bitstream.h

Go to the documentation of this file.

/* ******************************************************************
 * bitstream
 * Part of FSE library
 * Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
 *
 * You can contact the author at :
 * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
 *
 * This source code is licensed under both the BSD-style license (found in the
 * LICENSE file in the root directory of this source tree) and the GPLv2 (found
 * in the COPYING file in the root directory of this source tree).
 * You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#ifndef BITSTREAM_H_MODULE
#define BITSTREAM_H_MODULE
 
#if defined (__cplusplus)
extern "C" {
#endif
 
/*
*  This API consists of small unitary functions, which must be inlined for best performance.
*  Since link-time-optimization is not available for all compilers,
*  these functions are defined into a .h to be included.
*/
 
/*-****************************************
*  Dependencies
******************************************/
#include "mem.h"            /* unaligned access routines */
#include "compiler.h"       /* UNLIKELY() */
#include "debug.h"          /* assert(), DEBUGLOG(), RAWLOG() */
#include "error_private.h"  /* error codes and messages */
 
 
/*=========================================
*  Target specific
=========================================*/
#if defined(__BMI__) && defined(__GNUC__)
#  include <immintrin.h>   /* support for bextr (experimental) */
#elif defined(__ICCARM__)
#  include <intrinsics.h>
#endif
 
#define STREAM_ACCUMULATOR_MIN_32  25
#define STREAM_ACCUMULATOR_MIN_64  57
#define STREAM_ACCUMULATOR_MIN    ((U32)(MEM_32bits() ? STREAM_ACCUMULATOR_MIN_32 : STREAM_ACCUMULATOR_MIN_64))
 
 
/*-******************************************
*  bitStream encoding API (write forward)
********************************************/
/* bitStream can mix input from multiple sources.
 * A critical property of these streams is that they encode and decode in **reverse** direction.
 * So the first bit sequence you add will be the last to be read, like a LIFO stack.
 */
typedef struct {
    size_t bitContainer;
    unsigned bitPos;
    char*  startPtr;
    char*  ptr;
    char*  endPtr;
} BIT_CStream_t;
 
MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, void* dstBuffer, size_t dstCapacity);
MEM_STATIC void   BIT_addBits(BIT_CStream_t* bitC, size_t value, unsigned nbBits);
MEM_STATIC void   BIT_flushBits(BIT_CStream_t* bitC);
MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC);
 
/* Start with initCStream, providing the size of buffer to write into.
*  bitStream will never write outside of this buffer.
*  `dstCapacity` must be >= sizeof(bitD->bitContainer), otherwise @return will be an error code.
*
*  bits are first added to a local register.
*  Local register is size_t, hence 64-bits on 64-bits systems, or 32-bits on 32-bits systems.
*  Writing data into memory is an explicit operation, performed by the flushBits function.
*  Hence keep track how many bits are potentially stored into local register to avoid register overflow.
*  After a flushBits, a maximum of 7 bits might still be stored into local register.
*
*  Avoid storing elements of more than 24 bits if you want compatibility with 32-bits bitstream readers.
*
*  Last operation is to close the bitStream.
*  The function returns the final size of CStream in bytes.
*  If data couldn't fit into `dstBuffer`, it will return a 0 ( == not storable)
*/
 
 
/*-********************************************
*  bitStream decoding API (read backward)
**********************************************/
typedef struct {
    size_t   bitContainer;
    unsigned bitsConsumed;
    const char* ptr;
    const char* start;
    const char* limitPtr;
} BIT_DStream_t;
 
typedef enum { BIT_DStream_unfinished = 0,
               BIT_DStream_endOfBuffer = 1,
               BIT_DStream_completed = 2,
               BIT_DStream_overflow = 3 } BIT_DStream_status;  /* result of BIT_reloadDStream() */
               /* 1,2,4,8 would be better for bitmap combinations, but slows down performance a bit ... :( */
 
MEM_STATIC size_t   BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize);
MEM_STATIC size_t   BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits);
MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD);
MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* bitD);
 
 
/* Start by invoking BIT_initDStream().
*  A chunk of the bitStream is then stored into a local register.
*  Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t).
*  You can then retrieve bitFields stored into the local register, **in reverse order**.
*  Local register is explicitly reloaded from memory by the BIT_reloadDStream() method.
*  A reload guarantee a minimum of ((8*sizeof(bitD->bitContainer))-7) bits when its result is BIT_DStream_unfinished.
*  Otherwise, it can be less than that, so proceed accordingly.
*  Checking if DStream has reached its end can be performed with BIT_endOfDStream().
*/
 
 
/*-****************************************
*  unsafe API
******************************************/
MEM_STATIC void BIT_addBitsFast(BIT_CStream_t* bitC, size_t value, unsigned nbBits);
/* faster, but works only if value is "clean", meaning all high bits above nbBits are 0 */
 
MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC);
/* unsafe version; does not check buffer overflow */
 
MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits);
/* faster, but works only if nbBits >= 1 */
 
 
 
/*-**************************************************************
*  Internal functions
****************************************************************/
MEM_STATIC unsigned BIT_highbit32 (U32 val)
{
    assert(val != 0);
    {
#   if defined(_MSC_VER)   /* Visual */
        unsigned long r=0;
        return _BitScanReverse ( &r, val ) ? (unsigned)r : 0;
#   elif defined(__GNUC__) && (__GNUC__ >= 3)   /* Use GCC Intrinsic */
        return __builtin_clz (val) ^ 31;
#   elif defined(__ICCARM__)    /* IAR Intrinsic */
        return 31 - __CLZ(val);
#   else   /* Software version */
        static const unsigned DeBruijnClz[32] = { 0,  9,  1, 10, 13, 21,  2, 29,
                                                 11, 14, 16, 18, 22, 25,  3, 30,
                                                  8, 12, 20, 28, 15, 17, 24,  7,
                                                 19, 27, 23,  6, 26,  5,  4, 31 };
        U32 v = val;
        v |= v >> 1;
        v |= v >> 2;
        v |= v >> 4;
        v |= v >> 8;
        v |= v >> 16;
        return DeBruijnClz[ (U32) (v * 0x07C4ACDDU) >> 27];
#   endif
    }
}
 
/*=====    Local Constants   =====*/
static const unsigned BIT_mask[] = {
    0,          1,         3,         7,         0xF,       0x1F,
    0x3F,       0x7F,      0xFF,      0x1FF,     0x3FF,     0x7FF,
    0xFFF,      0x1FFF,    0x3FFF,    0x7FFF,    0xFFFF,    0x1FFFF,
    0x3FFFF,    0x7FFFF,   0xFFFFF,   0x1FFFFF,  0x3FFFFF,  0x7FFFFF,
    0xFFFFFF,   0x1FFFFFF, 0x3FFFFFF, 0x7FFFFFF, 0xFFFFFFF, 0x1FFFFFFF,
    0x3FFFFFFF, 0x7FFFFFFF}; /* up to 31 bits */
#define BIT_MASK_SIZE (sizeof(BIT_mask) / sizeof(BIT_mask[0]))
 
/*-**************************************************************
*  bitStream encoding
****************************************************************/
MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC,
                                  void* startPtr, size_t dstCapacity)
{
    bitC->bitContainer = 0;
    bitC->bitPos = 0;
    bitC->startPtr = (char*)startPtr;
    bitC->ptr = bitC->startPtr;
    bitC->endPtr = bitC->startPtr + dstCapacity - sizeof(bitC->bitContainer);
    if (dstCapacity <= sizeof(bitC->bitContainer)) return ERROR(dstSize_tooSmall);
    return 0;
}
 
MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC,
                            size_t value, unsigned nbBits)
{
    MEM_STATIC_ASSERT(BIT_MASK_SIZE == 32);
    assert(nbBits < BIT_MASK_SIZE);
    assert(nbBits + bitC->bitPos < sizeof(bitC->bitContainer) * 8);
    bitC->bitContainer |= (value & BIT_mask[nbBits]) << bitC->bitPos;
    bitC->bitPos += nbBits;
}
 
MEM_STATIC void BIT_addBitsFast(BIT_CStream_t* bitC,
                                size_t value, unsigned nbBits)
{
    assert((value>>nbBits) == 0);
    assert(nbBits + bitC->bitPos < sizeof(bitC->bitContainer) * 8);
    bitC->bitContainer |= value << bitC->bitPos;
    bitC->bitPos += nbBits;
}
 
MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC)
{
    size_t const nbBytes = bitC->bitPos >> 3;
    assert(bitC->bitPos < sizeof(bitC->bitContainer) * 8);
    assert(bitC->ptr <= bitC->endPtr);
    MEM_writeLEST(bitC->ptr, bitC->bitContainer);
    bitC->ptr += nbBytes;
    bitC->bitPos &= 7;
    bitC->bitContainer >>= nbBytes*8;
}
 
MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC)
{
    size_t const nbBytes = bitC->bitPos >> 3;
    assert(bitC->bitPos < sizeof(bitC->bitContainer) * 8);
    assert(bitC->ptr <= bitC->endPtr);
    MEM_writeLEST(bitC->ptr, bitC->bitContainer);
    bitC->ptr += nbBytes;
    if (bitC->ptr > bitC->endPtr) bitC->ptr = bitC->endPtr;
    bitC->bitPos &= 7;
    bitC->bitContainer >>= nbBytes*8;
}
 
MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC)
{
    BIT_addBitsFast(bitC, 1, 1);   /* endMark */
    BIT_flushBits(bitC);
    if (bitC->ptr >= bitC->endPtr) return 0; /* overflow detected */
    return (bitC->ptr - bitC->startPtr) + (bitC->bitPos > 0);
}
 
 
/*-********************************************************
*  bitStream decoding
**********************************************************/
MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize)
{
    if (srcSize < 1) { memset(bitD, 0, sizeof(*bitD)); return ERROR(srcSize_wrong); }
 
    bitD->start = (const char*)srcBuffer;
    bitD->limitPtr = bitD->start + sizeof(bitD->bitContainer);
 
    if (srcSize >=  sizeof(bitD->bitContainer)) {  /* normal case */
        bitD->ptr   = (const char*)srcBuffer + srcSize - sizeof(bitD->bitContainer);
        bitD->bitContainer = MEM_readLEST(bitD->ptr);
        { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1];
          bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0;  /* ensures bitsConsumed is always set */
          if (lastByte == 0) return ERROR(GENERIC); /* endMark not present */ }
    } else {
        bitD->ptr   = bitD->start;
        bitD->bitContainer = *(const BYTE*)(bitD->start);
        switch(srcSize)
        {
        case 7: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16);
                /* fall-through */
 
        case 6: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24);
                /* fall-through */
 
        case 5: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32);
                /* fall-through */
 
        case 4: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[3]) << 24;
                /* fall-through */
 
        case 3: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[2]) << 16;
                /* fall-through */
 
        case 2: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[1]) <<  8;
                /* fall-through */
 
        default: break;
        }
        {   BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1];
            bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0;
            if (lastByte == 0) return ERROR(corruption_detected);  /* endMark not present */
        }
        bitD->bitsConsumed += (U32)(sizeof(bitD->bitContainer) - srcSize)*8;
    }
 
    return srcSize;
}
 
MEM_STATIC size_t BIT_getUpperBits(size_t bitContainer, U32 const start)
{
    return bitContainer >> start;
}
 
MEM_STATIC size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits)
{
    U32 const regMask = sizeof(bitContainer)*8 - 1;
    /* if start > regMask, bitstream is corrupted, and result is undefined */
    assert(nbBits < BIT_MASK_SIZE);
    return (bitContainer >> (start & regMask)) & BIT_mask[nbBits];
}
 
MEM_STATIC size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
{
    assert(nbBits < BIT_MASK_SIZE);
    return bitContainer & BIT_mask[nbBits];
}
 
MEM_STATIC size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits)
{
    /* arbitrate between double-shift and shift+mask */
#if 1
    /* if bitD->bitsConsumed + nbBits > sizeof(bitD->bitContainer)*8,
     * bitstream is likely corrupted, and result is undefined */
    return BIT_getMiddleBits(bitD->bitContainer, (sizeof(bitD->bitContainer)*8) - bitD->bitsConsumed - nbBits, nbBits);
#else
    /* this code path is slower on my os-x laptop */
    U32 const regMask = sizeof(bitD->bitContainer)*8 - 1;
    return ((bitD->bitContainer << (bitD->bitsConsumed & regMask)) >> 1) >> ((regMask-nbBits) & regMask);
#endif
}
 
MEM_STATIC size_t BIT_lookBitsFast(const BIT_DStream_t* bitD, U32 nbBits)
{
    U32 const regMask = sizeof(bitD->bitContainer)*8 - 1;
    assert(nbBits >= 1);
    return (bitD->bitContainer << (bitD->bitsConsumed & regMask)) >> (((regMask+1)-nbBits) & regMask);
}
 
MEM_STATIC void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
{
    bitD->bitsConsumed += nbBits;
}
 
MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits)
{
    size_t const value = BIT_lookBits(bitD, nbBits);
    BIT_skipBits(bitD, nbBits);
    return value;
}
 
MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits)
{
    size_t const value = BIT_lookBitsFast(bitD, nbBits);
    assert(nbBits >= 1);
    BIT_skipBits(bitD, nbBits);
    return value;
}
 
MEM_STATIC BIT_DStream_status BIT_reloadDStreamFast(BIT_DStream_t* bitD)
{
    if (UNLIKELY(bitD->ptr < bitD->limitPtr))
        return BIT_DStream_overflow;
    assert(bitD->bitsConsumed <= sizeof(bitD->bitContainer)*8);
    bitD->ptr -= bitD->bitsConsumed >> 3;
    bitD->bitsConsumed &= 7;
    bitD->bitContainer = MEM_readLEST(bitD->ptr);
    return BIT_DStream_unfinished;
}
 
MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
{
    if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8))  /* overflow detected, like end of stream */
        return BIT_DStream_overflow;
 
    if (bitD->ptr >= bitD->limitPtr) {
        return BIT_reloadDStreamFast(bitD);
    }
    if (bitD->ptr == bitD->start) {
        if (bitD->bitsConsumed < sizeof(bitD->bitContainer)*8) return BIT_DStream_endOfBuffer;
        return BIT_DStream_completed;
    }
    /* start < ptr < limitPtr */
    {   U32 nbBytes = bitD->bitsConsumed >> 3;
        BIT_DStream_status result = BIT_DStream_unfinished;
        if (bitD->ptr - nbBytes < bitD->start) {
            nbBytes = (U32)(bitD->ptr - bitD->start);  /* ptr > start */
            result = BIT_DStream_endOfBuffer;
        }
        bitD->ptr -= nbBytes;
        bitD->bitsConsumed -= nbBytes*8;
        bitD->bitContainer = MEM_readLEST(bitD->ptr);   /* reminder : srcSize > sizeof(bitD->bitContainer), otherwise bitD->ptr == bitD->start */
        return result;
    }
}
 
MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* DStream)
{
    return ((DStream->ptr == DStream->start) && (DStream->bitsConsumed == sizeof(DStream->bitContainer)*8));
}
 
#if defined (__cplusplus)
}
#endif
 
#endif /* BITSTREAM_H_MODULE */