (initDic) + { + p->processedPos = 0; + p->checkDicSize = 0; + p->remainLen = kMatchSpecLenStart + 2; + } + if (initState) + p->remainLen = kMatchSpecLenStart + 2; +} + +void LzmaDec_Init(CLzmaDec *p) +{ + p->dicPos = 0; + LzmaDec_InitDicAndState(p, True, True); +} + + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *srcLen, + ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT inSize = *srcLen; + (*srcLen) = 0; + + *status = LZMA_STATUS_NOT_SPECIFIED; + + if (p->remainLen > kMatchSpecLenStart) + { + for (; inSize > 0 && p->tempBufSize < RC_INIT_SIZE; (*srcLen)++, inSize--) + p->tempBuf[p->tempBufSize++] = *src++; + if (p->tempBufSize != 0 && p->tempBuf[0] != 0) + return SZ_ERROR_DATA; + if (p->tempBufSize < RC_INIT_SIZE) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + p->code = + ((UInt32)p->tempBuf[1] << 24) + | ((UInt32)p->tempBuf[2] << 16) + | ((UInt32)p->tempBuf[3] << 8) + | ((UInt32)p->tempBuf[4]); + p->range = 0xFFFFFFFF; + p->tempBufSize = 0; + + if (p->remainLen > kMatchSpecLenStart + 1) + { + SizeT numProbs = LzmaProps_GetNumProbs(&p->prop); + SizeT i; + CLzmaProb *probs = p->probs; + for (i = 0; i < numProbs; i++) + probs[i] = kBitModelTotal >> 1; + p->reps[0] = p->reps[1] = p->reps[2] = p->reps[3] = 1; + p->state = 0; + } + + p->remainLen = 0; + } + + LzmaDec_WriteRem(p, dicLimit); + + while (p->remainLen != kMatchSpecLenStart) + { + int checkEndMarkNow = 0; + + if (p->dicPos >= dicLimit) + { + if (p->remainLen == 0 && p->code == 0) + { + *status = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK; + return SZ_OK; + } + if (finishMode == LZMA_FINISH_ANY) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_OK; + } + if (p->remainLen != 0) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_ERROR_DATA; + } + checkEndMarkNow = 1; + } + + if (p->tempBufSize == 0) + { + SizeT processed; + const Byte *bufLimit; + if (inSize < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + int dummyRes = LzmaDec_TryDummy(p, src, inSize); + if (dummyRes == DUMMY_ERROR) + { + memcpy(p->tempBuf, src, inSize); + p->tempBufSize = (unsigned)inSize; + (*srcLen) += inSize; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (checkEndMarkNow && dummyRes != DUMMY_MATCH) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_ERROR_DATA; + } + bufLimit = src; + } + else + bufLimit = src + inSize - LZMA_REQUIRED_INPUT_MAX; + p->buf = src; + if (LzmaDec_DecodeReal2(p, dicLimit, bufLimit) != 0) + return SZ_ERROR_DATA; + processed = (SizeT)(p->buf - src); + (*srcLen) += processed; + src += processed; + inSize -= processed; + } + else + { + unsigned rem = p->tempBufSize, lookAhead = 0; + while (rem < LZMA_REQUIRED_INPUT_MAX && lookAhead < inSize) + p->tempBuf[rem++] = src[lookAhead++]; + p->tempBufSize = rem; + if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + int dummyRes = LzmaDec_TryDummy(p, p->tempBuf, (SizeT)rem); + if (dummyRes == DUMMY_ERROR) + { + (*srcLen) += (SizeT)lookAhead; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (checkEndMarkNow && dummyRes != DUMMY_MATCH) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_ERROR_DATA; + } + } + p->buf = p->tempBuf; + if (LzmaDec_DecodeReal2(p, dicLimit, p->buf) != 0) + return SZ_ERROR_DATA; + + { + unsigned kkk = (unsigned)(p->buf - p->tempBuf); + if (rem < kkk) + return SZ_ERROR_FAIL; /* some internal error */ + rem -= kkk; + if (lookAhead < rem) + return SZ_ERROR_FAIL; /* some internal error */ + lookAhead -= rem; + } + (*srcLen) += (SizeT)lookAhead; + src += lookAhead; + inSize -= (SizeT)lookAhead; + p->tempBufSize = 0; + } + } + + if (p->code != 0) + return SZ_ERROR_DATA; + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return SZ_OK; +} + + +SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT outSize = *destLen; + SizeT inSize = *srcLen; + *srcLen = *destLen = 0; + for (;;) + { + SizeT inSizeCur = inSize, outSizeCur, dicPos; + ELzmaFinishMode curFinishMode; + SRes res; + if (p->dicPos == p->dicBufSize) + p->dicPos = 0; + dicPos = p->dicPos; + if (outSize > p->dicBufSize - dicPos) + { + outSizeCur = p->dicBufSize; + curFinishMode = LZMA_FINISH_ANY; + } + else + { + outSizeCur = dicPos + outSize; + curFinishMode = finishMode; + } + + res = LzmaDec_DecodeToDic(p, outSizeCur, src, &inSizeCur, curFinishMode, status); + src += inSizeCur; + inSize -= inSizeCur; + *srcLen += inSizeCur; + outSizeCur = p->dicPos - dicPos; + memcpy(dest, p->dic + dicPos, outSizeCur); + dest += outSizeCur; + outSize -= outSizeCur; + *destLen += outSizeCur; + if (res != 0) + return res; + if (outSizeCur == 0 || outSize == 0) + return SZ_OK; + } +} + +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->probs); + p->probs = NULL; +} + +static void LzmaDec_FreeDict(CLzmaDec *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->dic); + p->dic = NULL; +} + +void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc) +{ + LzmaDec_FreeProbs(p, alloc); + LzmaDec_FreeDict(p, alloc); +} + +SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size) +{ + UInt32 dicSize; + Byte d; + + if (size < LZMA_PROPS_SIZE) + return SZ_ERROR_UNSUPPORTED; + else + dicSize = data[1] | ((UInt32)data[2] << 8) | ((UInt32)data[3] << 16) | ((UInt32)data[4] << 24); + + if (dicSize < LZMA_DIC_MIN) + dicSize = LZMA_DIC_MIN; + p->dicSize = dicSize; + + d = data[0]; + if (d >= (9 * 5 * 5)) + return SZ_ERROR_UNSUPPORTED; + + p->lc = (Byte)(d % 9); + d /= 9; + p->pb = (Byte)(d / 5); + p->lp = (Byte)(d % 5); + + return SZ_OK; +} + +static SRes LzmaDec_AllocateProbs2(CLzmaDec *p, const CLzmaProps *propNew, ISzAllocPtr alloc) +{ + UInt32 numProbs = LzmaProps_GetNumProbs(propNew); + if (!p->probs || numProbs != p->numProbs) + { + LzmaDec_FreeProbs(p, alloc); + p->probs = (CLzmaProb *)ISzAlloc_Alloc(alloc, numProbs * sizeof(CLzmaProb)); + if (!p->probs) + return SZ_ERROR_MEM; + p->probs_1664 = p->probs + 1664; + p->numProbs = numProbs; + } + return SZ_OK; +} + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) +{ + CLzmaProps propNew; + RINOK(LzmaProps_Decode(&propNew, props, propsSize)); + RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); + p->prop = propNew; + return SZ_OK; +} + +SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) +{ + CLzmaProps propNew; + SizeT dicBufSize; + RINOK(LzmaProps_Decode(&propNew, props, propsSize)); + RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); + + { + UInt32 dictSize = propNew.dicSize; + SizeT mask = ((UInt32)1 << 12) - 1; + if (dictSize >= ((UInt32)1 << 30)) mask = ((UInt32)1 << 22) - 1; + else if (dictSize >= ((UInt32)1 << 22)) mask = ((UInt32)1 << 20) - 1;; + dicBufSize = ((SizeT)dictSize + mask) & ~mask; + if (dicBufSize < dictSize) + dicBufSize = dictSize; + } + + if (!p->dic || dicBufSize != p->dicBufSize) + { + LzmaDec_FreeDict(p, alloc); + p->dic = (Byte *)ISzAlloc_Alloc(alloc, dicBufSize); + if (!p->dic) + { + LzmaDec_FreeProbs(p, alloc); + return SZ_ERROR_MEM; + } + } + p->dicBufSize = dicBufSize; + p->prop = propNew; + return SZ_OK; +} + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAllocPtr alloc) +{ + CLzmaDec p; + SRes res; + SizeT outSize = *destLen, inSize = *srcLen; + *destLen = *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + if (inSize < RC_INIT_SIZE) + return SZ_ERROR_INPUT_EOF; + LzmaDec_Construct(&p); + RINOK(LzmaDec_AllocateProbs(&p, propData, propSize, alloc)); + p.dic = dest; + p.dicBufSize = outSize; + LzmaDec_Init(&p); + *srcLen = inSize; + res = LzmaDec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); + *destLen = p.dicPos; + if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) + res = SZ_ERROR_INPUT_EOF; + LzmaDec_FreeProbs(&p, alloc); + return res; +} diff --git a/dep/lzma/src/LzmaEnc.c b/dep/lzma/src/LzmaEnc.c new file mode 100644 index 000000000..46a0db000 --- /dev/null +++ b/dep/lzma/src/LzmaEnc.c @@ -0,0 +1,2976 @@ +/* LzmaEnc.c -- LZMA Encoder +2019-01-10: Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define SHOW_STAT */ +/* #define SHOW_STAT2 */ + +#if defined(SHOW_STAT) || defined(SHOW_STAT2) +#include +#endif + +#include "LzmaEnc.h" + +#include "LzFind.h" +#ifndef _7ZIP_ST +#include "LzFindMt.h" +#endif + +#ifdef SHOW_STAT +static unsigned g_STAT_OFFSET = 0; +#endif + +#define kLzmaMaxHistorySize ((UInt32)3 << 29) +/* #define kLzmaMaxHistorySize ((UInt32)7 << 29) */ + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) +#define kNumMoveBits 5 +#define kProbInitValue (kBitModelTotal >> 1) + +#define kNumMoveReducingBits 4 +#define kNumBitPriceShiftBits 4 +#define kBitPrice (1 << kNumBitPriceShiftBits) + +#define REP_LEN_COUNT 64 + +void LzmaEncProps_Init(CLzmaEncProps *p) +{ + p->level = 5; + p->dictSize = p->mc = 0; + p->reduceSize = (UInt64)(Int64)-1; + p->lc = p->lp = p->pb = p->algo = p->fb = p->btMode = p->numHashBytes = p->numThreads = -1; + p->writeEndMark = 0; +} + +void LzmaEncProps_Normalize(CLzmaEncProps *p) +{ + int level = p->level; + if (level < 0) level = 5; + p->level = level; + + if (p->dictSize == 0) p->dictSize = (level <= 5 ? (1 << (level * 2 + 14)) : (level <= 7 ? (1 << 25) : (1 << 26))); + if (p->dictSize > p->reduceSize) + { + unsigned i; + UInt32 reduceSize = (UInt32)p->reduceSize; + for (i = 11; i <= 30; i++) + { + if (reduceSize <= ((UInt32)2 << i)) { p->dictSize = ((UInt32)2 << i); break; } + if (reduceSize <= ((UInt32)3 << i)) { p->dictSize = ((UInt32)3 << i); break; } + } + } + + if (p->lc < 0) p->lc = 3; + if (p->lp < 0) p->lp = 0; + if (p->pb < 0) p->pb = 2; + + if (p->algo < 0) p->algo = (level < 5 ? 0 : 1); + if (p->fb < 0) p->fb = (level < 7 ? 32 : 64); + if (p->btMode < 0) p->btMode = (p->algo == 0 ? 0 : 1); + if (p->numHashBytes < 0) p->numHashBytes = 4; + if (p->mc == 0) p->mc = (16 + (p->fb >> 1)) >> (p->btMode ? 0 : 1); + + if (p->numThreads < 0) + p->numThreads = + #ifndef _7ZIP_ST + ((p->btMode && p->algo) ? 2 : 1); + #else + 1; + #endif +} + +UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2) +{ + CLzmaEncProps props = *props2; + LzmaEncProps_Normalize(&props); + return props.dictSize; +} + +#if (_MSC_VER >= 1400) +/* BSR code is fast for some new CPUs */ +/* #define LZMA_LOG_BSR */ +#endif + +#ifdef LZMA_LOG_BSR + +#define kDicLogSizeMaxCompress 32 + +#define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); res = (zz + zz) + ((pos >> (zz - 1)) & 1); } + +static unsigned GetPosSlot1(UInt32 pos) +{ + unsigned res; + BSR2_RET(pos, res); + return res; +} +#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } +#define GetPosSlot(pos, res) { if (pos < 2) res = pos; else BSR2_RET(pos, res); } + +#else + +#define kNumLogBits (9 + sizeof(size_t) / 2) +/* #define kNumLogBits (11 + sizeof(size_t) / 8 * 3) */ + +#define kDicLogSizeMaxCompress ((kNumLogBits - 1) * 2 + 7) + +static void LzmaEnc_FastPosInit(Byte *g_FastPos) +{ + unsigned slot; + g_FastPos[0] = 0; + g_FastPos[1] = 1; + g_FastPos += 2; + + for (slot = 2; slot < kNumLogBits * 2; slot++) + { + size_t k = ((size_t)1 << ((slot >> 1) - 1)); + size_t j; + for (j = 0; j < k; j++) + g_FastPos[j] = (Byte)slot; + g_FastPos += k; + } +} + +/* we can use ((limit - pos) >> 31) only if (pos < ((UInt32)1 << 31)) */ +/* +#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ + (0 - (((((UInt32)1 << (kNumLogBits + 6)) - 1) - pos) >> 31))); \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } +*/ + +/* +#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ + (0 - (((((UInt32)1 << (kNumLogBits)) - 1) - (pos >> 6)) >> 31))); \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } +*/ + +#define BSR2_RET(pos, res) { unsigned zz = (pos < (1 << (kNumLogBits + 6))) ? 6 : 6 + kNumLogBits - 1; \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } + +/* +#define BSR2_RET(pos, res) { res = (pos < (1 << (kNumLogBits + 6))) ? \ + p->g_FastPos[pos >> 6] + 12 : \ + p->g_FastPos[pos >> (6 + kNumLogBits - 1)] + (6 + (kNumLogBits - 1)) * 2; } +*/ + +#define GetPosSlot1(pos) p->g_FastPos[pos] +#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } +#define GetPosSlot(pos, res) { if (pos < kNumFullDistances) res = p->g_FastPos[pos & (kNumFullDistances - 1)]; else BSR2_RET(pos, res); } + +#endif + + +#define LZMA_NUM_REPS 4 + +typedef UInt16 CState; +typedef UInt16 CExtra; + +typedef struct +{ + UInt32 price; + CState state; + CExtra extra; + // 0 : normal + // 1 : LIT : MATCH + // > 1 : MATCH (extra-1) : LIT : REP0 (len) + UInt32 len; + UInt32 dist; + UInt32 reps[LZMA_NUM_REPS]; +} COptimal; + + +// 18.06 +#define kNumOpts (1 << 11) +#define kPackReserve (kNumOpts * 8) +// #define kNumOpts (1 << 12) +// #define kPackReserve (1 + kNumOpts * 2) + +#define kNumLenToPosStates 4 +#define kNumPosSlotBits 6 +#define kDicLogSizeMin 0 +#define kDicLogSizeMax 32 +#define kDistTableSizeMax (kDicLogSizeMax * 2) + +#define kNumAlignBits 4 +#define kAlignTableSize (1 << kNumAlignBits) +#define kAlignMask (kAlignTableSize - 1) + +#define kStartPosModelIndex 4 +#define kEndPosModelIndex 14 +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +typedef +#ifdef _LZMA_PROB32 + UInt32 +#else + UInt16 +#endif + CLzmaProb; + +#define LZMA_PB_MAX 4 +#define LZMA_LC_MAX 8 +#define LZMA_LP_MAX 4 + +#define LZMA_NUM_PB_STATES_MAX (1 << LZMA_PB_MAX) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) +#define kLenNumSymbolsTotal (kLenNumLowSymbols * 2 + kLenNumHighSymbols) + +#define LZMA_MATCH_LEN_MIN 2 +#define LZMA_MATCH_LEN_MAX (LZMA_MATCH_LEN_MIN + kLenNumSymbolsTotal - 1) + +#define kNumStates 12 + + +typedef struct +{ + CLzmaProb low[LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)]; + CLzmaProb high[kLenNumHighSymbols]; +} CLenEnc; + + +typedef struct +{ + unsigned tableSize; + UInt32 prices[LZMA_NUM_PB_STATES_MAX][kLenNumSymbolsTotal]; + // UInt32 prices1[LZMA_NUM_PB_STATES_MAX][kLenNumLowSymbols * 2]; + // UInt32 prices2[kLenNumSymbolsTotal]; +} CLenPriceEnc; + +#define GET_PRICE_LEN(p, posState, len) \ + ((p)->prices[posState][(size_t)(len) - LZMA_MATCH_LEN_MIN]) + +/* +#define GET_PRICE_LEN(p, posState, len) \ + ((p)->prices2[(size_t)(len) - 2] + ((p)->prices1[posState][((len) - 2) & (kLenNumLowSymbols * 2 - 1)] & (((len) - 2 - kLenNumLowSymbols * 2) >> 9))) +*/ + +typedef struct +{ + UInt32 range; + unsigned cache; + UInt64 low; + UInt64 cacheSize; + Byte *buf; + Byte *bufLim; + Byte *bufBase; + ISeqOutStream *outStream; + UInt64 processed; + SRes res; +} CRangeEnc; + + +typedef struct +{ + CLzmaProb *litProbs; + + unsigned state; + UInt32 reps[LZMA_NUM_REPS]; + + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances]; + + CLenEnc lenProbs; + CLenEnc repLenProbs; + +} CSaveState; + + +typedef UInt32 CProbPrice; + + +typedef struct +{ + void *matchFinderObj; + IMatchFinder matchFinder; + + unsigned optCur; + unsigned optEnd; + + unsigned longestMatchLen; + unsigned numPairs; + UInt32 numAvail; + + unsigned state; + unsigned numFastBytes; + unsigned additionalOffset; + UInt32 reps[LZMA_NUM_REPS]; + unsigned lpMask, pbMask; + CLzmaProb *litProbs; + CRangeEnc rc; + + UInt32 backRes; + + unsigned lc, lp, pb; + unsigned lclp; + + BoolInt fastMode; + BoolInt writeEndMark; + BoolInt finished; + BoolInt multiThread; + BoolInt needInit; + // BoolInt _maxMode; + + UInt64 nowPos64; + + unsigned matchPriceCount; + // unsigned alignPriceCount; + int repLenEncCounter; + + unsigned distTableSize; + + UInt32 dictSize; + SRes result; + + #ifndef _7ZIP_ST + BoolInt mtMode; + // begin of CMatchFinderMt is used in LZ thread + CMatchFinderMt matchFinderMt; + // end of CMatchFinderMt is used in BT and HASH threads + #endif + + CMatchFinder matchFinderBase; + + #ifndef _7ZIP_ST + Byte pad[128]; + #endif + + // LZ thread + CProbPrice ProbPrices[kBitModelTotal >> kNumMoveReducingBits]; + + UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2 + 1]; + + UInt32 alignPrices[kAlignTableSize]; + UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax]; + UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances]; + + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances]; + + CLenEnc lenProbs; + CLenEnc repLenProbs; + + #ifndef LZMA_LOG_BSR + Byte g_FastPos[1 << kNumLogBits]; + #endif + + CLenPriceEnc lenEnc; + CLenPriceEnc repLenEnc; + + COptimal opt[kNumOpts]; + + CSaveState saveState; + + #ifndef _7ZIP_ST + Byte pad2[128]; + #endif +} CLzmaEnc; + + + +#define COPY_ARR(dest, src, arr) memcpy(dest->arr, src->arr, sizeof(src->arr)); + +void LzmaEnc_SaveState(CLzmaEncHandle pp) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CSaveState *dest = &p->saveState; + + dest->state = p->state; + + dest->lenProbs = p->lenProbs; + dest->repLenProbs = p->repLenProbs; + + COPY_ARR(dest, p, reps); + + COPY_ARR(dest, p, posAlignEncoder); + COPY_ARR(dest, p, isRep); + COPY_ARR(dest, p, isRepG0); + COPY_ARR(dest, p, isRepG1); + COPY_ARR(dest, p, isRepG2); + COPY_ARR(dest, p, isMatch); + COPY_ARR(dest, p, isRep0Long); + COPY_ARR(dest, p, posSlotEncoder); + COPY_ARR(dest, p, posEncoders); + + memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << p->lclp) * sizeof(CLzmaProb)); +} + + +void LzmaEnc_RestoreState(CLzmaEncHandle pp) +{ + CLzmaEnc *dest = (CLzmaEnc *)pp; + const CSaveState *p = &dest->saveState; + + dest->state = p->state; + + dest->lenProbs = p->lenProbs; + dest->repLenProbs = p->repLenProbs; + + COPY_ARR(dest, p, reps); + + COPY_ARR(dest, p, posAlignEncoder); + COPY_ARR(dest, p, isRep); + COPY_ARR(dest, p, isRepG0); + COPY_ARR(dest, p, isRepG1); + COPY_ARR(dest, p, isRepG2); + COPY_ARR(dest, p, isMatch); + COPY_ARR(dest, p, isRep0Long); + COPY_ARR(dest, p, posSlotEncoder); + COPY_ARR(dest, p, posEncoders); + + memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << dest->lclp) * sizeof(CLzmaProb)); +} + + + +SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CLzmaEncProps props = *props2; + LzmaEncProps_Normalize(&props); + + if (props.lc > LZMA_LC_MAX + || props.lp > LZMA_LP_MAX + || props.pb > LZMA_PB_MAX + || props.dictSize > ((UInt64)1 << kDicLogSizeMaxCompress) + || props.dictSize > kLzmaMaxHistorySize) + return SZ_ERROR_PARAM; + + p->dictSize = props.dictSize; + { + unsigned fb = props.fb; + if (fb < 5) + fb = 5; + if (fb > LZMA_MATCH_LEN_MAX) + fb = LZMA_MATCH_LEN_MAX; + p->numFastBytes = fb; + } + p->lc = props.lc; + p->lp = props.lp; + p->pb = props.pb; + p->fastMode = (props.algo == 0); + // p->_maxMode = True; + p->matchFinderBase.btMode = (Byte)(props.btMode ? 1 : 0); + { + unsigned numHashBytes = 4; + if (props.btMode) + { + if (props.numHashBytes < 2) + numHashBytes = 2; + else if (props.numHashBytes < 4) + numHashBytes = props.numHashBytes; + } + p->matchFinderBase.numHashBytes = numHashBytes; + } + + p->matchFinderBase.cutValue = props.mc; + + p->writeEndMark = props.writeEndMark; + + #ifndef _7ZIP_ST + /* + if (newMultiThread != _multiThread) + { + ReleaseMatchFinder(); + _multiThread = newMultiThread; + } + */ + p->multiThread = (props.numThreads > 1); + #endif + + return SZ_OK; +} + + +void LzmaEnc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + p->matchFinderBase.expectedDataSize = expectedDataSiize; +} + + +#define kState_Start 0 +#define kState_LitAfterMatch 4 +#define kState_LitAfterRep 5 +#define kState_MatchAfterLit 7 +#define kState_RepAfterLit 8 + +static const Byte kLiteralNextStates[kNumStates] = {0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 4, 5}; +static const Byte kMatchNextStates[kNumStates] = {7, 7, 7, 7, 7, 7, 7, 10, 10, 10, 10, 10}; +static const Byte kRepNextStates[kNumStates] = {8, 8, 8, 8, 8, 8, 8, 11, 11, 11, 11, 11}; +static const Byte kShortRepNextStates[kNumStates]= {9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11}; + +#define IsLitState(s) ((s) < 7) +#define GetLenToPosState2(len) (((len) < kNumLenToPosStates - 1) ? (len) : kNumLenToPosStates - 1) +#define GetLenToPosState(len) (((len) < kNumLenToPosStates + 1) ? (len) - 2 : kNumLenToPosStates - 1) + +#define kInfinityPrice (1 << 30) + +static void RangeEnc_Construct(CRangeEnc *p) +{ + p->outStream = NULL; + p->bufBase = NULL; +} + +#define RangeEnc_GetProcessed(p) ((p)->processed + ((p)->buf - (p)->bufBase) + (p)->cacheSize) +#define RangeEnc_GetProcessed_sizet(p) ((size_t)(p)->processed + ((p)->buf - (p)->bufBase) + (size_t)(p)->cacheSize) + +#define RC_BUF_SIZE (1 << 16) + +static int RangeEnc_Alloc(CRangeEnc *p, ISzAllocPtr alloc) +{ + if (!p->bufBase) + { + p->bufBase = (Byte *)ISzAlloc_Alloc(alloc, RC_BUF_SIZE); + if (!p->bufBase) + return 0; + p->bufLim = p->bufBase + RC_BUF_SIZE; + } + return 1; +} + +static void RangeEnc_Free(CRangeEnc *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->bufBase); + p->bufBase = 0; +} + +static void RangeEnc_Init(CRangeEnc *p) +{ + /* Stream.Init(); */ + p->range = 0xFFFFFFFF; + p->cache = 0; + p->low = 0; + p->cacheSize = 0; + + p->buf = p->bufBase; + + p->processed = 0; + p->res = SZ_OK; +} + +MY_NO_INLINE static void RangeEnc_FlushStream(CRangeEnc *p) +{ + size_t num; + if (p->res != SZ_OK) + return; + num = p->buf - p->bufBase; + if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num)) + p->res = SZ_ERROR_WRITE; + p->processed += num; + p->buf = p->bufBase; +} + +MY_NO_INLINE static void MY_FAST_CALL RangeEnc_ShiftLow(CRangeEnc *p) +{ + UInt32 low = (UInt32)p->low; + unsigned high = (unsigned)(p->low >> 32); + p->low = (UInt32)(low << 8); + if (low < (UInt32)0xFF000000 || high != 0) + { + { + Byte *buf = p->buf; + *buf++ = (Byte)(p->cache + high); + p->cache = (unsigned)(low >> 24); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + if (p->cacheSize == 0) + return; + } + high += 0xFF; + for (;;) + { + Byte *buf = p->buf; + *buf++ = (Byte)(high); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + if (--p->cacheSize == 0) + return; + } + } + p->cacheSize++; +} + +static void RangeEnc_FlushData(CRangeEnc *p) +{ + int i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + +#define RC_NORM(p) if (range < kTopValue) { range <<= 8; RangeEnc_ShiftLow(p); } + +#define RC_BIT_PRE(p, prob) \ + ttt = *(prob); \ + newBound = (range >> kNumBitModelTotalBits) * ttt; + +// #define _LZMA_ENC_USE_BRANCH + +#ifdef _LZMA_ENC_USE_BRANCH + +#define RC_BIT(p, prob, bit) { \ + RC_BIT_PRE(p, prob) \ + if (bit == 0) { range = newBound; ttt += (kBitModelTotal - ttt) >> kNumMoveBits; } \ + else { (p)->low += newBound; range -= newBound; ttt -= ttt >> kNumMoveBits; } \ + *(prob) = (CLzmaProb)ttt; \ + RC_NORM(p) \ + } + +#else + +#define RC_BIT(p, prob, bit) { \ + UInt32 mask; \ + RC_BIT_PRE(p, prob) \ + mask = 0 - (UInt32)bit; \ + range &= mask; \ + mask &= newBound; \ + range -= mask; \ + (p)->low += mask; \ + mask = (UInt32)bit - 1; \ + range += newBound & mask; \ + mask &= (kBitModelTotal - ((1 << kNumMoveBits) - 1)); \ + mask += ((1 << kNumMoveBits) - 1); \ + ttt += (Int32)(mask - ttt) >> kNumMoveBits; \ + *(prob) = (CLzmaProb)ttt; \ + RC_NORM(p) \ + } + +#endif + + + + +#define RC_BIT_0_BASE(p, prob) \ + range = newBound; *(prob) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); + +#define RC_BIT_1_BASE(p, prob) \ + range -= newBound; (p)->low += newBound; *(prob) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); \ + +#define RC_BIT_0(p, prob) \ + RC_BIT_0_BASE(p, prob) \ + RC_NORM(p) + +#define RC_BIT_1(p, prob) \ + RC_BIT_1_BASE(p, prob) \ + RC_NORM(p) + +static void RangeEnc_EncodeBit_0(CRangeEnc *p, CLzmaProb *prob) +{ + UInt32 range, ttt, newBound; + range = p->range; + RC_BIT_PRE(p, prob) + RC_BIT_0(p, prob) + p->range = range; +} + +static void LitEnc_Encode(CRangeEnc *p, CLzmaProb *probs, UInt32 sym) +{ + UInt32 range = p->range; + sym |= 0x100; + do + { + UInt32 ttt, newBound; + // RangeEnc_EncodeBit(p, probs + (sym >> 8), (sym >> 7) & 1); + CLzmaProb *prob = probs + (sym >> 8); + UInt32 bit = (sym >> 7) & 1; + sym <<= 1; + RC_BIT(p, prob, bit); + } + while (sym < 0x10000); + p->range = range; +} + +static void LitEnc_EncodeMatched(CRangeEnc *p, CLzmaProb *probs, UInt32 sym, UInt32 matchByte) +{ + UInt32 range = p->range; + UInt32 offs = 0x100; + sym |= 0x100; + do + { + UInt32 ttt, newBound; + CLzmaProb *prob; + UInt32 bit; + matchByte <<= 1; + // RangeEnc_EncodeBit(p, probs + (offs + (matchByte & offs) + (sym >> 8)), (sym >> 7) & 1); + prob = probs + (offs + (matchByte & offs) + (sym >> 8)); + bit = (sym >> 7) & 1; + sym <<= 1; + offs &= ~(matchByte ^ sym); + RC_BIT(p, prob, bit); + } + while (sym < 0x10000); + p->range = range; +} + + + +static void LzmaEnc_InitPriceTables(CProbPrice *ProbPrices) +{ + UInt32 i; + for (i = 0; i < (kBitModelTotal >> kNumMoveReducingBits); i++) + { + const unsigned kCyclesBits = kNumBitPriceShiftBits; + UInt32 w = (i << kNumMoveReducingBits) + (1 << (kNumMoveReducingBits - 1)); + unsigned bitCount = 0; + unsigned j; + for (j = 0; j < kCyclesBits; j++) + { + w = w * w; + bitCount <<= 1; + while (w >= ((UInt32)1 << 16)) + { + w >>= 1; + bitCount++; + } + } + ProbPrices[i] = (CProbPrice)((kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount); + // printf("\n%3d: %5d", i, ProbPrices[i]); + } +} + + +#define GET_PRICE(prob, bit) \ + p->ProbPrices[((prob) ^ (unsigned)(((-(int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICEa(prob, bit) \ + ProbPrices[((prob) ^ (unsigned)((-((int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICE_0(prob) p->ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICE_1(prob) p->ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + +#define GET_PRICEa_0(prob) ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICEa_1(prob) ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + + +static UInt32 LitEnc_GetPrice(const CLzmaProb *probs, UInt32 sym, const CProbPrice *ProbPrices) +{ + UInt32 price = 0; + sym |= 0x100; + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[sym], bit); + } + while (sym >= 2); + return price; +} + + +static UInt32 LitEnc_Matched_GetPrice(const CLzmaProb *probs, UInt32 sym, UInt32 matchByte, const CProbPrice *ProbPrices) +{ + UInt32 price = 0; + UInt32 offs = 0x100; + sym |= 0x100; + do + { + matchByte <<= 1; + price += GET_PRICEa(probs[offs + (matchByte & offs) + (sym >> 8)], (sym >> 7) & 1); + sym <<= 1; + offs &= ~(matchByte ^ sym); + } + while (sym < 0x10000); + return price; +} + + +static void RcTree_ReverseEncode(CRangeEnc *rc, CLzmaProb *probs, unsigned numBits, unsigned sym) +{ + UInt32 range = rc->range; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + unsigned bit = sym & 1; + // RangeEnc_EncodeBit(rc, probs + m, bit); + sym >>= 1; + RC_BIT(rc, probs + m, bit); + m = (m << 1) | bit; + } + while (--numBits); + rc->range = range; +} + + + +static void LenEnc_Init(CLenEnc *p) +{ + unsigned i; + for (i = 0; i < (LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)); i++) + p->low[i] = kProbInitValue; + for (i = 0; i < kLenNumHighSymbols; i++) + p->high[i] = kProbInitValue; +} + +static void LenEnc_Encode(CLenEnc *p, CRangeEnc *rc, unsigned sym, unsigned posState) +{ + UInt32 range, ttt, newBound; + CLzmaProb *probs = p->low; + range = rc->range; + RC_BIT_PRE(rc, probs); + if (sym >= kLenNumLowSymbols) + { + RC_BIT_1(rc, probs); + probs += kLenNumLowSymbols; + RC_BIT_PRE(rc, probs); + if (sym >= kLenNumLowSymbols * 2) + { + RC_BIT_1(rc, probs); + rc->range = range; + // RcTree_Encode(rc, p->high, kLenNumHighBits, sym - kLenNumLowSymbols * 2); + LitEnc_Encode(rc, p->high, sym - kLenNumLowSymbols * 2); + return; + } + sym -= kLenNumLowSymbols; + } + + // RcTree_Encode(rc, probs + (posState << kLenNumLowBits), kLenNumLowBits, sym); + { + unsigned m; + unsigned bit; + RC_BIT_0(rc, probs); + probs += (posState << (1 + kLenNumLowBits)); + bit = (sym >> 2) ; RC_BIT(rc, probs + 1, bit); m = (1 << 1) + bit; + bit = (sym >> 1) & 1; RC_BIT(rc, probs + m, bit); m = (m << 1) + bit; + bit = sym & 1; RC_BIT(rc, probs + m, bit); + rc->range = range; + } +} + +static void SetPrices_3(const CLzmaProb *probs, UInt32 startPrice, UInt32 *prices, const CProbPrice *ProbPrices) +{ + unsigned i; + for (i = 0; i < 8; i += 2) + { + UInt32 price = startPrice; + UInt32 prob; + price += GET_PRICEa(probs[1 ], (i >> 2)); + price += GET_PRICEa(probs[2 + (i >> 2)], (i >> 1) & 1); + prob = probs[4 + (i >> 1)]; + prices[i ] = price + GET_PRICEa_0(prob); + prices[i + 1] = price + GET_PRICEa_1(prob); + } +} + + +MY_NO_INLINE static void MY_FAST_CALL LenPriceEnc_UpdateTables( + CLenPriceEnc *p, + unsigned numPosStates, + const CLenEnc *enc, + const CProbPrice *ProbPrices) +{ + UInt32 b; + + { + unsigned prob = enc->low[0]; + UInt32 a, c; + unsigned posState; + b = GET_PRICEa_1(prob); + a = GET_PRICEa_0(prob); + c = b + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); + for (posState = 0; posState < numPosStates; posState++) + { + UInt32 *prices = p->prices[posState]; + const CLzmaProb *probs = enc->low + (posState << (1 + kLenNumLowBits)); + SetPrices_3(probs, a, prices, ProbPrices); + SetPrices_3(probs + kLenNumLowSymbols, c, prices + kLenNumLowSymbols, ProbPrices); + } + } + + /* + { + unsigned i; + UInt32 b; + a = GET_PRICEa_0(enc->low[0]); + for (i = 0; i < kLenNumLowSymbols; i++) + p->prices2[i] = a; + a = GET_PRICEa_1(enc->low[0]); + b = a + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); + for (i = kLenNumLowSymbols; i < kLenNumLowSymbols * 2; i++) + p->prices2[i] = b; + a += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); + } + */ + + // p->counter = numSymbols; + // p->counter = 64; + + { + unsigned i = p->tableSize; + + if (i > kLenNumLowSymbols * 2) + { + const CLzmaProb *probs = enc->high; + UInt32 *prices = p->prices[0] + kLenNumLowSymbols * 2; + i -= kLenNumLowSymbols * 2 - 1; + i >>= 1; + b += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); + do + { + /* + p->prices2[i] = a + + // RcTree_GetPrice(enc->high, kLenNumHighBits, i - kLenNumLowSymbols * 2, ProbPrices); + LitEnc_GetPrice(probs, i - kLenNumLowSymbols * 2, ProbPrices); + */ + // UInt32 price = a + RcTree_GetPrice(probs, kLenNumHighBits - 1, sym, ProbPrices); + unsigned sym = --i + (1 << (kLenNumHighBits - 1)); + UInt32 price = b; + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[sym], bit); + } + while (sym >= 2); + + { + unsigned prob = probs[(size_t)i + (1 << (kLenNumHighBits - 1))]; + prices[(size_t)i * 2 ] = price + GET_PRICEa_0(prob); + prices[(size_t)i * 2 + 1] = price + GET_PRICEa_1(prob); + } + } + while (i); + + { + unsigned posState; + size_t num = (p->tableSize - kLenNumLowSymbols * 2) * sizeof(p->prices[0][0]); + for (posState = 1; posState < numPosStates; posState++) + memcpy(p->prices[posState] + kLenNumLowSymbols * 2, p->prices[0] + kLenNumLowSymbols * 2, num); + } + } + } +} + +/* + #ifdef SHOW_STAT + g_STAT_OFFSET += num; + printf("\n MovePos %u", num); + #endif +*/ + +#define MOVE_POS(p, num) { \ + p->additionalOffset += (num); \ + p->matchFinder.Skip(p->matchFinderObj, (UInt32)(num)); } + + +static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) +{ + unsigned numPairs; + + p->additionalOffset++; + p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); + numPairs = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); + *numPairsRes = numPairs; + + #ifdef SHOW_STAT + printf("\n i = %u numPairs = %u ", g_STAT_OFFSET, numPairs / 2); + g_STAT_OFFSET++; + { + unsigned i; + for (i = 0; i < numPairs; i += 2) + printf("%2u %6u | ", p->matches[i], p->matches[i + 1]); + } + #endif + + if (numPairs == 0) + return 0; + { + unsigned len = p->matches[(size_t)numPairs - 2]; + if (len != p->numFastBytes) + return len; + { + UInt32 numAvail = p->numAvail; + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + { + const Byte *p1 = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + const Byte *p2 = p1 + len; + ptrdiff_t dif = (ptrdiff_t)-1 - p->matches[(size_t)numPairs - 1]; + const Byte *lim = p1 + numAvail; + for (; p2 != lim && *p2 == p2[dif]; p2++) + {} + return (unsigned)(p2 - p1); + } + } + } +} + +#define MARK_LIT ((UInt32)(Int32)-1) + +#define MakeAs_Lit(p) { (p)->dist = MARK_LIT; (p)->extra = 0; } +#define MakeAs_ShortRep(p) { (p)->dist = 0; (p)->extra = 0; } +#define IsShortRep(p) ((p)->dist == 0) + + +#define GetPrice_ShortRep(p, state, posState) \ + ( GET_PRICE_0(p->isRepG0[state]) + GET_PRICE_0(p->isRep0Long[state][posState])) + +#define GetPrice_Rep_0(p, state, posState) ( \ + GET_PRICE_1(p->isMatch[state][posState]) \ + + GET_PRICE_1(p->isRep0Long[state][posState])) \ + + GET_PRICE_1(p->isRep[state]) \ + + GET_PRICE_0(p->isRepG0[state]) + +MY_FORCE_INLINE +static UInt32 GetPrice_PureRep(const CLzmaEnc *p, unsigned repIndex, size_t state, size_t posState) +{ + UInt32 price; + UInt32 prob = p->isRepG0[state]; + if (repIndex == 0) + { + price = GET_PRICE_0(prob); + price += GET_PRICE_1(p->isRep0Long[state][posState]); + } + else + { + price = GET_PRICE_1(prob); + prob = p->isRepG1[state]; + if (repIndex == 1) + price += GET_PRICE_0(prob); + else + { + price += GET_PRICE_1(prob); + price += GET_PRICE(p->isRepG2[state], repIndex - 2); + } + } + return price; +} + + +static unsigned Backward(CLzmaEnc *p, unsigned cur) +{ + unsigned wr = cur + 1; + p->optEnd = wr; + + for (;;) + { + UInt32 dist = p->opt[cur].dist; + unsigned len = (unsigned)p->opt[cur].len; + unsigned extra = (unsigned)p->opt[cur].extra; + cur -= len; + + if (extra) + { + wr--; + p->opt[wr].len = (UInt32)len; + cur -= extra; + len = extra; + if (extra == 1) + { + p->opt[wr].dist = dist; + dist = MARK_LIT; + } + else + { + p->opt[wr].dist = 0; + len--; + wr--; + p->opt[wr].dist = MARK_LIT; + p->opt[wr].len = 1; + } + } + + if (cur == 0) + { + p->backRes = dist; + p->optCur = wr; + return len; + } + + wr--; + p->opt[wr].dist = dist; + p->opt[wr].len = (UInt32)len; + } +} + + + +#define LIT_PROBS(pos, prevByte) \ + (p->litProbs + (UInt32)3 * (((((pos) << 8) + (prevByte)) & p->lpMask) << p->lc)) + + +static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) +{ + unsigned last, cur; + UInt32 reps[LZMA_NUM_REPS]; + unsigned repLens[LZMA_NUM_REPS]; + UInt32 *matches; + + { + UInt32 numAvail; + unsigned numPairs, mainLen, repMaxIndex, i, posState; + UInt32 matchPrice, repMatchPrice; + const Byte *data; + Byte curByte, matchByte; + + p->optCur = p->optEnd = 0; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLen; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + if (numAvail < 2) + { + p->backRes = MARK_LIT; + return 1; + } + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + repMaxIndex = 0; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len; + const Byte *data2; + reps[i] = p->reps[i]; + data2 = data - reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + { + repLens[i] = 0; + continue; + } + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + repLens[i] = len; + if (len > repLens[repMaxIndex]) + repMaxIndex = i; + } + + if (repLens[repMaxIndex] >= p->numFastBytes) + { + unsigned len; + p->backRes = (UInt32)repMaxIndex; + len = repLens[repMaxIndex]; + MOVE_POS(p, len - 1) + return len; + } + + matches = p->matches; + + if (mainLen >= p->numFastBytes) + { + p->backRes = matches[(size_t)numPairs - 1] + LZMA_NUM_REPS; + MOVE_POS(p, mainLen - 1) + return mainLen; + } + + curByte = *data; + matchByte = *(data - reps[0]); + + last = repLens[repMaxIndex]; + if (last <= mainLen) + last = mainLen; + + if (last < 2 && curByte != matchByte) + { + p->backRes = MARK_LIT; + return 1; + } + + p->opt[0].state = (CState)p->state; + + posState = (position & p->pbMask); + + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + p->opt[1].price = GET_PRICE_0(p->isMatch[p->state][posState]) + + (!IsLitState(p->state) ? + LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + } + + MakeAs_Lit(&p->opt[1]); + + matchPrice = GET_PRICE_1(p->isMatch[p->state][posState]); + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[p->state]); + + // 18.06 + if (matchByte == curByte && repLens[0] == 0) + { + UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, p->state, posState); + if (shortRepPrice < p->opt[1].price) + { + p->opt[1].price = shortRepPrice; + MakeAs_ShortRep(&p->opt[1]); + } + if (last < 2) + { + p->backRes = p->opt[1].dist; + return 1; + } + } + + p->opt[1].len = 1; + + p->opt[0].reps[0] = reps[0]; + p->opt[0].reps[1] = reps[1]; + p->opt[0].reps[2] = reps[2]; + p->opt[0].reps[3] = reps[3]; + + // ---------- REP ---------- + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned repLen = repLens[i]; + UInt32 price; + if (repLen < 2) + continue; + price = repMatchPrice + GetPrice_PureRep(p, i, p->state, posState); + do + { + UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, repLen); + COptimal *opt = &p->opt[repLen]; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)repLen; + opt->dist = (UInt32)i; + opt->extra = 0; + } + } + while (--repLen >= 2); + } + + + // ---------- MATCH ---------- + { + unsigned len = repLens[0] + 1; + if (len <= mainLen) + { + unsigned offs = 0; + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[p->state]); + + if (len < 2) + len = 2; + else + while (len > matches[offs]) + offs += 2; + + for (; ; len++) + { + COptimal *opt; + UInt32 dist = matches[(size_t)offs + 1]; + UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); + unsigned lenToPosState = GetLenToPosState(len); + + if (dist < kNumFullDistances) + price += p->distancesPrices[lenToPosState][dist & (kNumFullDistances - 1)]; + else + { + unsigned slot; + GetPosSlot2(dist, slot); + price += p->alignPrices[dist & kAlignMask]; + price += p->posSlotPrices[lenToPosState][slot]; + } + + opt = &p->opt[len]; + + if (price < opt->price) + { + opt->price = price; + opt->len = (UInt32)len; + opt->dist = dist + LZMA_NUM_REPS; + opt->extra = 0; + } + + if (len == matches[offs]) + { + offs += 2; + if (offs == numPairs) + break; + } + } + } + } + + + cur = 0; + + #ifdef SHOW_STAT2 + /* if (position >= 0) */ + { + unsigned i; + printf("\n pos = %4X", position); + for (i = cur; i <= last; i++) + printf("\nprice[%4X] = %u", position - cur + i, p->opt[i].price); + } + #endif + } + + + + // ---------- Optimal Parsing ---------- + + for (;;) + { + unsigned numAvail; + UInt32 numAvailFull; + unsigned newLen, numPairs, prev, state, posState, startLen; + UInt32 litPrice, matchPrice, repMatchPrice; + BoolInt nextIsLit; + Byte curByte, matchByte; + const Byte *data; + COptimal *curOpt, *nextOpt; + + if (++cur == last) + break; + + // 18.06 + if (cur >= kNumOpts - 64) + { + unsigned j, best; + UInt32 price = p->opt[cur].price; + best = cur; + for (j = cur + 1; j <= last; j++) + { + UInt32 price2 = p->opt[j].price; + if (price >= price2) + { + price = price2; + best = j; + } + } + { + unsigned delta = best - cur; + if (delta != 0) + { + MOVE_POS(p, delta); + } + } + cur = best; + break; + } + + newLen = ReadMatchDistances(p, &numPairs); + + if (newLen >= p->numFastBytes) + { + p->numPairs = numPairs; + p->longestMatchLen = newLen; + break; + } + + curOpt = &p->opt[cur]; + + position++; + + // we need that check here, if skip_items in p->opt are possible + /* + if (curOpt->price >= kInfinityPrice) + continue; + */ + + prev = cur - curOpt->len; + + if (curOpt->len == 1) + { + state = (unsigned)p->opt[prev].state; + if (IsShortRep(curOpt)) + state = kShortRepNextStates[state]; + else + state = kLiteralNextStates[state]; + } + else + { + const COptimal *prevOpt; + UInt32 b0; + UInt32 dist = curOpt->dist; + + if (curOpt->extra) + { + prev -= (unsigned)curOpt->extra; + state = kState_RepAfterLit; + if (curOpt->extra == 1) + state = (dist < LZMA_NUM_REPS ? kState_RepAfterLit : kState_MatchAfterLit); + } + else + { + state = (unsigned)p->opt[prev].state; + if (dist < LZMA_NUM_REPS) + state = kRepNextStates[state]; + else + state = kMatchNextStates[state]; + } + + prevOpt = &p->opt[prev]; + b0 = prevOpt->reps[0]; + + if (dist < LZMA_NUM_REPS) + { + if (dist == 0) + { + reps[0] = b0; + reps[1] = prevOpt->reps[1]; + reps[2] = prevOpt->reps[2]; + reps[3] = prevOpt->reps[3]; + } + else + { + reps[1] = b0; + b0 = prevOpt->reps[1]; + if (dist == 1) + { + reps[0] = b0; + reps[2] = prevOpt->reps[2]; + reps[3] = prevOpt->reps[3]; + } + else + { + reps[2] = b0; + reps[0] = prevOpt->reps[dist]; + reps[3] = prevOpt->reps[dist ^ 1]; + } + } + } + else + { + reps[0] = (dist - LZMA_NUM_REPS + 1); + reps[1] = b0; + reps[2] = prevOpt->reps[1]; + reps[3] = prevOpt->reps[2]; + } + } + + curOpt->state = (CState)state; + curOpt->reps[0] = reps[0]; + curOpt->reps[1] = reps[1]; + curOpt->reps[2] = reps[2]; + curOpt->reps[3] = reps[3]; + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + curByte = *data; + matchByte = *(data - reps[0]); + + posState = (position & p->pbMask); + + /* + The order of Price checks: + < LIT + <= SHORT_REP + < LIT : REP_0 + < REP [ : LIT : REP_0 ] + < MATCH [ : LIT : REP_0 ] + */ + + { + UInt32 curPrice = curOpt->price; + unsigned prob = p->isMatch[state][posState]; + matchPrice = curPrice + GET_PRICE_1(prob); + litPrice = curPrice + GET_PRICE_0(prob); + } + + nextOpt = &p->opt[(size_t)cur + 1]; + nextIsLit = False; + + // here we can allow skip_items in p->opt, if we don't check (nextOpt->price < kInfinityPrice) + // 18.new.06 + if ((nextOpt->price < kInfinityPrice + // && !IsLitState(state) + && matchByte == curByte) + || litPrice > nextOpt->price + ) + litPrice = 0; + else + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + litPrice += (!IsLitState(state) ? + LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + + if (litPrice < nextOpt->price) + { + nextOpt->price = litPrice; + nextOpt->len = 1; + MakeAs_Lit(nextOpt); + nextIsLit = True; + } + } + + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[state]); + + numAvailFull = p->numAvail; + { + unsigned temp = kNumOpts - 1 - cur; + if (numAvailFull > temp) + numAvailFull = (UInt32)temp; + } + + // 18.06 + // ---------- SHORT_REP ---------- + if (IsLitState(state)) // 18.new + if (matchByte == curByte) + if (repMatchPrice < nextOpt->price) // 18.new + // if (numAvailFull < 2 || data[1] != *(data - reps[0] + 1)) + if ( + // nextOpt->price >= kInfinityPrice || + nextOpt->len < 2 // we can check nextOpt->len, if skip items are not allowed in p->opt + || (nextOpt->dist != 0 + // && nextOpt->extra <= 1 // 17.old + ) + ) + { + UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, state, posState); + // if (shortRepPrice <= nextOpt->price) // 17.old + if (shortRepPrice < nextOpt->price) // 18.new + { + nextOpt->price = shortRepPrice; + nextOpt->len = 1; + MakeAs_ShortRep(nextOpt); + nextIsLit = False; + } + } + + if (numAvailFull < 2) + continue; + numAvail = (numAvailFull <= p->numFastBytes ? numAvailFull : p->numFastBytes); + + // numAvail <= p->numFastBytes + + // ---------- LIT : REP_0 ---------- + + if (!nextIsLit + && litPrice != 0 // 18.new + && matchByte != curByte + && numAvailFull > 2) + { + const Byte *data2 = data - reps[0]; + if (data[1] == data2[1] && data[2] == data2[2]) + { + unsigned len; + unsigned limit = p->numFastBytes + 1; + if (limit > numAvailFull) + limit = numAvailFull; + for (len = 3; len < limit && data[len] == data2[len]; len++) + {} + + { + unsigned state2 = kLiteralNextStates[state]; + unsigned posState2 = (position + 1) & p->pbMask; + UInt32 price = litPrice + GetPrice_Rep_0(p, state2, posState2); + { + unsigned offset = cur + len; + + if (last < offset) + last = offset; + + // do + { + UInt32 price2; + COptimal *opt; + len--; + // price2 = price + GetPrice_Len_Rep_0(p, len, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len); + + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len; + opt->dist = 0; + opt->extra = 1; + } + } + // while (len >= 3); + } + } + } + } + + startLen = 2; /* speed optimization */ + + { + // ---------- REP ---------- + unsigned repIndex = 0; // 17.old + // unsigned repIndex = IsLitState(state) ? 0 : 1; // 18.notused + for (; repIndex < LZMA_NUM_REPS; repIndex++) + { + unsigned len; + UInt32 price; + const Byte *data2 = data - reps[repIndex]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + + // if (len < startLen) continue; // 18.new: speed optimization + + { + unsigned offset = cur + len; + if (last < offset) + last = offset; + } + { + unsigned len2 = len; + price = repMatchPrice + GetPrice_PureRep(p, repIndex, state, posState); + do + { + UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, len2); + COptimal *opt = &p->opt[cur + len2]; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->dist = (UInt32)repIndex; + opt->extra = 0; + } + } + while (--len2 >= 2); + } + + if (repIndex == 0) startLen = len + 1; // 17.old + // startLen = len + 1; // 18.new + + /* if (_maxMode) */ + { + // ---------- REP : LIT : REP_0 ---------- + // numFastBytes + 1 + numFastBytes + + unsigned len2 = len + 1; + unsigned limit = len2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + + len2 += 2; + if (len2 <= limit) + if (data[len2 - 2] == data2[len2 - 2]) + if (data[len2 - 1] == data2[len2 - 1]) + { + unsigned state2 = kRepNextStates[state]; + unsigned posState2 = (position + len) & p->pbMask; + price += GET_PRICE_LEN(&p->repLenEnc, posState, len) + + GET_PRICE_0(p->isMatch[state2][posState2]) + + LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), + data[len], data2[len], p->ProbPrices); + + // state2 = kLiteralNextStates[state2]; + state2 = kState_LitAfterRep; + posState2 = (posState2 + 1) & p->pbMask; + + + price += GetPrice_Rep_0(p, state2, posState2); + + for (; len2 < limit && data[len2] == data2[len2]; len2++) + {} + + len2 -= len; + // if (len2 >= 3) + { + { + unsigned offset = cur + len + len2; + + if (last < offset) + last = offset; + // do + { + UInt32 price2; + COptimal *opt; + len2--; + // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); + + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->extra = (CExtra)(len + 1); + opt->dist = (UInt32)repIndex; + } + } + // while (len2 >= 3); + } + } + } + } + } + } + + + // ---------- MATCH ---------- + /* for (unsigned len = 2; len <= newLen; len++) */ + if (newLen > numAvail) + { + newLen = numAvail; + for (numPairs = 0; newLen > matches[numPairs]; numPairs += 2); + matches[numPairs] = (UInt32)newLen; + numPairs += 2; + } + + // startLen = 2; /* speed optimization */ + + if (newLen >= startLen) + { + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[state]); + UInt32 dist; + unsigned offs, posSlot, len; + + { + unsigned offset = cur + newLen; + if (last < offset) + last = offset; + } + + offs = 0; + while (startLen > matches[offs]) + offs += 2; + dist = matches[(size_t)offs + 1]; + + // if (dist >= kNumFullDistances) + GetPosSlot2(dist, posSlot); + + for (len = /*2*/ startLen; ; len++) + { + UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); + { + COptimal *opt; + unsigned lenNorm = len - 2; + lenNorm = GetLenToPosState2(lenNorm); + if (dist < kNumFullDistances) + price += p->distancesPrices[lenNorm][dist & (kNumFullDistances - 1)]; + else + price += p->posSlotPrices[lenNorm][posSlot] + p->alignPrices[dist & kAlignMask]; + + opt = &p->opt[cur + len]; + if (price < opt->price) + { + opt->price = price; + opt->len = (UInt32)len; + opt->dist = dist + LZMA_NUM_REPS; + opt->extra = 0; + } + } + + if (len == matches[offs]) + { + // if (p->_maxMode) { + // MATCH : LIT : REP_0 + + const Byte *data2 = data - dist - 1; + unsigned len2 = len + 1; + unsigned limit = len2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + + len2 += 2; + if (len2 <= limit) + if (data[len2 - 2] == data2[len2 - 2]) + if (data[len2 - 1] == data2[len2 - 1]) + { + for (; len2 < limit && data[len2] == data2[len2]; len2++) + {} + + len2 -= len; + + // if (len2 >= 3) + { + unsigned state2 = kMatchNextStates[state]; + unsigned posState2 = (position + len) & p->pbMask; + unsigned offset; + price += GET_PRICE_0(p->isMatch[state2][posState2]); + price += LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), + data[len], data2[len], p->ProbPrices); + + // state2 = kLiteralNextStates[state2]; + state2 = kState_LitAfterMatch; + + posState2 = (posState2 + 1) & p->pbMask; + price += GetPrice_Rep_0(p, state2, posState2); + + offset = cur + len + len2; + + if (last < offset) + last = offset; + // do + { + UInt32 price2; + COptimal *opt; + len2--; + // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->extra = (CExtra)(len + 1); + opt->dist = dist + LZMA_NUM_REPS; + } + } + // while (len2 >= 3); + } + + } + + offs += 2; + if (offs == numPairs) + break; + dist = matches[(size_t)offs + 1]; + // if (dist >= kNumFullDistances) + GetPosSlot2(dist, posSlot); + } + } + } + } + + do + p->opt[last].price = kInfinityPrice; + while (--last); + + return Backward(p, cur); +} + + + +#define ChangePair(smallDist, bigDist) (((bigDist) >> 7) > (smallDist)) + + + +static unsigned GetOptimumFast(CLzmaEnc *p) +{ + UInt32 numAvail, mainDist; + unsigned mainLen, numPairs, repIndex, repLen, i; + const Byte *data; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLen; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + p->backRes = MARK_LIT; + if (numAvail < 2) + return 1; + // if (mainLen < 2 && p->state == 0) return 1; // 18.06.notused + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + repLen = repIndex = 0; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len; + const Byte *data2 = data - p->reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + if (len >= p->numFastBytes) + { + p->backRes = (UInt32)i; + MOVE_POS(p, len - 1) + return len; + } + if (len > repLen) + { + repIndex = i; + repLen = len; + } + } + + if (mainLen >= p->numFastBytes) + { + p->backRes = p->matches[(size_t)numPairs - 1] + LZMA_NUM_REPS; + MOVE_POS(p, mainLen - 1) + return mainLen; + } + + mainDist = 0; /* for GCC */ + + if (mainLen >= 2) + { + mainDist = p->matches[(size_t)numPairs - 1]; + while (numPairs > 2) + { + UInt32 dist2; + if (mainLen != p->matches[(size_t)numPairs - 4] + 1) + break; + dist2 = p->matches[(size_t)numPairs - 3]; + if (!ChangePair(dist2, mainDist)) + break; + numPairs -= 2; + mainLen--; + mainDist = dist2; + } + if (mainLen == 2 && mainDist >= 0x80) + mainLen = 1; + } + + if (repLen >= 2) + if ( repLen + 1 >= mainLen + || (repLen + 2 >= mainLen && mainDist >= (1 << 9)) + || (repLen + 3 >= mainLen && mainDist >= (1 << 15))) + { + p->backRes = (UInt32)repIndex; + MOVE_POS(p, repLen - 1) + return repLen; + } + + if (mainLen < 2 || numAvail <= 2) + return 1; + + { + unsigned len1 = ReadMatchDistances(p, &p->numPairs); + p->longestMatchLen = len1; + + if (len1 >= 2) + { + UInt32 newDist = p->matches[(size_t)p->numPairs - 1]; + if ( (len1 >= mainLen && newDist < mainDist) + || (len1 == mainLen + 1 && !ChangePair(mainDist, newDist)) + || (len1 > mainLen + 1) + || (len1 + 1 >= mainLen && mainLen >= 3 && ChangePair(newDist, mainDist))) + return 1; + } + } + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len, limit; + const Byte *data2 = data - p->reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + limit = mainLen - 1; + for (len = 2;; len++) + { + if (len >= limit) + return 1; + if (data[len] != data2[len]) + break; + } + } + + p->backRes = mainDist + LZMA_NUM_REPS; + if (mainLen != 2) + { + MOVE_POS(p, mainLen - 2) + } + return mainLen; +} + + + + +static void WriteEndMarker(CLzmaEnc *p, unsigned posState) +{ + UInt32 range; + range = p->rc.range; + { + UInt32 ttt, newBound; + CLzmaProb *prob = &p->isMatch[p->state][posState]; + RC_BIT_PRE(&p->rc, prob) + RC_BIT_1(&p->rc, prob) + prob = &p->isRep[p->state]; + RC_BIT_PRE(&p->rc, prob) + RC_BIT_0(&p->rc, prob) + } + p->state = kMatchNextStates[p->state]; + + p->rc.range = range; + LenEnc_Encode(&p->lenProbs, &p->rc, 0, posState); + range = p->rc.range; + + { + // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[0], (1 << kNumPosSlotBits) - 1); + CLzmaProb *probs = p->posSlotEncoder[0]; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + RC_BIT_PRE(p, probs + m) + RC_BIT_1(&p->rc, probs + m); + m = (m << 1) + 1; + } + while (m < (1 << kNumPosSlotBits)); + } + { + // RangeEnc_EncodeDirectBits(&p->rc, ((UInt32)1 << (30 - kNumAlignBits)) - 1, 30 - kNumAlignBits); UInt32 range = p->range; + unsigned numBits = 30 - kNumAlignBits; + do + { + range >>= 1; + p->rc.low += range; + RC_NORM(&p->rc) + } + while (--numBits); + } + + { + // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, kAlignMask); + CLzmaProb *probs = p->posAlignEncoder; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + RC_BIT_PRE(p, probs + m) + RC_BIT_1(&p->rc, probs + m); + m = (m << 1) + 1; + } + while (m < kAlignTableSize); + } + p->rc.range = range; +} + + +static SRes CheckErrors(CLzmaEnc *p) +{ + if (p->result != SZ_OK) + return p->result; + if (p->rc.res != SZ_OK) + p->result = SZ_ERROR_WRITE; + if (p->matchFinderBase.result != SZ_OK) + p->result = SZ_ERROR_READ; + if (p->result != SZ_OK) + p->finished = True; + return p->result; +} + + +MY_NO_INLINE static SRes Flush(CLzmaEnc *p, UInt32 nowPos) +{ + /* ReleaseMFStream(); */ + p->finished = True; + if (p->writeEndMark) + WriteEndMarker(p, nowPos & p->pbMask); + RangeEnc_FlushData(&p->rc); + RangeEnc_FlushStream(&p->rc); + return CheckErrors(p); +} + + +MY_NO_INLINE static void FillAlignPrices(CLzmaEnc *p) +{ + unsigned i; + const CProbPrice *ProbPrices = p->ProbPrices; + const CLzmaProb *probs = p->posAlignEncoder; + // p->alignPriceCount = 0; + for (i = 0; i < kAlignTableSize / 2; i++) + { + UInt32 price = 0; + unsigned sym = i; + unsigned m = 1; + unsigned bit; + UInt32 prob; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + prob = probs[m]; + p->alignPrices[i ] = price + GET_PRICEa_0(prob); + p->alignPrices[i + 8] = price + GET_PRICEa_1(prob); + // p->alignPrices[i] = RcTree_ReverseGetPrice(p->posAlignEncoder, kNumAlignBits, i, p->ProbPrices); + } +} + + +MY_NO_INLINE static void FillDistancesPrices(CLzmaEnc *p) +{ + // int y; for (y = 0; y < 100; y++) { + + UInt32 tempPrices[kNumFullDistances]; + unsigned i, lps; + + const CProbPrice *ProbPrices = p->ProbPrices; + p->matchPriceCount = 0; + + for (i = kStartPosModelIndex / 2; i < kNumFullDistances / 2; i++) + { + unsigned posSlot = GetPosSlot1(i); + unsigned footerBits = (posSlot >> 1) - 1; + unsigned base = ((2 | (posSlot & 1)) << footerBits); + const CLzmaProb *probs = p->posEncoders + (size_t)base * 2; + // tempPrices[i] = RcTree_ReverseGetPrice(p->posEncoders + base, footerBits, i - base, p->ProbPrices); + UInt32 price = 0; + unsigned m = 1; + unsigned sym = i; + unsigned offset = (unsigned)1 << footerBits; + base += i; + + if (footerBits) + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[m], bit); + m = (m << 1) + bit; + } + while (--footerBits); + + { + unsigned prob = probs[m]; + tempPrices[base ] = price + GET_PRICEa_0(prob); + tempPrices[base + offset] = price + GET_PRICEa_1(prob); + } + } + + for (lps = 0; lps < kNumLenToPosStates; lps++) + { + unsigned slot; + unsigned distTableSize2 = (p->distTableSize + 1) >> 1; + UInt32 *posSlotPrices = p->posSlotPrices[lps]; + const CLzmaProb *probs = p->posSlotEncoder[lps]; + + for (slot = 0; slot < distTableSize2; slot++) + { + // posSlotPrices[slot] = RcTree_GetPrice(encoder, kNumPosSlotBits, slot, p->ProbPrices); + UInt32 price; + unsigned bit; + unsigned sym = slot + (1 << (kNumPosSlotBits - 1)); + unsigned prob; + bit = sym & 1; sym >>= 1; price = GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + prob = probs[(size_t)slot + (1 << (kNumPosSlotBits - 1))]; + posSlotPrices[(size_t)slot * 2 ] = price + GET_PRICEa_0(prob); + posSlotPrices[(size_t)slot * 2 + 1] = price + GET_PRICEa_1(prob); + } + + { + UInt32 delta = ((UInt32)((kEndPosModelIndex / 2 - 1) - kNumAlignBits) << kNumBitPriceShiftBits); + for (slot = kEndPosModelIndex / 2; slot < distTableSize2; slot++) + { + posSlotPrices[(size_t)slot * 2 ] += delta; + posSlotPrices[(size_t)slot * 2 + 1] += delta; + delta += ((UInt32)1 << kNumBitPriceShiftBits); + } + } + + { + UInt32 *dp = p->distancesPrices[lps]; + + dp[0] = posSlotPrices[0]; + dp[1] = posSlotPrices[1]; + dp[2] = posSlotPrices[2]; + dp[3] = posSlotPrices[3]; + + for (i = 4; i < kNumFullDistances; i += 2) + { + UInt32 slotPrice = posSlotPrices[GetPosSlot1(i)]; + dp[i ] = slotPrice + tempPrices[i]; + dp[i + 1] = slotPrice + tempPrices[i + 1]; + } + } + } + // } +} + + + +void LzmaEnc_Construct(CLzmaEnc *p) +{ + RangeEnc_Construct(&p->rc); + MatchFinder_Construct(&p->matchFinderBase); + + #ifndef _7ZIP_ST + MatchFinderMt_Construct(&p->matchFinderMt); + p->matchFinderMt.MatchFinder = &p->matchFinderBase; + #endif + + { + CLzmaEncProps props; + LzmaEncProps_Init(&props); + LzmaEnc_SetProps(p, &props); + } + + #ifndef LZMA_LOG_BSR + LzmaEnc_FastPosInit(p->g_FastPos); + #endif + + LzmaEnc_InitPriceTables(p->ProbPrices); + p->litProbs = NULL; + p->saveState.litProbs = NULL; + +} + +CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc) +{ + void *p; + p = ISzAlloc_Alloc(alloc, sizeof(CLzmaEnc)); + if (p) + LzmaEnc_Construct((CLzmaEnc *)p); + return p; +} + +void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->litProbs); + ISzAlloc_Free(alloc, p->saveState.litProbs); + p->litProbs = NULL; + p->saveState.litProbs = NULL; +} + +void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + #ifndef _7ZIP_ST + MatchFinderMt_Destruct(&p->matchFinderMt, allocBig); + #endif + + MatchFinder_Free(&p->matchFinderBase, allocBig); + LzmaEnc_FreeLits(p, alloc); + RangeEnc_Free(&p->rc, alloc); +} + +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + LzmaEnc_Destruct((CLzmaEnc *)p, alloc, allocBig); + ISzAlloc_Free(alloc, p); +} + + +static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpackSize) +{ + UInt32 nowPos32, startPos32; + if (p->needInit) + { + p->matchFinder.Init(p->matchFinderObj); + p->needInit = 0; + } + + if (p->finished) + return p->result; + RINOK(CheckErrors(p)); + + nowPos32 = (UInt32)p->nowPos64; + startPos32 = nowPos32; + + if (p->nowPos64 == 0) + { + unsigned numPairs; + Byte curByte; + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + return Flush(p, nowPos32); + ReadMatchDistances(p, &numPairs); + RangeEnc_EncodeBit_0(&p->rc, &p->isMatch[kState_Start][0]); + // p->state = kLiteralNextStates[p->state]; + curByte = *(p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset); + LitEnc_Encode(&p->rc, p->litProbs, curByte); + p->additionalOffset--; + nowPos32++; + } + + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) != 0) + + for (;;) + { + UInt32 dist; + unsigned len, posState; + UInt32 range, ttt, newBound; + CLzmaProb *probs; + + if (p->fastMode) + len = GetOptimumFast(p); + else + { + unsigned oci = p->optCur; + if (p->optEnd == oci) + len = GetOptimum(p, nowPos32); + else + { + const COptimal *opt = &p->opt[oci]; + len = opt->len; + p->backRes = opt->dist; + p->optCur = oci + 1; + } + } + + posState = (unsigned)nowPos32 & p->pbMask; + range = p->rc.range; + probs = &p->isMatch[p->state][posState]; + + RC_BIT_PRE(&p->rc, probs) + + dist = p->backRes; + + #ifdef SHOW_STAT2 + printf("\n pos = %6X, len = %3u pos = %6u", nowPos32, len, dist); + #endif + + if (dist == MARK_LIT) + { + Byte curByte; + const Byte *data; + unsigned state; + + RC_BIT_0(&p->rc, probs); + p->rc.range = range; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; + probs = LIT_PROBS(nowPos32, *(data - 1)); + curByte = *data; + state = p->state; + p->state = kLiteralNextStates[state]; + if (IsLitState(state)) + LitEnc_Encode(&p->rc, probs, curByte); + else + LitEnc_EncodeMatched(&p->rc, probs, curByte, *(data - p->reps[0])); + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRep[p->state]; + RC_BIT_PRE(&p->rc, probs) + + if (dist < LZMA_NUM_REPS) + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG0[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 0) + { + RC_BIT_0(&p->rc, probs); + probs = &p->isRep0Long[p->state][posState]; + RC_BIT_PRE(&p->rc, probs) + if (len != 1) + { + RC_BIT_1_BASE(&p->rc, probs); + } + else + { + RC_BIT_0_BASE(&p->rc, probs); + p->state = kShortRepNextStates[p->state]; + } + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG1[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 1) + { + RC_BIT_0_BASE(&p->rc, probs); + dist = p->reps[1]; + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG2[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 2) + { + RC_BIT_0_BASE(&p->rc, probs); + dist = p->reps[2]; + } + else + { + RC_BIT_1_BASE(&p->rc, probs); + dist = p->reps[3]; + p->reps[3] = p->reps[2]; + } + p->reps[2] = p->reps[1]; + } + p->reps[1] = p->reps[0]; + p->reps[0] = dist; + } + + RC_NORM(&p->rc) + + p->rc.range = range; + + if (len != 1) + { + LenEnc_Encode(&p->repLenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); + --p->repLenEncCounter; + p->state = kRepNextStates[p->state]; + } + } + else + { + unsigned posSlot; + RC_BIT_0(&p->rc, probs); + p->rc.range = range; + p->state = kMatchNextStates[p->state]; + + LenEnc_Encode(&p->lenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); + // --p->lenEnc.counter; + + dist -= LZMA_NUM_REPS; + p->reps[3] = p->reps[2]; + p->reps[2] = p->reps[1]; + p->reps[1] = p->reps[0]; + p->reps[0] = dist + 1; + + p->matchPriceCount++; + GetPosSlot(dist, posSlot); + // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[GetLenToPosState(len)], posSlot); + { + UInt32 sym = (UInt32)posSlot + (1 << kNumPosSlotBits); + range = p->rc.range; + probs = p->posSlotEncoder[GetLenToPosState(len)]; + do + { + CLzmaProb *prob = probs + (sym >> kNumPosSlotBits); + UInt32 bit = (sym >> (kNumPosSlotBits - 1)) & 1; + sym <<= 1; + RC_BIT(&p->rc, prob, bit); + } + while (sym < (1 << kNumPosSlotBits * 2)); + p->rc.range = range; + } + + if (dist >= kStartPosModelIndex) + { + unsigned footerBits = ((posSlot >> 1) - 1); + + if (dist < kNumFullDistances) + { + unsigned base = ((2 | (posSlot & 1)) << footerBits); + RcTree_ReverseEncode(&p->rc, p->posEncoders + base, footerBits, (unsigned)(dist /* - base */)); + } + else + { + UInt32 pos2 = (dist | 0xF) << (32 - footerBits); + range = p->rc.range; + // RangeEnc_EncodeDirectBits(&p->rc, posReduced >> kNumAlignBits, footerBits - kNumAlignBits); + /* + do + { + range >>= 1; + p->rc.low += range & (0 - ((dist >> --footerBits) & 1)); + RC_NORM(&p->rc) + } + while (footerBits > kNumAlignBits); + */ + do + { + range >>= 1; + p->rc.low += range & (0 - (pos2 >> 31)); + pos2 += pos2; + RC_NORM(&p->rc) + } + while (pos2 != 0xF0000000); + + + // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, posReduced & kAlignMask); + + { + unsigned m = 1; + unsigned bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); + p->rc.range = range; + // p->alignPriceCount++; + } + } + } + } + } + + nowPos32 += (UInt32)len; + p->additionalOffset -= len; + + if (p->additionalOffset == 0) + { + UInt32 processed; + + if (!p->fastMode) + { + /* + if (p->alignPriceCount >= 16) // kAlignTableSize + FillAlignPrices(p); + if (p->matchPriceCount >= 128) + FillDistancesPrices(p); + if (p->lenEnc.counter <= 0) + LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices); + */ + if (p->matchPriceCount >= 64) + { + FillAlignPrices(p); + // { int y; for (y = 0; y < 100; y++) { + FillDistancesPrices(p); + // }} + LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices); + } + if (p->repLenEncCounter <= 0) + { + p->repLenEncCounter = REP_LEN_COUNT; + LenPriceEnc_UpdateTables(&p->repLenEnc, 1 << p->pb, &p->repLenProbs, p->ProbPrices); + } + } + + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + break; + processed = nowPos32 - startPos32; + + if (maxPackSize) + { + if (processed + kNumOpts + 300 >= maxUnpackSize + || RangeEnc_GetProcessed_sizet(&p->rc) + kPackReserve >= maxPackSize) + break; + } + else if (processed >= (1 << 17)) + { + p->nowPos64 += nowPos32 - startPos32; + return CheckErrors(p); + } + } + } + + p->nowPos64 += nowPos32 - startPos32; + return Flush(p, nowPos32); +} + + + +#define kBigHashDicLimit ((UInt32)1 << 24) + +static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + UInt32 beforeSize = kNumOpts; + if (!RangeEnc_Alloc(&p->rc, alloc)) + return SZ_ERROR_MEM; + + #ifndef _7ZIP_ST + p->mtMode = (p->multiThread && !p->fastMode && (p->matchFinderBase.btMode != 0)); + #endif + + { + unsigned lclp = p->lc + p->lp; + if (!p->litProbs || !p->saveState.litProbs || p->lclp != lclp) + { + LzmaEnc_FreeLits(p, alloc); + p->litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + p->saveState.litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + if (!p->litProbs || !p->saveState.litProbs) + { + LzmaEnc_FreeLits(p, alloc); + return SZ_ERROR_MEM; + } + p->lclp = lclp; + } + } + + p->matchFinderBase.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); + + if (beforeSize + p->dictSize < keepWindowSize) + beforeSize = keepWindowSize - p->dictSize; + + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_Create(&p->matchFinderMt, p->dictSize, beforeSize, p->numFastBytes, + LZMA_MATCH_LEN_MAX + + 1 /* 18.04 */ + , allocBig)); + p->matchFinderObj = &p->matchFinderMt; + p->matchFinderBase.bigHash = (Byte)( + (p->dictSize > kBigHashDicLimit && p->matchFinderBase.hashMask >= 0xFFFFFF) ? 1 : 0); + MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder); + } + else + #endif + { + if (!MatchFinder_Create(&p->matchFinderBase, p->dictSize, beforeSize, p->numFastBytes, LZMA_MATCH_LEN_MAX, allocBig)) + return SZ_ERROR_MEM; + p->matchFinderObj = &p->matchFinderBase; + MatchFinder_CreateVTable(&p->matchFinderBase, &p->matchFinder); + } + + return SZ_OK; +} + +void LzmaEnc_Init(CLzmaEnc *p) +{ + unsigned i; + p->state = 0; + p->reps[0] = + p->reps[1] = + p->reps[2] = + p->reps[3] = 1; + + RangeEnc_Init(&p->rc); + + for (i = 0; i < (1 << kNumAlignBits); i++) + p->posAlignEncoder[i] = kProbInitValue; + + for (i = 0; i < kNumStates; i++) + { + unsigned j; + for (j = 0; j < LZMA_NUM_PB_STATES_MAX; j++) + { + p->isMatch[i][j] = kProbInitValue; + p->isRep0Long[i][j] = kProbInitValue; + } + p->isRep[i] = kProbInitValue; + p->isRepG0[i] = kProbInitValue; + p->isRepG1[i] = kProbInitValue; + p->isRepG2[i] = kProbInitValue; + } + + { + for (i = 0; i < kNumLenToPosStates; i++) + { + CLzmaProb *probs = p->posSlotEncoder[i]; + unsigned j; + for (j = 0; j < (1 << kNumPosSlotBits); j++) + probs[j] = kProbInitValue; + } + } + { + for (i = 0; i < kNumFullDistances; i++) + p->posEncoders[i] = kProbInitValue; + } + + { + UInt32 num = (UInt32)0x300 << (p->lp + p->lc); + UInt32 k; + CLzmaProb *probs = p->litProbs; + for (k = 0; k < num; k++) + probs[k] = kProbInitValue; + } + + + LenEnc_Init(&p->lenProbs); + LenEnc_Init(&p->repLenProbs); + + p->optEnd = 0; + p->optCur = 0; + + { + for (i = 0; i < kNumOpts; i++) + p->opt[i].price = kInfinityPrice; + } + + p->additionalOffset = 0; + + p->pbMask = (1 << p->pb) - 1; + p->lpMask = ((UInt32)0x100 << p->lp) - ((unsigned)0x100 >> p->lc); +} + + +void LzmaEnc_InitPrices(CLzmaEnc *p) +{ + if (!p->fastMode) + { + FillDistancesPrices(p); + FillAlignPrices(p); + } + + p->lenEnc.tableSize = + p->repLenEnc.tableSize = + p->numFastBytes + 1 - LZMA_MATCH_LEN_MIN; + + p->repLenEncCounter = REP_LEN_COUNT; + + LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices); + LenPriceEnc_UpdateTables(&p->repLenEnc, 1 << p->pb, &p->repLenProbs, p->ProbPrices); +} + +static SRes LzmaEnc_AllocAndInit(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + unsigned i; + for (i = kEndPosModelIndex / 2; i < kDicLogSizeMax; i++) + if (p->dictSize <= ((UInt32)1 << i)) + break; + p->distTableSize = i * 2; + + p->finished = False; + p->result = SZ_OK; + RINOK(LzmaEnc_Alloc(p, keepWindowSize, alloc, allocBig)); + LzmaEnc_Init(p); + LzmaEnc_InitPrices(p); + p->nowPos64 = 0; + return SZ_OK; +} + +static SRes LzmaEnc_Prepare(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + p->matchFinderBase.stream = inStream; + p->needInit = 1; + p->rc.outStream = outStream; + return LzmaEnc_AllocAndInit(p, 0, alloc, allocBig); +} + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, + ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + p->matchFinderBase.stream = inStream; + p->needInit = 1; + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen) +{ + p->matchFinderBase.directInput = 1; + p->matchFinderBase.bufferBase = (Byte *)src; + p->matchFinderBase.directInputRem = srcLen; +} + +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + LzmaEnc_SetInputBuf(p, src, srcLen); + p->needInit = 1; + + LzmaEnc_SetDataSize(pp, srcLen); + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +void LzmaEnc_Finish(CLzmaEncHandle pp) +{ + #ifndef _7ZIP_ST + CLzmaEnc *p = (CLzmaEnc *)pp; + if (p->mtMode) + MatchFinderMt_ReleaseStream(&p->matchFinderMt); + #else + UNUSED_VAR(pp); + #endif +} + + +typedef struct +{ + ISeqOutStream vt; + Byte *data; + SizeT rem; + BoolInt overflow; +} CLzmaEnc_SeqOutStreamBuf; + +static size_t SeqOutStreamBuf_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CLzmaEnc_SeqOutStreamBuf *p = CONTAINER_FROM_VTBL(pp, CLzmaEnc_SeqOutStreamBuf, vt); + if (p->rem < size) + { + size = p->rem; + p->overflow = True; + } + memcpy(p->data, data, size); + p->rem -= size; + p->data += size; + return size; +} + + +UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + return p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); +} + + +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + return p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; +} + + +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + UInt64 nowPos64; + SRes res; + CLzmaEnc_SeqOutStreamBuf outStream; + + outStream.vt.Write = SeqOutStreamBuf_Write; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = False; + p->finished = False; + p->result = SZ_OK; + + if (reInit) + LzmaEnc_Init(p); + LzmaEnc_InitPrices(p); + + nowPos64 = p->nowPos64; + RangeEnc_Init(&p->rc); + p->rc.outStream = &outStream.vt; + + if (desiredPackSize == 0) + return SZ_ERROR_OUTPUT_EOF; + + res = LzmaEnc_CodeOneBlock(p, desiredPackSize, *unpackSize); + + *unpackSize = (UInt32)(p->nowPos64 - nowPos64); + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + + return res; +} + + +static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress) +{ + SRes res = SZ_OK; + + #ifndef _7ZIP_ST + Byte allocaDummy[0x300]; + allocaDummy[0] = 0; + allocaDummy[1] = allocaDummy[0]; + #endif + + for (;;) + { + res = LzmaEnc_CodeOneBlock(p, 0, 0); + if (res != SZ_OK || p->finished) + break; + if (progress) + { + res = ICompressProgress_Progress(progress, p->nowPos64, RangeEnc_GetProcessed(&p->rc)); + if (res != SZ_OK) + { + res = SZ_ERROR_PROGRESS; + break; + } + } + } + + LzmaEnc_Finish(p); + + /* + if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&p->matchFinderBase)) + res = SZ_ERROR_FAIL; + } + */ + + return res; +} + + +SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + RINOK(LzmaEnc_Prepare(pp, outStream, inStream, alloc, allocBig)); + return LzmaEnc_Encode2((CLzmaEnc *)pp, progress); +} + + +SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + unsigned i; + UInt32 dictSize = p->dictSize; + if (*size < LZMA_PROPS_SIZE) + return SZ_ERROR_PARAM; + *size = LZMA_PROPS_SIZE; + props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); + + if (dictSize >= ((UInt32)1 << 22)) + { + UInt32 kDictMask = ((UInt32)1 << 20) - 1; + if (dictSize < (UInt32)0xFFFFFFFF - kDictMask) + dictSize = (dictSize + kDictMask) & ~kDictMask; + } + else for (i = 11; i <= 30; i++) + { + if (dictSize <= ((UInt32)2 << i)) { dictSize = (2 << i); break; } + if (dictSize <= ((UInt32)3 << i)) { dictSize = (3 << i); break; } + } + + for (i = 0; i < 4; i++) + props[1 + i] = (Byte)(dictSize >> (8 * i)); + return SZ_OK; +} + + +unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle pp) +{ + return ((CLzmaEnc *)pp)->writeEndMark; +} + + +SRes LzmaEnc_MemEncode(CLzmaEncHandle pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + SRes res; + CLzmaEnc *p = (CLzmaEnc *)pp; + + CLzmaEnc_SeqOutStreamBuf outStream; + + outStream.vt.Write = SeqOutStreamBuf_Write; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = writeEndMark; + p->rc.outStream = &outStream.vt; + + res = LzmaEnc_MemPrepare(pp, src, srcLen, 0, alloc, allocBig); + + if (res == SZ_OK) + { + res = LzmaEnc_Encode2(p, progress); + if (res == SZ_OK && p->nowPos64 != srcLen) + res = SZ_ERROR_FAIL; + } + + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + return res; +} + + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)LzmaEnc_Create(alloc); + SRes res; + if (!p) + return SZ_ERROR_MEM; + + res = LzmaEnc_SetProps(p, props); + if (res == SZ_OK) + { + res = LzmaEnc_WriteProperties(p, propsEncoded, propsSize); + if (res == SZ_OK) + res = LzmaEnc_MemEncode(p, dest, destLen, src, srcLen, + writeEndMark, progress, alloc, allocBig); + } + + LzmaEnc_Destroy(p, alloc, allocBig); + return res; +} diff --git a/dep/lzma/src/LzmaLib.c b/dep/lzma/src/LzmaLib.c new file mode 100644 index 000000000..706e9e58c --- /dev/null +++ b/dep/lzma/src/LzmaLib.c @@ -0,0 +1,40 @@ +/* LzmaLib.c -- LZMA library wrapper +2015-06-13 : Igor Pavlov : Public domain */ + +#include "Alloc.h" +#include "LzmaDec.h" +#include "LzmaEnc.h" +#include "LzmaLib.h" + +MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, + unsigned char *outProps, size_t *outPropsSize, + int level, /* 0 <= level <= 9, default = 5 */ + unsigned dictSize, /* use (1 << N) or (3 << N). 4 KB < dictSize <= 128 MB */ + int lc, /* 0 <= lc <= 8, default = 3 */ + int lp, /* 0 <= lp <= 4, default = 0 */ + int pb, /* 0 <= pb <= 4, default = 2 */ + int fb, /* 5 <= fb <= 273, default = 32 */ + int numThreads /* 1 or 2, default = 2 */ +) +{ + CLzmaEncProps props; + LzmaEncProps_Init(&props); + props.level = level; + props.dictSize = dictSize; + props.lc = lc; + props.lp = lp; + props.pb = pb; + props.fb = fb; + props.numThreads = numThreads; + + return LzmaEncode(dest, destLen, src, srcLen, &props, outProps, outPropsSize, 0, + NULL, &g_Alloc, &g_Alloc); +} + + +MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t *srcLen, + const unsigned char *props, size_t propsSize) +{ + ELzmaStatus status; + return LzmaDecode(dest, destLen, src, srcLen, props, (unsigned)propsSize, LZMA_FINISH_ANY, &status, &g_Alloc); +} diff --git a/dep/lzma/src/Sort.c b/dep/lzma/src/Sort.c new file mode 100644 index 000000000..e1097e380 --- /dev/null +++ b/dep/lzma/src/Sort.c @@ -0,0 +1,141 @@ +/* Sort.c -- Sort functions +2014-04-05 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Sort.h" + +#define HeapSortDown(p, k, size, temp) \ + { for (;;) { \ + size_t s = (k << 1); \ + if (s > size) break; \ + if (s < size && p[s + 1] > p[s]) s++; \ + if (temp >= p[s]) break; \ + p[k] = p[s]; k = s; \ + } p[k] = temp; } + +void HeapSort(UInt32 *p, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt32 temp = p[i]; + size_t k = i; + HeapSortDown(p, k, size, temp) + } + while (--i != 0); + } + /* + do + { + size_t k = 1; + UInt32 temp = p[size]; + p[size--] = p[1]; + HeapSortDown(p, k, size, temp) + } + while (size > 1); + */ + while (size > 3) + { + UInt32 temp = p[size]; + size_t k = (p[3] > p[2]) ? 3 : 2; + p[size--] = p[1]; + p[1] = p[k]; + HeapSortDown(p, k, size, temp) + } + { + UInt32 temp = p[size]; + p[size] = p[1]; + if (size > 2 && p[2] < temp) + { + p[1] = p[2]; + p[2] = temp; + } + else + p[1] = temp; + } +} + +void HeapSort64(UInt64 *p, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt64 temp = p[i]; + size_t k = i; + HeapSortDown(p, k, size, temp) + } + while (--i != 0); + } + /* + do + { + size_t k = 1; + UInt64 temp = p[size]; + p[size--] = p[1]; + HeapSortDown(p, k, size, temp) + } + while (size > 1); + */ + while (size > 3) + { + UInt64 temp = p[size]; + size_t k = (p[3] > p[2]) ? 3 : 2; + p[size--] = p[1]; + p[1] = p[k]; + HeapSortDown(p, k, size, temp) + } + { + UInt64 temp = p[size]; + p[size] = p[1]; + if (size > 2 && p[2] < temp) + { + p[1] = p[2]; + p[2] = temp; + } + else + p[1] = temp; + } +} + +/* +#define HeapSortRefDown(p, vals, n, size, temp) \ + { size_t k = n; UInt32 val = vals[temp]; for (;;) { \ + size_t s = (k << 1); \ + if (s > size) break; \ + if (s < size && vals[p[s + 1]] > vals[p[s]]) s++; \ + if (val >= vals[p[s]]) break; \ + p[k] = p[s]; k = s; \ + } p[k] = temp; } + +void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt32 temp = p[i]; + HeapSortRefDown(p, vals, i, size, temp); + } + while (--i != 0); + } + do + { + UInt32 temp = p[size]; + p[size--] = p[1]; + HeapSortRefDown(p, vals, 1, size, temp); + } + while (size > 1); +} +*/ diff --git a/duckstation.sln b/duckstation.sln index 061dff69e..47a72edbe 100644 --- a/duckstation.sln +++ b/duckstation.sln @@ -33,6 +33,8 @@ Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "minizip", "dep\minizip\mini EndProject Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "zlib", "dep\zlib\zlib.vcxproj", "{7FF9FDB9-D504-47DB-A16A-B08071999620}" EndProject +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "lzma", "dep\lzma\lzma.vcxproj", "{DD944834-7899-4C1C-A4C1-064B5009D239}" +EndProject Global GlobalSection(SolutionConfigurationPlatforms) = preSolution Debug|x64 = Debug|x64 @@ -269,6 +271,22 @@ Global {7FF9FDB9-D504-47DB-A16A-B08071999620}.ReleaseLTCG|x64.Build.0 = ReleaseLTCG|x64 {7FF9FDB9-D504-47DB-A16A-B08071999620}.ReleaseLTCG|x86.ActiveCfg = ReleaseLTCG|Win32 {7FF9FDB9-D504-47DB-A16A-B08071999620}.ReleaseLTCG|x86.Build.0 = ReleaseLTCG|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Debug|x64.ActiveCfg = Debug|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Debug|x64.Build.0 = Debug|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Debug|x86.ActiveCfg = Debug|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Debug|x86.Build.0 = Debug|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.DebugFast|x64.ActiveCfg = DebugFast|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.DebugFast|x64.Build.0 = DebugFast|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.DebugFast|x86.ActiveCfg = DebugFast|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.DebugFast|x86.Build.0 = DebugFast|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Release|x64.ActiveCfg = Release|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Release|x64.Build.0 = Release|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Release|x86.ActiveCfg = Release|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.Release|x86.Build.0 = Release|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.ReleaseLTCG|x64.ActiveCfg = ReleaseLTCG|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.ReleaseLTCG|x64.Build.0 = ReleaseLTCG|x64 + {DD944834-7899-4C1C-A4C1-064B5009D239}.ReleaseLTCG|x86.ActiveCfg = ReleaseLTCG|Win32 + {DD944834-7899-4C1C-A4C1-064B5009D239}.ReleaseLTCG|x86.Build.0 = ReleaseLTCG|Win32 EndGlobalSection GlobalSection(SolutionProperties) = preSolution HideSolutionNode = FALSE @@ -284,6 +302,7 @@ Global {72F9423C-91EE-4487-AAC6-555ED6F61AA1} = {BA490C0E-497D-4634-A21E-E65012006385} {8BDA439C-6358-45FB-9994-2FF083BABE06} = {BA490C0E-497D-4634-A21E-E65012006385} {7FF9FDB9-D504-47DB-A16A-B08071999620} = {BA490C0E-497D-4634-A21E-E65012006385} + {DD944834-7899-4C1C-A4C1-064B5009D239} = {BA490C0E-497D-4634-A21E-E65012006385} EndGlobalSection GlobalSection(ExtensibilityGlobals) = postSolution SolutionGuid = {26E40B32-7C1D-48D0-95F4-1A500E054028}