contrib/python-zstandard/zstd/common/entropy_common.c
author Gregory Szorc <gregory.szorc@gmail.com>
Mon, 24 Sep 2018 14:31:31 -0700
changeset 39878 3e896b51aa5d
parent 37495 b1fb341d8a61
child 40121 73fef626dae3
permissions -rw-r--r--
storageutil: move metadata parsing and packing from revlog (API) Parsing and writing of revision text metadata is likely identical across storage backends. Let's move the code out of revlog so we don't need to import the revlog module in order to use it. Differential Revision: https://phab.mercurial-scm.org/D4754
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     1
/*
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     2
   Common functions of New Generation Entropy library
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     3
   Copyright (C) 2016, Yann Collet.
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     4
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     5
   BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     6
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     7
   Redistribution and use in source and binary forms, with or without
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     8
   modification, are permitted provided that the following conditions are
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     9
   met:
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    10
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    11
       * Redistributions of source code must retain the above copyright
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    12
   notice, this list of conditions and the following disclaimer.
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    13
       * Redistributions in binary form must reproduce the above
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    14
   copyright notice, this list of conditions and the following disclaimer
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    15
   in the documentation and/or other materials provided with the
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    16
   distribution.
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    17
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    18
   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    19
   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    20
   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    21
   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    22
   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    23
   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    24
   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    25
   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    26
   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    27
   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    28
   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    29
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    30
    You can contact the author at :
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    31
    - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    32
    - Public forum : https://groups.google.com/forum/#!forum/lz4c
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    33
*************************************************************************** */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    34
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    35
/* *************************************
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    36
*  Dependencies
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    37
***************************************/
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    38
#include "mem.h"
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    39
#include "error_private.h"       /* ERR_*, ERROR */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    40
#define FSE_STATIC_LINKING_ONLY  /* FSE_MIN_TABLELOG */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    41
#include "fse.h"
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    42
#define HUF_STATIC_LINKING_ONLY  /* HUF_TABLELOG_ABSOLUTEMAX */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    43
#include "huf.h"
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    44
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    45
37495
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
    46
/*===   Version   ===*/
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
    47
unsigned FSE_versionNumber(void) { return FSE_VERSION_NUMBER; }
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
    48
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
    49
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
    50
/*===   Error Management   ===*/
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    51
unsigned FSE_isError(size_t code) { return ERR_isError(code); }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    52
const char* FSE_getErrorName(size_t code) { return ERR_getErrorName(code); }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    53
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    54
unsigned HUF_isError(size_t code) { return ERR_isError(code); }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    55
const char* HUF_getErrorName(size_t code) { return ERR_getErrorName(code); }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    56
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    57
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    58
/*-**************************************************************
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    59
*  FSE NCount encoding-decoding
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    60
****************************************************************/
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    61
size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    62
                 const void* headerBuffer, size_t hbSize)
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    63
{
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    64
    const BYTE* const istart = (const BYTE*) headerBuffer;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    65
    const BYTE* const iend = istart + hbSize;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    66
    const BYTE* ip = istart;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    67
    int nbBits;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    68
    int remaining;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    69
    int threshold;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    70
    U32 bitStream;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    71
    int bitCount;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    72
    unsigned charnum = 0;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    73
    int previous0 = 0;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    74
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    75
    if (hbSize < 4) return ERROR(srcSize_wrong);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    76
    bitStream = MEM_readLE32(ip);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    77
    nbBits = (bitStream & 0xF) + FSE_MIN_TABLELOG;   /* extract tableLog */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    78
    if (nbBits > FSE_TABLELOG_ABSOLUTE_MAX) return ERROR(tableLog_tooLarge);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    79
    bitStream >>= 4;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    80
    bitCount = 4;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    81
    *tableLogPtr = nbBits;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    82
    remaining = (1<<nbBits)+1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    83
    threshold = 1<<nbBits;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    84
    nbBits++;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    85
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    86
    while ((remaining>1) & (charnum<=*maxSVPtr)) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    87
        if (previous0) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    88
            unsigned n0 = charnum;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    89
            while ((bitStream & 0xFFFF) == 0xFFFF) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    90
                n0 += 24;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    91
                if (ip < iend-5) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    92
                    ip += 2;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    93
                    bitStream = MEM_readLE32(ip) >> bitCount;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    94
                } else {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    95
                    bitStream >>= 16;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    96
                    bitCount   += 16;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    97
            }   }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    98
            while ((bitStream & 3) == 3) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    99
                n0 += 3;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   100
                bitStream >>= 2;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   101
                bitCount += 2;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   102
            }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   103
            n0 += bitStream & 3;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   104
            bitCount += 2;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   105
            if (n0 > *maxSVPtr) return ERROR(maxSymbolValue_tooSmall);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   106
            while (charnum < n0) normalizedCounter[charnum++] = 0;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   107
            if ((ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   108
                ip += bitCount>>3;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   109
                bitCount &= 7;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   110
                bitStream = MEM_readLE32(ip) >> bitCount;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   111
            } else {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   112
                bitStream >>= 2;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   113
        }   }
37495
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   114
        {   int const max = (2*threshold-1) - remaining;
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   115
            int count;
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   116
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   117
            if ((bitStream & (threshold-1)) < (U32)max) {
37495
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   118
                count = bitStream & (threshold-1);
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   119
                bitCount += nbBits-1;
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   120
            } else {
37495
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   121
                count = bitStream & (2*threshold-1);
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   122
                if (count >= threshold) count -= max;
37495
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   123
                bitCount += nbBits;
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   124
            }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   125
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   126
            count--;   /* extra accuracy */
37495
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   127
            remaining -= count < 0 ? -count : count;   /* -1 means +1 */
b1fb341d8a61 zstandard: vendor python-zstandard 0.9.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30822
diff changeset
   128
            normalizedCounter[charnum++] = (short)count;
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   129
            previous0 = !count;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   130
            while (remaining < threshold) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   131
                nbBits--;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   132
                threshold >>= 1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   133
            }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   134
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   135
            if ((ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   136
                ip += bitCount>>3;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   137
                bitCount &= 7;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   138
            } else {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   139
                bitCount -= (int)(8 * (iend - 4 - ip));
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   140
                ip = iend - 4;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   141
            }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   142
            bitStream = MEM_readLE32(ip) >> (bitCount & 31);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   143
    }   }   /* while ((remaining>1) & (charnum<=*maxSVPtr)) */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   144
    if (remaining != 1) return ERROR(corruption_detected);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   145
    if (bitCount > 32) return ERROR(corruption_detected);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   146
    *maxSVPtr = charnum-1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   147
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   148
    ip += (bitCount+7)>>3;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   149
    return ip-istart;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   150
}
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   151
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   152
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   153
/*! HUF_readStats() :
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   154
    Read compact Huffman tree, saved by HUF_writeCTable().
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   155
    `huffWeight` is destination buffer.
30822
b54a2984cdd4 zstd: vendor python-zstandard 0.6.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30434
diff changeset
   156
    `rankStats` is assumed to be a table of at least HUF_TABLELOG_MAX U32.
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   157
    @return : size read from `src` , or an error Code .
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   158
    Note : Needed by HUF_readCTable() and HUF_readDTableX?() .
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   159
*/
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   160
size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   161
                     U32* nbSymbolsPtr, U32* tableLogPtr,
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   162
                     const void* src, size_t srcSize)
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   163
{
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   164
    U32 weightTotal;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   165
    const BYTE* ip = (const BYTE*) src;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   166
    size_t iSize;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   167
    size_t oSize;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   168
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   169
    if (!srcSize) return ERROR(srcSize_wrong);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   170
    iSize = ip[0];
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   171
    /* memset(huffWeight, 0, hwSize);   *//* is not necessary, even though some analyzer complain ... */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   172
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   173
    if (iSize >= 128) {  /* special header */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   174
        oSize = iSize - 127;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   175
        iSize = ((oSize+1)/2);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   176
        if (iSize+1 > srcSize) return ERROR(srcSize_wrong);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   177
        if (oSize >= hwSize) return ERROR(corruption_detected);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   178
        ip += 1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   179
        {   U32 n;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   180
            for (n=0; n<oSize; n+=2) {
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   181
                huffWeight[n]   = ip[n/2] >> 4;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   182
                huffWeight[n+1] = ip[n/2] & 15;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   183
    }   }   }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   184
    else  {   /* header compressed with FSE (normal case) */
30822
b54a2984cdd4 zstd: vendor python-zstandard 0.6.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30434
diff changeset
   185
        FSE_DTable fseWorkspace[FSE_DTABLE_SIZE_U32(6)];  /* 6 is max possible tableLog for HUF header (maybe even 5, to be tested) */
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   186
        if (iSize+1 > srcSize) return ERROR(srcSize_wrong);
30822
b54a2984cdd4 zstd: vendor python-zstandard 0.6.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30434
diff changeset
   187
        oSize = FSE_decompress_wksp(huffWeight, hwSize-1, ip+1, iSize, fseWorkspace, 6);   /* max (hwSize-1) values decoded, as last one is implied */
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   188
        if (FSE_isError(oSize)) return oSize;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   189
    }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   190
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   191
    /* collect weight stats */
30822
b54a2984cdd4 zstd: vendor python-zstandard 0.6.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30434
diff changeset
   192
    memset(rankStats, 0, (HUF_TABLELOG_MAX + 1) * sizeof(U32));
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   193
    weightTotal = 0;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   194
    {   U32 n; for (n=0; n<oSize; n++) {
30822
b54a2984cdd4 zstd: vendor python-zstandard 0.6.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30434
diff changeset
   195
            if (huffWeight[n] >= HUF_TABLELOG_MAX) return ERROR(corruption_detected);
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   196
            rankStats[huffWeight[n]]++;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   197
            weightTotal += (1 << huffWeight[n]) >> 1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   198
    }   }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   199
    if (weightTotal == 0) return ERROR(corruption_detected);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   200
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   201
    /* get last non-null symbol weight (implied, total must be 2^n) */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   202
    {   U32 const tableLog = BIT_highbit32(weightTotal) + 1;
30822
b54a2984cdd4 zstd: vendor python-zstandard 0.6.0
Gregory Szorc <gregory.szorc@gmail.com>
parents: 30434
diff changeset
   203
        if (tableLog > HUF_TABLELOG_MAX) return ERROR(corruption_detected);
30434
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   204
        *tableLogPtr = tableLog;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   205
        /* determine last weight */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   206
        {   U32 const total = 1 << tableLog;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   207
            U32 const rest = total - weightTotal;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   208
            U32 const verif = 1 << BIT_highbit32(rest);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   209
            U32 const lastWeight = BIT_highbit32(rest) + 1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   210
            if (verif != rest) return ERROR(corruption_detected);    /* last value must be a clean power of 2 */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   211
            huffWeight[oSize] = (BYTE)lastWeight;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   212
            rankStats[lastWeight]++;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   213
    }   }
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   214
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   215
    /* check tree construction validity */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   216
    if ((rankStats[1] < 2) || (rankStats[1] & 1)) return ERROR(corruption_detected);   /* by construction : at least 2 elts of rank 1, must be even */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   217
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   218
    /* results */
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   219
    *nbSymbolsPtr = (U32)(oSize+1);
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   220
    return iSize+1;
2e484bdea8c4 zstd: vendor zstd 1.1.1
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
   221
}