1 |
/* |
2 |
* Copyright (c) 2000-2002 Marc Alexander Lehmann <pcg@goof.com> |
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 |
|
29 |
#include "lzfP.h" |
30 |
|
31 |
#define HSIZE (1 << (HLOG)) |
32 |
|
33 |
/* |
34 |
* don't play with this unless you benchmark! |
35 |
* decompression is not dependent on the hash function |
36 |
* the hashing function might seem strange, just believe me |
37 |
* it works ;) |
38 |
*/ |
39 |
#define FRST(p) (((p[0]) << 8) + p[1]) |
40 |
#define NEXT(v,p) (((v) << 8) + p[2]) |
41 |
#define IDX(h) ((((h ^ (h << 5)) >> (3*8 - HLOG)) + h*3) & (HSIZE - 1)) |
42 |
/* |
43 |
* IDX works because it is very similar to a multiplicative hash, e.g. |
44 |
* (h * 57321 >> (3*8 - HLOG)) |
45 |
* the next one is also quite good, albeit slow ;) |
46 |
* (int)(cos(h & 0xffffff) * 1e6) |
47 |
*/ |
48 |
|
49 |
#if 0 |
50 |
/* original lzv-like hash function */ |
51 |
# define FRST(p) (p[0] << 5) ^ p[1] |
52 |
# define NEXT(v,p) ((v) << 5) ^ p[2] |
53 |
# define IDX(h) ((h) & (HSIZE - 1)) |
54 |
#endif |
55 |
|
56 |
#define MAX_LIT (1 << 5) |
57 |
#define MAX_OFF (1 << 13) |
58 |
#define MAX_REF ((1 << 8) + (1 << 3)) |
59 |
|
60 |
/* |
61 |
* compressed format |
62 |
* |
63 |
* 000LLLLL <L+1> ; literal |
64 |
* LLLOOOOO oooooooo ; backref L |
65 |
* 111OOOOO LLLLLLLL oooooooo ; backref L+7 |
66 |
* |
67 |
*/ |
68 |
|
69 |
unsigned int |
70 |
lzf_compress (const void *const in_data, unsigned int in_len, |
71 |
void *out_data, unsigned int out_len) |
72 |
{ |
73 |
const u8 *htab[HSIZE]; |
74 |
const u8 **hslot; |
75 |
const u8 *ip = in_data; |
76 |
u8 *op = out_data; |
77 |
const u8 *in_end = ip + in_len; |
78 |
u8 *out_end = op + out_len; |
79 |
const u8 *ref; |
80 |
|
81 |
unsigned int hval = FRST (ip); |
82 |
unsigned long off; |
83 |
int lit = 0; |
84 |
|
85 |
#if INIT_HTAB |
86 |
# if USE_MEMCPY |
87 |
memset (htab, 0, sizeof (htab)); |
88 |
# else |
89 |
for (hslot = htab; hslot < htab + HSIZE; hslot++) |
90 |
*hslot++ = ip; |
91 |
# endif |
92 |
#endif |
93 |
|
94 |
do |
95 |
{ |
96 |
hval = NEXT (hval, ip); |
97 |
hslot = htab + IDX (hval); |
98 |
ref = *hslot; *hslot = ip; |
99 |
|
100 |
if (1 |
101 |
#if INIT_HTAB && !USE_MEMCPY |
102 |
&& ref < ip /* the next test will actually take care of this, but it is faster */ |
103 |
#endif |
104 |
&& (off = ip - ref - 1) < MAX_OFF |
105 |
&& ip + 4 < in_end |
106 |
&& ref > (u8 *)in_data |
107 |
#if STRICT_ALIGN |
108 |
&& ref[0] == ip[0] |
109 |
&& ref[1] == ip[1] |
110 |
&& ref[2] == ip[2] |
111 |
#else |
112 |
&& *(u16 *)ref == *(u16 *)ip |
113 |
&& ref[2] == ip[2] |
114 |
#endif |
115 |
) |
116 |
{ |
117 |
/* match found at *ref++ */ |
118 |
unsigned int len = 2; |
119 |
unsigned int maxlen = in_end - ip - len; |
120 |
maxlen = maxlen > MAX_REF ? MAX_REF : maxlen; |
121 |
|
122 |
do |
123 |
len++; |
124 |
while (len < maxlen && ref[len] == ip[len]); |
125 |
|
126 |
if (op + lit + 1 + 3 >= out_end) |
127 |
return 0; |
128 |
|
129 |
if (lit) |
130 |
{ |
131 |
*op++ = lit - 1; |
132 |
lit = -lit; |
133 |
do |
134 |
*op++ = ip[lit]; |
135 |
while (++lit); |
136 |
} |
137 |
|
138 |
len -= 2; |
139 |
ip++; |
140 |
|
141 |
if (len < 7) |
142 |
{ |
143 |
*op++ = (off >> 8) + (len << 5); |
144 |
} |
145 |
else |
146 |
{ |
147 |
*op++ = (off >> 8) + ( 7 << 5); |
148 |
*op++ = len - 7; |
149 |
} |
150 |
|
151 |
*op++ = off; |
152 |
|
153 |
#if ULTRA_FAST |
154 |
ip += len; |
155 |
hval = FRST (ip); |
156 |
hval = NEXT (hval, ip); |
157 |
htab[IDX (hval)] = ip; |
158 |
ip++; |
159 |
#else |
160 |
do |
161 |
{ |
162 |
hval = NEXT (hval, ip); |
163 |
htab[IDX (hval)] = ip; |
164 |
ip++; |
165 |
} |
166 |
while (len--); |
167 |
#endif |
168 |
} |
169 |
else |
170 |
{ |
171 |
/* one more literal byte we must copy */ |
172 |
lit++; |
173 |
ip++; |
174 |
|
175 |
if (lit == MAX_LIT) |
176 |
{ |
177 |
if (op + 1 + MAX_LIT >= out_end) |
178 |
return 0; |
179 |
|
180 |
*op++ = MAX_LIT - 1; |
181 |
#if USE_MEMCPY |
182 |
memcpy (op, ip - MAX_LIT, MAX_LIT); |
183 |
op += MAX_LIT; |
184 |
lit = 0; |
185 |
#else |
186 |
lit = -lit; |
187 |
do |
188 |
*op++ = ip[lit]; |
189 |
while (++lit); |
190 |
#endif |
191 |
} |
192 |
} |
193 |
} |
194 |
while (ip < in_end); |
195 |
|
196 |
if (lit) |
197 |
{ |
198 |
if (op + lit + 1 >= out_end) |
199 |
return 0; |
200 |
|
201 |
*op++ = lit - 1; |
202 |
lit = -lit; |
203 |
do |
204 |
*op++ = ip[lit]; |
205 |
while (++lit); |
206 |
} |
207 |
|
208 |
return op - (u8 *) out_data; |
209 |
} |