| 1 |
/* |
| 2 |
* Copyright (c) 2000-2005 Marc Alexander Lehmann <schmorp@schmorp.de> |
| 3 |
* |
| 4 |
* Redistribution and use in source and binary forms, with or without modifica- |
| 5 |
* tion, are permitted provided that the following conditions are met: |
| 6 |
* |
| 7 |
* 1. Redistributions of source code must retain the above copyright notice, |
| 8 |
* this list of conditions and the following disclaimer. |
| 9 |
* |
| 10 |
* 2. Redistributions in binary form must reproduce the above copyright |
| 11 |
* notice, this list of conditions and the following disclaimer in the |
| 12 |
* documentation and/or other materials provided with the distribution. |
| 13 |
* |
| 14 |
* 3. The name of the author may not be used to endorse or promote products |
| 15 |
* derived from this software without specific prior written permission. |
| 16 |
* |
| 17 |
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED |
| 18 |
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER- |
| 19 |
* CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO |
| 20 |
* EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPE- |
| 21 |
* CIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, |
| 22 |
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; |
| 23 |
* OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, |
| 24 |
* WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTH- |
| 25 |
* ERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED |
| 26 |
* OF THE POSSIBILITY OF SUCH DAMAGE. |
| 27 |
* |
| 28 |
* Alternatively, the contents of this file may be used under the terms of |
| 29 |
* the GNU General Public License version 2 (the "GPL"), in which case the |
| 30 |
* provisions of the GPL are applicable instead of the above. If you wish to |
| 31 |
* allow the use of your version of this file only under the terms of the |
| 32 |
* GPL and not to allow others to use your version of this file under the |
| 33 |
* BSD license, indicate your decision by deleting the provisions above and |
| 34 |
* replace them with the notice and other provisions required by the GPL. If |
| 35 |
* you do not delete the provisions above, a recipient may use your version |
| 36 |
* of this file under either the BSD or the GPL. |
| 37 |
*/ |
| 38 |
|
| 39 |
#include "lzfP.h" |
| 40 |
|
| 41 |
#define HSIZE (1 << (HLOG)) |
| 42 |
|
| 43 |
/* |
| 44 |
* don't play with this unless you benchmark! |
| 45 |
* decompression is not dependent on the hash function |
| 46 |
* the hashing function might seem strange, just believe me |
| 47 |
* it works ;) |
| 48 |
*/ |
| 49 |
#ifndef FRST |
| 50 |
# define FRST(p) (((p[0]) << 8) | p[1]) |
| 51 |
# define NEXT(v,p) (((v) << 8) | p[2]) |
| 52 |
# define IDX(h) ((((h ^ (h << 5)) >> (3*8 - HLOG)) - h*5) & (HSIZE - 1)) |
| 53 |
/*# define IDX(h) ((ip[0] * 121 ^ ip[1] * 33 ^ ip[2] * 1) & (HSIZE-1))*/ |
| 54 |
#endif |
| 55 |
/* |
| 56 |
* IDX works because it is very similar to a multiplicative hash, e.g. |
| 57 |
* ((h * 57321 >> (3*8 - HLOG)) & (HSIZE - 1)) |
| 58 |
* the latter is also quite fast on newer CPUs, and compresses similarly. |
| 59 |
* |
| 60 |
* the next one is also quite good, albeit slow ;) |
| 61 |
* (int)(cos(h & 0xffffff) * 1e6) |
| 62 |
*/ |
| 63 |
|
| 64 |
#if 0 |
| 65 |
/* original lzv-like hash function, much worse and thus slower */ |
| 66 |
# define FRST(p) (p[0] << 5) ^ p[1] |
| 67 |
# define NEXT(v,p) ((v) << 5) ^ p[2] |
| 68 |
# define IDX(h) ((h) & (HSIZE - 1)) |
| 69 |
#endif |
| 70 |
|
| 71 |
#define MAX_LIT (1 << 5) |
| 72 |
#define MAX_OFF (1 << 13) |
| 73 |
#define MAX_REF ((1 << 8) + (1 << 3)) |
| 74 |
|
| 75 |
/* |
| 76 |
* compressed format |
| 77 |
* |
| 78 |
* 000LLLLL <L+1> ; literal |
| 79 |
* LLLooooo oooooooo ; backref L |
| 80 |
* 111ooooo LLLLLLLL oooooooo ; backref L+7 |
| 81 |
* |
| 82 |
*/ |
| 83 |
|
| 84 |
unsigned int |
| 85 |
lzf_compress (const void *const in_data, unsigned int in_len, |
| 86 |
void *out_data, unsigned int out_len |
| 87 |
#if LZF_STATE_ARG |
| 88 |
, LZF_STATE htab |
| 89 |
#endif |
| 90 |
) |
| 91 |
{ |
| 92 |
#if !LZF_STATE_ARG |
| 93 |
LZF_STATE htab; |
| 94 |
#endif |
| 95 |
const u8 **hslot; |
| 96 |
const u8 *ip = (const u8 *)in_data; |
| 97 |
u8 *op = (u8 *)out_data; |
| 98 |
const u8 *in_end = ip + in_len; |
| 99 |
u8 *out_end = op + out_len; |
| 100 |
const u8 *ref; |
| 101 |
|
| 102 |
unsigned int hval = FRST (ip); |
| 103 |
unsigned long off; |
| 104 |
int lit = 0; |
| 105 |
|
| 106 |
#if INIT_HTAB |
| 107 |
# if USE_MEMCPY |
| 108 |
memset (htab, 0, sizeof (htab)); |
| 109 |
# else |
| 110 |
for (hslot = htab; hslot < htab + HSIZE; hslot++) |
| 111 |
*hslot++ = ip; |
| 112 |
# endif |
| 113 |
#endif |
| 114 |
|
| 115 |
for (;;) |
| 116 |
{ |
| 117 |
if (ip < in_end - 2) |
| 118 |
{ |
| 119 |
hval = NEXT (hval, ip); |
| 120 |
hslot = htab + IDX (hval); |
| 121 |
ref = *hslot; *hslot = ip; |
| 122 |
|
| 123 |
if (1 |
| 124 |
#if INIT_HTAB && !USE_MEMCPY |
| 125 |
&& ref < ip /* the next test will actually take care of this, but this is faster */ |
| 126 |
#endif |
| 127 |
&& (off = ip - ref - 1) < MAX_OFF |
| 128 |
&& ip + 4 < in_end |
| 129 |
&& ref > (u8 *)in_data |
| 130 |
#if STRICT_ALIGN |
| 131 |
&& ref[0] == ip[0] |
| 132 |
&& ref[1] == ip[1] |
| 133 |
&& ref[2] == ip[2] |
| 134 |
#else |
| 135 |
&& *(u16 *)ref == *(u16 *)ip |
| 136 |
&& ref[2] == ip[2] |
| 137 |
#endif |
| 138 |
) |
| 139 |
{ |
| 140 |
/* match found at *ref++ */ |
| 141 |
unsigned int len = 2; |
| 142 |
unsigned int maxlen = in_end - ip - len; |
| 143 |
maxlen = maxlen > MAX_REF ? MAX_REF : maxlen; |
| 144 |
|
| 145 |
if (op + lit + 1 + 3 >= out_end) |
| 146 |
return 0; |
| 147 |
|
| 148 |
do |
| 149 |
len++; |
| 150 |
while (len < maxlen && ref[len] == ip[len]); |
| 151 |
|
| 152 |
if (lit) |
| 153 |
{ |
| 154 |
*op++ = lit - 1; |
| 155 |
lit = -lit; |
| 156 |
do |
| 157 |
*op++ = ip[lit]; |
| 158 |
while (++lit); |
| 159 |
} |
| 160 |
|
| 161 |
len -= 2; |
| 162 |
ip++; |
| 163 |
|
| 164 |
if (len < 7) |
| 165 |
{ |
| 166 |
*op++ = (off >> 8) + (len << 5); |
| 167 |
} |
| 168 |
else |
| 169 |
{ |
| 170 |
*op++ = (off >> 8) + ( 7 << 5); |
| 171 |
*op++ = len - 7; |
| 172 |
} |
| 173 |
|
| 174 |
*op++ = off; |
| 175 |
|
| 176 |
#if ULTRA_FAST || VERY_FAST |
| 177 |
ip += len; |
| 178 |
#if VERY_FAST && !ULTRA_FAST |
| 179 |
--ip; |
| 180 |
#endif |
| 181 |
hval = FRST (ip); |
| 182 |
|
| 183 |
hval = NEXT (hval, ip); |
| 184 |
htab[IDX (hval)] = ip; |
| 185 |
ip++; |
| 186 |
|
| 187 |
#if VERY_FAST && !ULTRA_FAST |
| 188 |
hval = NEXT (hval, ip); |
| 189 |
htab[IDX (hval)] = ip; |
| 190 |
ip++; |
| 191 |
#endif |
| 192 |
#else |
| 193 |
do |
| 194 |
{ |
| 195 |
hval = NEXT (hval, ip); |
| 196 |
htab[IDX (hval)] = ip; |
| 197 |
ip++; |
| 198 |
} |
| 199 |
while (len--); |
| 200 |
#endif |
| 201 |
continue; |
| 202 |
} |
| 203 |
} |
| 204 |
else if (ip == in_end) |
| 205 |
break; |
| 206 |
|
| 207 |
/* one more literal byte we must copy */ |
| 208 |
lit++; |
| 209 |
ip++; |
| 210 |
|
| 211 |
if (lit == MAX_LIT) |
| 212 |
{ |
| 213 |
if (op + 1 + MAX_LIT >= out_end) |
| 214 |
return 0; |
| 215 |
|
| 216 |
*op++ = MAX_LIT - 1; |
| 217 |
#if USE_MEMCPY |
| 218 |
memcpy (op, ip - MAX_LIT, MAX_LIT); |
| 219 |
op += MAX_LIT; |
| 220 |
lit = 0; |
| 221 |
#else |
| 222 |
lit = -lit; |
| 223 |
do |
| 224 |
*op++ = ip[lit]; |
| 225 |
while (++lit); |
| 226 |
#endif |
| 227 |
} |
| 228 |
} |
| 229 |
|
| 230 |
if (lit) |
| 231 |
{ |
| 232 |
if (op + lit + 1 >= out_end) |
| 233 |
return 0; |
| 234 |
|
| 235 |
*op++ = lit - 1; |
| 236 |
lit = -lit; |
| 237 |
do |
| 238 |
*op++ = ip[lit]; |
| 239 |
while (++lit); |
| 240 |
} |
| 241 |
|
| 242 |
return op - (u8 *) out_data; |
| 243 |
} |
