CroakFang
525 lines
14 KiB
C#
525 lines
14 KiB
C#
#region LZ4 original
|
|
|
|
/*
|
|
LZ4 - Fast LZ compression algorithm
|
|
Copyright (C) 2011-2012, Yann Collet.
|
|
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
|
|
|
|
Redistribution and use in source and binary forms, with or without
|
|
modification, are permitted provided that the following conditions are
|
|
met:
|
|
|
|
* Redistributions of source code must retain the above copyright
|
|
notice, this list of conditions and the following disclaimer.
|
|
* Redistributions in binary form must reproduce the above
|
|
copyright notice, this list of conditions and the following disclaimer
|
|
in the documentation and/or other materials provided with the
|
|
distribution.
|
|
|
|
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
|
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
|
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
|
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
|
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
|
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
|
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
You can contact the author at :
|
|
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
|
|
- LZ4 source repository : http://code.google.com/p/lz4/
|
|
*/
|
|
|
|
#endregion
|
|
|
|
#region LZ4 port
|
|
|
|
/*
|
|
Copyright (c) 2013, Milosz Krajewski
|
|
All rights reserved.
|
|
|
|
Redistribution and use in source and binary forms, with or without modification, are permitted provided
|
|
that the following conditions are met:
|
|
|
|
* Redistributions of source code must retain the above copyright notice, this list of conditions
|
|
and the following disclaimer.
|
|
|
|
* Redistributions in binary form must reproduce the above copyright notice, this list of conditions
|
|
and the following disclaimer in the documentation and/or other materials provided with the distribution.
|
|
|
|
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED
|
|
WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
|
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
|
|
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
|
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
|
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
|
|
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN
|
|
IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
*/
|
|
|
|
#endregion
|
|
|
|
// ReSharper disable InconsistentNaming
|
|
|
|
namespace LZ4ps
|
|
{
|
|
public static partial class LZ4Codec
|
|
{
|
|
// Update chains up to ip (excluded)
|
|
private static void LZ4HC_Insert_32(LZ4HC_Data_Structure ctx, int src_p)
|
|
{
|
|
var chainTable = ctx.chainTable;
|
|
var hashTable = ctx.hashTable;
|
|
var nextToUpdate = ctx.nextToUpdate;
|
|
var src = ctx.src;
|
|
var src_base = ctx.src_base;
|
|
|
|
while (nextToUpdate < src_p)
|
|
{
|
|
var p = nextToUpdate;
|
|
var delta = (p) - (hashTable[(((Peek4(src, p)) * 2654435761u) >> HASHHC_ADJUST)] + src_base);
|
|
if (delta > MAX_DISTANCE) delta = MAX_DISTANCE;
|
|
chainTable[(p) & MAXD_MASK] = (ushort)delta;
|
|
hashTable[(((Peek4(src, p)) * 2654435761u) >> HASHHC_ADJUST)] = ((p) - src_base);
|
|
nextToUpdate++;
|
|
}
|
|
|
|
ctx.nextToUpdate = nextToUpdate;
|
|
}
|
|
|
|
private static int LZ4HC_CommonLength_32(LZ4HC_Data_Structure ctx, int p1, int p2)
|
|
{
|
|
var debruijn32 = DEBRUIJN_TABLE_32;
|
|
var src = ctx.src;
|
|
var src_LASTLITERALS = ctx.src_LASTLITERALS;
|
|
|
|
var p1t = p1;
|
|
|
|
while (p1t < src_LASTLITERALS - (STEPSIZE_32 - 1))
|
|
{
|
|
var diff = (int)Xor4(src, p2, p1t);
|
|
if (diff == 0)
|
|
{
|
|
p1t += STEPSIZE_32;
|
|
p2 += STEPSIZE_32;
|
|
continue;
|
|
}
|
|
p1t += debruijn32[((uint)((diff) & -(diff)) * 0x077CB531u) >> 27];
|
|
return (p1t - p1);
|
|
}
|
|
if ((p1t < (src_LASTLITERALS - 1)) && (Equal2(src, p2, p1t)))
|
|
{
|
|
p1t += 2;
|
|
p2 += 2;
|
|
}
|
|
if ((p1t < src_LASTLITERALS) && (src[p2] == src[p1t])) p1t++;
|
|
return (p1t - p1);
|
|
}
|
|
|
|
private static int LZ4HC_InsertAndFindBestMatch_32(LZ4HC_Data_Structure ctx, int src_p, ref int src_match)
|
|
{
|
|
var chainTable = ctx.chainTable;
|
|
var hashTable = ctx.hashTable;
|
|
var src = ctx.src;
|
|
var src_base = ctx.src_base;
|
|
|
|
var nbAttempts = MAX_NB_ATTEMPTS;
|
|
int repl = 0, ml = 0;
|
|
ushort delta = 0;
|
|
|
|
// HC4 match finder
|
|
LZ4HC_Insert_32(ctx, src_p);
|
|
var src_ref = (hashTable[(((Peek4(src, src_p)) * 2654435761u) >> HASHHC_ADJUST)] + src_base);
|
|
|
|
|
|
// Detect repetitive sequences of length <= 4
|
|
if (src_ref >= src_p - 4) // potential repetition
|
|
{
|
|
if (Equal4(src, src_ref, src_p)) // confirmed
|
|
{
|
|
delta = (ushort)(src_p - src_ref);
|
|
repl = ml = LZ4HC_CommonLength_32(ctx, src_p + MINMATCH, src_ref + MINMATCH) + MINMATCH;
|
|
src_match = src_ref;
|
|
}
|
|
src_ref = ((src_ref) - chainTable[(src_ref) & MAXD_MASK]);
|
|
}
|
|
|
|
while ((src_ref >= src_p - MAX_DISTANCE) && (nbAttempts != 0))
|
|
{
|
|
nbAttempts--;
|
|
if (src[(src_ref + ml)] == src[(src_p + ml)])
|
|
{
|
|
if (Equal4(src, src_ref, src_p))
|
|
{
|
|
var mlt = LZ4HC_CommonLength_32(ctx, src_p + MINMATCH, src_ref + MINMATCH) + MINMATCH;
|
|
if (mlt > ml)
|
|
{
|
|
ml = mlt;
|
|
src_match = src_ref;
|
|
}
|
|
}
|
|
}
|
|
src_ref = ((src_ref) - chainTable[(src_ref) & MAXD_MASK]);
|
|
}
|
|
|
|
|
|
// Complete table
|
|
if (repl != 0)
|
|
{
|
|
var src_ptr = src_p;
|
|
|
|
var end = src_p + repl - (MINMATCH - 1);
|
|
while (src_ptr < end - delta)
|
|
{
|
|
chainTable[(src_ptr) & MAXD_MASK] = delta; // Pre-Load
|
|
src_ptr++;
|
|
}
|
|
do
|
|
{
|
|
chainTable[(src_ptr) & MAXD_MASK] = delta;
|
|
hashTable[(((Peek4(src, src_ptr)) * 2654435761u) >> HASHHC_ADJUST)] = ((src_ptr) - src_base); // Head of chain
|
|
src_ptr++;
|
|
} while (src_ptr < end);
|
|
ctx.nextToUpdate = end;
|
|
}
|
|
|
|
return ml;
|
|
}
|
|
|
|
private static int LZ4HC_InsertAndGetWiderMatch_32(
|
|
LZ4HC_Data_Structure ctx, int src_p, int startLimit, int longest, ref int matchpos, ref int startpos)
|
|
{
|
|
var chainTable = ctx.chainTable;
|
|
var hashTable = ctx.hashTable;
|
|
var src = ctx.src;
|
|
var src_base = ctx.src_base;
|
|
var src_LASTLITERALS = ctx.src_LASTLITERALS;
|
|
var debruijn32 = DEBRUIJN_TABLE_32;
|
|
var nbAttempts = MAX_NB_ATTEMPTS;
|
|
var delta = (src_p - startLimit);
|
|
|
|
// First Match
|
|
LZ4HC_Insert_32(ctx, src_p);
|
|
var src_ref = (hashTable[(((Peek4(src, src_p)) * 2654435761u) >> HASHHC_ADJUST)] + src_base);
|
|
|
|
while ((src_ref >= src_p - MAX_DISTANCE) && (nbAttempts != 0))
|
|
{
|
|
nbAttempts--;
|
|
if (src[(startLimit + longest)] == src[(src_ref - delta + longest)])
|
|
{
|
|
if (Equal4(src, src_ref, src_p))
|
|
{
|
|
var reft = src_ref + MINMATCH;
|
|
var ipt = src_p + MINMATCH;
|
|
var startt = src_p;
|
|
|
|
while (ipt < src_LASTLITERALS - (STEPSIZE_32 - 1))
|
|
{
|
|
var diff = (int)Xor4(src, reft, ipt);
|
|
if (diff == 0)
|
|
{
|
|
ipt += STEPSIZE_32;
|
|
reft += STEPSIZE_32;
|
|
continue;
|
|
}
|
|
ipt += debruijn32[((uint)((diff) & -(diff)) * 0x077CB531u) >> 27];
|
|
goto _endCount;
|
|
}
|
|
if ((ipt < (src_LASTLITERALS - 1)) && (Equal2(src, reft, ipt)))
|
|
{
|
|
ipt += 2;
|
|
reft += 2;
|
|
}
|
|
if ((ipt < src_LASTLITERALS) && (src[reft] == src[ipt])) ipt++;
|
|
|
|
_endCount:
|
|
reft = src_ref;
|
|
|
|
while ((startt > startLimit) && (reft > src_base) && (src[startt - 1] == src[reft - 1]))
|
|
{
|
|
startt--;
|
|
reft--;
|
|
}
|
|
|
|
if ((ipt - startt) > longest)
|
|
{
|
|
longest = (ipt - startt);
|
|
matchpos = reft;
|
|
startpos = startt;
|
|
}
|
|
}
|
|
}
|
|
src_ref = ((src_ref) - chainTable[(src_ref) & MAXD_MASK]);
|
|
}
|
|
|
|
return longest;
|
|
}
|
|
|
|
private static int LZ4_encodeSequence_32(
|
|
LZ4HC_Data_Structure ctx, ref int src_p, ref int dst_p, ref int src_anchor, int matchLength, int src_ref, int dst_end)
|
|
{
|
|
int len;
|
|
var src = ctx.src;
|
|
var dst = ctx.dst;
|
|
|
|
// Encode Literal length
|
|
var length = src_p - src_anchor;
|
|
var dst_token = dst_p++;
|
|
if ((dst_p + length + (2 + 1 + LASTLITERALS) + (length >> 8)) > dst_end) return 1; // Check output limit
|
|
if (length >= RUN_MASK)
|
|
{
|
|
dst[dst_token] = (RUN_MASK << ML_BITS);
|
|
len = length - RUN_MASK;
|
|
for (; len > 254; len -= 255) dst[dst_p++] = 255;
|
|
dst[dst_p++] = (byte)len;
|
|
}
|
|
else
|
|
{
|
|
dst[dst_token] = (byte)(length << ML_BITS);
|
|
}
|
|
|
|
// Copy Literals
|
|
if (length > 0)
|
|
{
|
|
var _i = dst_p + length;
|
|
src_anchor += WildCopy(src, src_anchor, dst, dst_p, _i);
|
|
dst_p = _i;
|
|
}
|
|
|
|
// Encode Offset
|
|
Poke2(dst, dst_p, (ushort)(src_p - src_ref));
|
|
dst_p += 2;
|
|
|
|
// Encode MatchLength
|
|
len = (matchLength - MINMATCH);
|
|
if (dst_p + (1 + LASTLITERALS) + (length >> 8) > dst_end) return 1; // Check output limit
|
|
if (len >= ML_MASK)
|
|
{
|
|
dst[dst_token] += ML_MASK;
|
|
len -= ML_MASK;
|
|
for (; len > 509; len -= 510)
|
|
{
|
|
dst[(dst_p)++] = 255;
|
|
dst[(dst_p)++] = 255;
|
|
}
|
|
if (len > 254)
|
|
{
|
|
len -= 255;
|
|
dst[(dst_p)++] = 255;
|
|
}
|
|
dst[(dst_p)++] = (byte)len;
|
|
}
|
|
else
|
|
{
|
|
dst[dst_token] += (byte)len;
|
|
}
|
|
|
|
// Prepare next loop
|
|
src_p += matchLength;
|
|
src_anchor = src_p;
|
|
|
|
return 0;
|
|
}
|
|
|
|
private static int LZ4_compressHCCtx_32(LZ4HC_Data_Structure ctx)
|
|
{
|
|
var src = ctx.src;
|
|
var dst = ctx.dst;
|
|
var src_0 = ctx.src_base;
|
|
var src_end = ctx.src_end;
|
|
var dst_0 = ctx.dst_base;
|
|
var dst_len = ctx.dst_len;
|
|
var dst_end = ctx.dst_end;
|
|
|
|
var src_p = src_0;
|
|
var src_anchor = src_p;
|
|
var src_mflimit = src_end - MFLIMIT;
|
|
|
|
var dst_p = dst_0;
|
|
|
|
var src_ref = 0;
|
|
var start2 = 0;
|
|
var ref2 = 0;
|
|
var start3 = 0;
|
|
var ref3 = 0;
|
|
|
|
src_p++;
|
|
|
|
// Main Loop
|
|
while (src_p < src_mflimit)
|
|
{
|
|
var ml = LZ4HC_InsertAndFindBestMatch_32(ctx, src_p, ref src_ref);
|
|
if (ml == 0)
|
|
{
|
|
src_p++;
|
|
continue;
|
|
}
|
|
|
|
// saved, in case we would skip too much
|
|
var start0 = src_p;
|
|
var ref0 = src_ref;
|
|
var ml0 = ml;
|
|
|
|
_Search2:
|
|
var ml2 = src_p + ml < src_mflimit
|
|
? LZ4HC_InsertAndGetWiderMatch_32(ctx, src_p + ml - 2, src_p + 1, ml, ref ref2, ref start2)
|
|
: ml;
|
|
|
|
if (ml2 == ml) // No better match
|
|
{
|
|
if (LZ4_encodeSequence_32(ctx, ref src_p, ref dst_p, ref src_anchor, ml, src_ref, dst_end) != 0) return 0;
|
|
continue;
|
|
}
|
|
|
|
if (start0 < src_p)
|
|
{
|
|
if (start2 < src_p + ml0) // empirical
|
|
{
|
|
src_p = start0;
|
|
src_ref = ref0;
|
|
ml = ml0;
|
|
}
|
|
}
|
|
|
|
// Here, start0==ip
|
|
if ((start2 - src_p) < 3) // First Match too small : removed
|
|
{
|
|
ml = ml2;
|
|
src_p = start2;
|
|
src_ref = ref2;
|
|
goto _Search2;
|
|
}
|
|
|
|
_Search3:
|
|
// Currently we have :
|
|
// ml2 > ml1, and
|
|
// ip1+3 <= ip2 (usually < ip1+ml1)
|
|
if ((start2 - src_p) < OPTIMAL_ML)
|
|
{
|
|
var new_ml = ml;
|
|
if (new_ml > OPTIMAL_ML) new_ml = OPTIMAL_ML;
|
|
if (src_p + new_ml > start2 + ml2 - MINMATCH) new_ml = (start2 - src_p) + ml2 - MINMATCH;
|
|
var correction = new_ml - (start2 - src_p);
|
|
if (correction > 0)
|
|
{
|
|
start2 += correction;
|
|
ref2 += correction;
|
|
ml2 -= correction;
|
|
}
|
|
}
|
|
// Now, we have start2 = ip+new_ml, with new_ml=min(ml, OPTIMAL_ML=18)
|
|
|
|
var ml3 = start2 + ml2 < src_mflimit
|
|
? LZ4HC_InsertAndGetWiderMatch_32(ctx, start2 + ml2 - 3, start2, ml2, ref ref3, ref start3)
|
|
: ml2;
|
|
|
|
if (ml3 == ml2) // No better match : 2 sequences to encode
|
|
{
|
|
// ip & ref are known; Now for ml
|
|
if (start2 < src_p + ml) ml = (start2 - src_p);
|
|
// Now, encode 2 sequences
|
|
if (LZ4_encodeSequence_32(ctx, ref src_p, ref dst_p, ref src_anchor, ml, src_ref, dst_end) != 0) return 0;
|
|
src_p = start2;
|
|
if (LZ4_encodeSequence_32(ctx, ref src_p, ref dst_p, ref src_anchor, ml2, ref2, dst_end) != 0) return 0;
|
|
continue;
|
|
}
|
|
|
|
if (start3 < src_p + ml + 3) // Not enough space for match 2 : remove it
|
|
{
|
|
if (start3 >= (src_p + ml)) // can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1
|
|
{
|
|
if (start2 < src_p + ml)
|
|
{
|
|
var correction = (src_p + ml - start2);
|
|
start2 += correction;
|
|
ref2 += correction;
|
|
ml2 -= correction;
|
|
if (ml2 < MINMATCH)
|
|
{
|
|
start2 = start3;
|
|
ref2 = ref3;
|
|
ml2 = ml3;
|
|
}
|
|
}
|
|
|
|
if (LZ4_encodeSequence_32(ctx, ref src_p, ref dst_p, ref src_anchor, ml, src_ref, dst_end) != 0) return 0;
|
|
src_p = start3;
|
|
src_ref = ref3;
|
|
ml = ml3;
|
|
|
|
start0 = start2;
|
|
ref0 = ref2;
|
|
ml0 = ml2;
|
|
goto _Search2;
|
|
}
|
|
|
|
start2 = start3;
|
|
ref2 = ref3;
|
|
ml2 = ml3;
|
|
goto _Search3;
|
|
}
|
|
|
|
// OK, now we have 3 ascending matches; let's write at least the first one
|
|
// ip & ref are known; Now for ml
|
|
if (start2 < src_p + ml)
|
|
{
|
|
if ((start2 - src_p) < ML_MASK)
|
|
{
|
|
if (ml > OPTIMAL_ML) ml = OPTIMAL_ML;
|
|
if (src_p + ml > start2 + ml2 - MINMATCH) ml = (start2 - src_p) + ml2 - MINMATCH;
|
|
var correction = ml - (start2 - src_p);
|
|
if (correction > 0)
|
|
{
|
|
start2 += correction;
|
|
ref2 += correction;
|
|
ml2 -= correction;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
ml = (start2 - src_p);
|
|
}
|
|
}
|
|
if (LZ4_encodeSequence_32(ctx, ref src_p, ref dst_p, ref src_anchor, ml, src_ref, dst_end) != 0) return 0;
|
|
|
|
src_p = start2;
|
|
src_ref = ref2;
|
|
ml = ml2;
|
|
|
|
start2 = start3;
|
|
ref2 = ref3;
|
|
ml2 = ml3;
|
|
|
|
goto _Search3;
|
|
}
|
|
|
|
// Encode Last Literals
|
|
{
|
|
var lastRun = (src_end - src_anchor);
|
|
if ((dst_p - dst_0) + lastRun + 1 + ((lastRun + 255 - RUN_MASK) / 255) > (uint)dst_len) return 0; // Check output limit
|
|
if (lastRun >= RUN_MASK)
|
|
{
|
|
dst[dst_p++] = (RUN_MASK << ML_BITS);
|
|
lastRun -= RUN_MASK;
|
|
for (; lastRun > 254; lastRun -= 255) dst[dst_p++] = 255;
|
|
dst[dst_p++] = (byte)lastRun;
|
|
}
|
|
else
|
|
{
|
|
dst[dst_p++] = (byte)(lastRun << ML_BITS);
|
|
}
|
|
BlockCopy(src, src_anchor, dst, dst_p, src_end - src_anchor);
|
|
dst_p += src_end - src_anchor;
|
|
}
|
|
|
|
// End
|
|
return (dst_p - dst_0);
|
|
}
|
|
}
|
|
}
|
|
|
|
// ReSharper restore InconsistentNaming
|