1 | /* lzo_mchw.ch -- matching functions using a window |
---|
2 | |
---|
3 | This file is part of the LZO real-time data compression library. |
---|
4 | |
---|
5 | Copyright (C) 1996-2014 Markus Franz Xaver Johannes Oberhumer |
---|
6 | All Rights Reserved. |
---|
7 | |
---|
8 | The LZO library is free software; you can redistribute it and/or |
---|
9 | modify it under the terms of the GNU General Public License as |
---|
10 | published by the Free Software Foundation; either version 2 of |
---|
11 | the License, or (at your option) any later version. |
---|
12 | |
---|
13 | The LZO library is distributed in the hope that it will be useful, |
---|
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
---|
16 | GNU General Public License for more details. |
---|
17 | |
---|
18 | You should have received a copy of the GNU General Public License |
---|
19 | along with the LZO library; see the file COPYING. |
---|
20 | If not, write to the Free Software Foundation, Inc., |
---|
21 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
---|
22 | |
---|
23 | Markus F.X.J. Oberhumer |
---|
24 | <markus@oberhumer.com> |
---|
25 | http://www.oberhumer.com/opensource/lzo/ |
---|
26 | */ |
---|
27 | |
---|
28 | |
---|
29 | /*********************************************************************** |
---|
30 | // |
---|
31 | ************************************************************************/ |
---|
32 | |
---|
33 | typedef struct |
---|
34 | { |
---|
35 | unsigned init; |
---|
36 | |
---|
37 | lzo_uint look; /* bytes in lookahead buffer */ |
---|
38 | |
---|
39 | lzo_uint m_len; |
---|
40 | lzo_uint m_off; |
---|
41 | |
---|
42 | lzo_uint last_m_len; |
---|
43 | lzo_uint last_m_off; |
---|
44 | |
---|
45 | const lzo_bytep bp; |
---|
46 | const lzo_bytep ip; |
---|
47 | const lzo_bytep in; |
---|
48 | const lzo_bytep in_end; |
---|
49 | lzo_bytep out; |
---|
50 | |
---|
51 | lzo_callback_p cb; |
---|
52 | |
---|
53 | lzo_uint textsize; /* text size counter */ |
---|
54 | lzo_uint codesize; /* code size counter */ |
---|
55 | lzo_uint printcount; /* counter for reporting progress every 1K bytes */ |
---|
56 | |
---|
57 | /* some stats */ |
---|
58 | unsigned long lit_bytes; |
---|
59 | unsigned long match_bytes; |
---|
60 | unsigned long rep_bytes; |
---|
61 | unsigned long lazy; |
---|
62 | |
---|
63 | #if defined(LZO1B) |
---|
64 | lzo_uint r1_m_len; |
---|
65 | |
---|
66 | /* some stats */ |
---|
67 | unsigned long r1_r, m3_r, m2_m, m3_m; |
---|
68 | #endif |
---|
69 | |
---|
70 | #if defined(LZO1C) |
---|
71 | lzo_uint r1_m_len; |
---|
72 | lzo_bytep m3; |
---|
73 | |
---|
74 | /* some stats */ |
---|
75 | unsigned long r1_r, m3_r, m2_m, m3_m; |
---|
76 | #endif |
---|
77 | |
---|
78 | #if defined(LZO1F) |
---|
79 | lzo_uint r1_lit; |
---|
80 | lzo_uint r1_m_len; |
---|
81 | |
---|
82 | /* some stats */ |
---|
83 | unsigned long r1_r, m2_m, m3_m; |
---|
84 | #endif |
---|
85 | |
---|
86 | #if defined(LZO1X) || defined(LZO1Y) || defined(LZO1Z) |
---|
87 | lzo_uint r1_lit; |
---|
88 | lzo_uint r1_m_len; |
---|
89 | |
---|
90 | /* some stats */ |
---|
91 | unsigned long m1a_m, m1b_m, m2_m, m3_m, m4_m; |
---|
92 | unsigned long lit1_r, lit2_r, lit3_r; |
---|
93 | #endif |
---|
94 | |
---|
95 | #if defined(LZO2A) |
---|
96 | /* some stats */ |
---|
97 | unsigned long m1, m2, m3, m4; |
---|
98 | #endif |
---|
99 | } |
---|
100 | LZO_COMPRESS_T; |
---|
101 | |
---|
102 | |
---|
103 | #define getbyte(c) ((c).ip < (c).in_end ? *((c).ip)++ : (-1)) |
---|
104 | |
---|
105 | #include "lzo_swd.ch" |
---|
106 | |
---|
107 | |
---|
108 | /*********************************************************************** |
---|
109 | // |
---|
110 | ************************************************************************/ |
---|
111 | |
---|
112 | static int |
---|
113 | init_match ( LZO_COMPRESS_T *c, lzo_swd_p s, |
---|
114 | const lzo_bytep dict, lzo_uint dict_len, |
---|
115 | lzo_uint32_t flags ) |
---|
116 | { |
---|
117 | int r; |
---|
118 | |
---|
119 | assert(!c->init); |
---|
120 | c->init = 1; |
---|
121 | |
---|
122 | s->c = c; |
---|
123 | |
---|
124 | c->last_m_len = c->last_m_off = 0; |
---|
125 | |
---|
126 | c->textsize = c->codesize = c->printcount = 0; |
---|
127 | c->lit_bytes = c->match_bytes = c->rep_bytes = 0; |
---|
128 | c->lazy = 0; |
---|
129 | |
---|
130 | r = swd_init(s,dict,dict_len); |
---|
131 | if (r != LZO_E_OK) |
---|
132 | { |
---|
133 | swd_exit(s); |
---|
134 | return r; |
---|
135 | } |
---|
136 | |
---|
137 | s->use_best_off = (flags & 1) ? 1 : 0; |
---|
138 | return LZO_E_OK; |
---|
139 | } |
---|
140 | |
---|
141 | |
---|
142 | /*********************************************************************** |
---|
143 | // |
---|
144 | ************************************************************************/ |
---|
145 | |
---|
146 | static int |
---|
147 | find_match ( LZO_COMPRESS_T *c, lzo_swd_p s, |
---|
148 | lzo_uint this_len, lzo_uint skip ) |
---|
149 | { |
---|
150 | assert(c->init); |
---|
151 | |
---|
152 | if (skip > 0) |
---|
153 | { |
---|
154 | assert(this_len >= skip); |
---|
155 | swd_accept(s, this_len - skip); |
---|
156 | c->textsize += this_len - skip + 1; |
---|
157 | } |
---|
158 | else |
---|
159 | { |
---|
160 | assert(this_len <= 1); |
---|
161 | c->textsize += this_len - skip; |
---|
162 | } |
---|
163 | |
---|
164 | s->m_len = SWD_THRESHOLD; |
---|
165 | s->m_off = 0; |
---|
166 | #ifdef SWD_BEST_OFF |
---|
167 | if (s->use_best_off) |
---|
168 | lzo_memset(s->best_pos,0,sizeof(s->best_pos)); |
---|
169 | #endif |
---|
170 | swd_findbest(s); |
---|
171 | c->m_len = s->m_len; |
---|
172 | c->m_off = s->m_off; |
---|
173 | |
---|
174 | swd_getbyte(s); |
---|
175 | |
---|
176 | if (s->b_char < 0) |
---|
177 | { |
---|
178 | c->look = 0; |
---|
179 | c->m_len = 0; |
---|
180 | swd_exit(s); |
---|
181 | } |
---|
182 | else |
---|
183 | { |
---|
184 | c->look = s->look + 1; |
---|
185 | } |
---|
186 | c->bp = c->ip - c->look; |
---|
187 | |
---|
188 | #if 0 |
---|
189 | /* brute force match search */ |
---|
190 | if (c->m_len > SWD_THRESHOLD && c->m_len + 1 <= c->look) |
---|
191 | { |
---|
192 | const lzo_bytep ip = c->bp; |
---|
193 | const lzo_bytep m = c->bp - c->m_off; |
---|
194 | const lzo_bytep in = c->in; |
---|
195 | |
---|
196 | if (ip - in > s->swd_n) |
---|
197 | in = ip - s->swd_n; |
---|
198 | for (;;) |
---|
199 | { |
---|
200 | while (*in != *ip) |
---|
201 | in++; |
---|
202 | if (in == ip) |
---|
203 | break; |
---|
204 | if (in != m) |
---|
205 | if (lzo_memcmp(in,ip,c->m_len+1) == 0) |
---|
206 | printf("%p %p %p %5d\n",in,ip,m,c->m_len); |
---|
207 | in++; |
---|
208 | } |
---|
209 | } |
---|
210 | #endif |
---|
211 | |
---|
212 | if (c->cb && c->cb->nprogress && c->textsize > c->printcount) |
---|
213 | { |
---|
214 | (*c->cb->nprogress)(c->cb, c->textsize, c->codesize, 0); |
---|
215 | c->printcount += 1024; |
---|
216 | } |
---|
217 | |
---|
218 | return LZO_E_OK; |
---|
219 | } |
---|
220 | |
---|
221 | |
---|
222 | /* |
---|
223 | vi:ts=4:et |
---|
224 | */ |
---|
225 | |
---|