|
10
|
1 /* deflate.h -- internal compression state
|
|
|
2 * Copyright (C) 1995 Jean-loup Gailly
|
|
|
3 * For conditions of distribution and use, see copyright notice in zlib.h
|
|
|
4 */
|
|
|
5
|
|
|
6 /* WARNING: this file should *not* be used by applications. It is
|
|
|
7 part of the implementation of the compression library and is
|
|
|
8 subject to change. Applications should only use zlib.h.
|
|
|
9 */
|
|
|
10
|
|
|
11 /* $Id: deflate.h,v 1.1.1.1 1997/12/06 04:37:16 darius Exp $ */
|
|
|
12
|
|
|
13 #include "zutil.h"
|
|
|
14
|
|
|
15 /* ===========================================================================
|
|
|
16 * Internal compression state.
|
|
|
17 */
|
|
|
18
|
|
|
19 /* Data type */
|
|
|
20 #define BINARY 0
|
|
|
21 #define ASCII 1
|
|
|
22 #define UNKNOWN 2
|
|
|
23
|
|
|
24 #define LENGTH_CODES 29
|
|
|
25 /* number of length codes, not counting the special END_BLOCK code */
|
|
|
26
|
|
|
27 #define LITERALS 256
|
|
|
28 /* number of literal bytes 0..255 */
|
|
|
29
|
|
|
30 #define L_CODES (LITERALS+1+LENGTH_CODES)
|
|
|
31 /* number of Literal or Length codes, including the END_BLOCK code */
|
|
|
32
|
|
|
33 #define D_CODES 30
|
|
|
34 /* number of distance codes */
|
|
|
35
|
|
|
36 #define BL_CODES 19
|
|
|
37 /* number of codes used to transfer the bit lengths */
|
|
|
38
|
|
|
39 #define HEAP_SIZE (2*L_CODES+1)
|
|
|
40 /* maximum heap size */
|
|
|
41
|
|
|
42 #define MAX_BITS 15
|
|
|
43 /* All codes must not exceed MAX_BITS bits */
|
|
|
44
|
|
|
45 #define INIT_STATE 42
|
|
|
46 #define BUSY_STATE 113
|
|
|
47 #define FINISH_STATE 666
|
|
|
48 /* Stream status */
|
|
|
49
|
|
|
50
|
|
|
51 /* Data structure describing a single value and its code string. */
|
|
|
52 typedef struct ct_data_s {
|
|
|
53 union {
|
|
|
54 ush freq; /* frequency count */
|
|
|
55 ush code; /* bit string */
|
|
|
56 } fc;
|
|
|
57 union {
|
|
|
58 ush dad; /* father node in Huffman tree */
|
|
|
59 ush len; /* length of bit string */
|
|
|
60 } dl;
|
|
|
61 } ct_data;
|
|
|
62
|
|
|
63 #define Freq fc.freq
|
|
|
64 #define Code fc.code
|
|
|
65 #define Dad dl.dad
|
|
|
66 #define Len dl.len
|
|
|
67
|
|
|
68 typedef struct static_tree_desc_s static_tree_desc;
|
|
|
69
|
|
|
70 typedef struct tree_desc_s {
|
|
|
71 ct_data *dyn_tree; /* the dynamic tree */
|
|
|
72 int max_code; /* largest code with non zero frequency */
|
|
|
73 static_tree_desc *stat_desc; /* the corresponding static tree */
|
|
|
74 } tree_desc;
|
|
|
75
|
|
|
76 typedef ush Pos;
|
|
|
77 typedef unsigned IPos;
|
|
|
78 /* A Pos is an index in the character window. We use short instead of int to
|
|
|
79 * save space in the various tables. IPos is used only for parameter passing.
|
|
|
80 */
|
|
|
81
|
|
|
82 typedef struct internal_state {
|
|
|
83 z_stream *strm; /* pointer back to this zlib stream */
|
|
|
84 int status; /* as the name implies */
|
|
|
85 Byte *pending_buf; /* output still pending */
|
|
|
86 Byte *pending_out; /* next pending byte to output to the stream */
|
|
|
87 int pending; /* nb of bytes in the pending buffer */
|
|
|
88 uLong adler; /* adler32 of uncompressed data */
|
|
|
89 int noheader; /* suppress zlib header and adler32 */
|
|
|
90 Byte data_type; /* UNKNOWN, BINARY or ASCII */
|
|
|
91 Byte method; /* STORED (for zip only) or DEFLATED */
|
|
|
92
|
|
|
93 /* used by deflate.c: */
|
|
|
94
|
|
|
95 uInt w_size; /* LZ77 window size (32K by default) */
|
|
|
96 uInt w_bits; /* log2(w_size) (8..16) */
|
|
|
97 uInt w_mask; /* w_size - 1 */
|
|
|
98
|
|
|
99 Byte *window;
|
|
|
100 /* Sliding window. Input bytes are read into the second half of the window,
|
|
|
101 * and move to the first half later to keep a dictionary of at least wSize
|
|
|
102 * bytes. With this organization, matches are limited to a distance of
|
|
|
103 * wSize-MAX_MATCH bytes, but this ensures that IO is always
|
|
|
104 * performed with a length multiple of the block size. Also, it limits
|
|
|
105 * the window size to 64K, which is quite useful on MSDOS.
|
|
|
106 * To do: use the user input buffer as sliding window.
|
|
|
107 */
|
|
|
108
|
|
|
109 ulg window_size;
|
|
|
110 /* Actual size of window: 2*wSize, except when the user input buffer
|
|
|
111 * is directly used as sliding window.
|
|
|
112 */
|
|
|
113
|
|
|
114 Pos *prev;
|
|
|
115 /* Link to older string with same hash index. To limit the size of this
|
|
|
116 * array to 64K, this link is maintained only for the last 32K strings.
|
|
|
117 * An index in this array is thus a window index modulo 32K.
|
|
|
118 */
|
|
|
119
|
|
|
120 Pos *head; /* Heads of the hash chains or NIL. */
|
|
|
121
|
|
|
122 uInt ins_h; /* hash index of string to be inserted */
|
|
|
123 uInt hash_size; /* number of elements in hash table */
|
|
|
124 uInt hash_bits; /* log2(hash_size) */
|
|
|
125 uInt hash_mask; /* hash_size-1 */
|
|
|
126
|
|
|
127 uInt hash_shift;
|
|
|
128 /* Number of bits by which ins_h must be shifted at each input
|
|
|
129 * step. It must be such that after MIN_MATCH steps, the oldest
|
|
|
130 * byte no longer takes part in the hash key, that is:
|
|
|
131 * hash_shift * MIN_MATCH >= hash_bits
|
|
|
132 */
|
|
|
133
|
|
|
134 long block_start;
|
|
|
135 /* Window position at the beginning of the current output block. Gets
|
|
|
136 * negative when the window is moved backwards.
|
|
|
137 */
|
|
|
138
|
|
|
139 uInt match_length; /* length of best match */
|
|
|
140 IPos prev_match; /* previous match */
|
|
|
141 int match_available; /* set if previous match exists */
|
|
|
142 uInt strstart; /* start of string to insert */
|
|
|
143 uInt match_start; /* start of matching string */
|
|
|
144 uInt lookahead; /* number of valid bytes ahead in window */
|
|
|
145
|
|
|
146 uInt prev_length;
|
|
|
147 /* Length of the best match at previous step. Matches not greater than this
|
|
|
148 * are discarded. This is used in the lazy match evaluation.
|
|
|
149 */
|
|
|
150
|
|
|
151 uInt max_chain_length;
|
|
|
152 /* To speed up deflation, hash chains are never searched beyond this
|
|
|
153 * length. A higher limit improves compression ratio but degrades the
|
|
|
154 * speed.
|
|
|
155 */
|
|
|
156
|
|
|
157 uInt max_lazy_match;
|
|
|
158 /* Attempt to find a better match only when the current match is strictly
|
|
|
159 * smaller than this value. This mechanism is used only for compression
|
|
|
160 * levels >= 4.
|
|
|
161 */
|
|
|
162 # define max_insert_length max_lazy_match
|
|
|
163 /* Insert new strings in the hash table only if the match length is not
|
|
|
164 * greater than this length. This saves time but degrades compression.
|
|
|
165 * max_insert_length is used only for compression levels <= 3.
|
|
|
166 */
|
|
|
167
|
|
|
168 int level; /* compression level (1..9) */
|
|
|
169 int strategy; /* favor or force Huffman coding*/
|
|
|
170
|
|
|
171 uInt good_match;
|
|
|
172 /* Use a faster search when the previous match is longer than this */
|
|
|
173
|
|
|
174 int nice_match; /* Stop searching when current match exceeds this */
|
|
|
175
|
|
|
176 /* used by trees.c: */
|
|
|
177
|
|
|
178 ct_data dyn_ltree[HEAP_SIZE]; /* literal and length tree */
|
|
|
179 ct_data dyn_dtree[2*D_CODES+1]; /* distance tree */
|
|
|
180 ct_data bl_tree[2*BL_CODES+1]; /* Huffman tree for the bit lengths */
|
|
|
181
|
|
|
182 tree_desc l_desc; /* descriptor for literal tree */
|
|
|
183 tree_desc d_desc; /* descriptor for distance tree */
|
|
|
184 tree_desc bl_desc; /* descriptor for bit length tree */
|
|
|
185
|
|
|
186 ush bl_count[MAX_BITS+1];
|
|
|
187 /* number of codes at each bit length for an optimal tree */
|
|
|
188
|
|
|
189 int heap[2*L_CODES+1]; /* heap used to build the Huffman trees */
|
|
|
190 int heap_len; /* number of elements in the heap */
|
|
|
191 int heap_max; /* element of largest frequency */
|
|
|
192 /* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
|
|
|
193 * The same heap array is used to build all trees.
|
|
|
194 */
|
|
|
195
|
|
|
196 uch depth[2*L_CODES+1];
|
|
|
197 /* Depth of each subtree used as tie breaker for trees of equal frequency
|
|
|
198 */
|
|
|
199
|
|
|
200 uch *l_buf; /* buffer for literals or lengths */
|
|
|
201
|
|
|
202 uInt lit_bufsize;
|
|
|
203 /* Size of match buffer for literals/lengths. There are 4 reasons for
|
|
|
204 * limiting lit_bufsize to 64K:
|
|
|
205 * - frequencies can be kept in 16 bit counters
|
|
|
206 * - if compression is not successful for the first block, all input
|
|
|
207 * data is still in the window so we can still emit a stored block even
|
|
|
208 * when input comes from standard input. (This can also be done for
|
|
|
209 * all blocks if lit_bufsize is not greater than 32K.)
|
|
|
210 * - if compression is not successful for a file smaller than 64K, we can
|
|
|
211 * even emit a stored file instead of a stored block (saving 5 bytes).
|
|
|
212 * This is applicable only for zip (not gzip or zlib).
|
|
|
213 * - creating new Huffman trees less frequently may not provide fast
|
|
|
214 * adaptation to changes in the input data statistics. (Take for
|
|
|
215 * example a binary file with poorly compressible code followed by
|
|
|
216 * a highly compressible string table.) Smaller buffer sizes give
|
|
|
217 * fast adaptation but have of course the overhead of transmitting
|
|
|
218 * trees more frequently.
|
|
|
219 * - I can't count above 4
|
|
|
220 */
|
|
|
221
|
|
|
222 uInt last_lit; /* running index in l_buf */
|
|
|
223
|
|
|
224 ush *d_buf;
|
|
|
225 /* Buffer for distances. To simplify the code, d_buf and l_buf have
|
|
|
226 * the same number of elements. To use different lengths, an extra flag
|
|
|
227 * array would be necessary.
|
|
|
228 */
|
|
|
229
|
|
|
230 ulg opt_len; /* bit length of current block with optimal trees */
|
|
|
231 ulg static_len; /* bit length of current block with static trees */
|
|
|
232 ulg compressed_len; /* total bit length of compressed file */
|
|
|
233 uInt matches; /* number of string matches in current block */
|
|
|
234
|
|
|
235 #ifdef DEBUG
|
|
|
236 ulg bits_sent; /* bit length of the compressed data */
|
|
|
237 #endif
|
|
|
238
|
|
|
239 ush bi_buf;
|
|
|
240 /* Output buffer. bits are inserted starting at the bottom (least
|
|
|
241 * significant bits).
|
|
|
242 */
|
|
|
243 int bi_valid;
|
|
|
244 /* Number of valid bits in bi_buf. All bits above the last valid bit
|
|
|
245 * are always zero.
|
|
|
246 */
|
|
|
247
|
|
|
248 } deflate_state;
|
|
|
249
|
|
|
250
|
|
|
251 /* Output a byte on the stream.
|
|
|
252 * IN assertion: there is enough room in pending_buf.
|
|
|
253 */
|
|
|
254 #define put_byte(s, c) {s->pending_buf[s->pending++] = (c);}
|
|
|
255
|
|
|
256
|
|
|
257 #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1)
|
|
|
258 /* Minimum amount of lookahead, except at the end of the input file.
|
|
|
259 * See deflate.c for comments about the MIN_MATCH+1.
|
|
|
260 */
|
|
|
261
|
|
|
262 #define MAX_DIST(s) ((s)->w_size-MIN_LOOKAHEAD)
|
|
|
263 /* In order to simplify the code, particularly on 16 bit machines, match
|
|
|
264 * distances are limited to MAX_DIST instead of WSIZE.
|
|
|
265 */
|
|
|
266
|
|
|
267 /* in trees.c */
|
|
|
268 void ct_init __P((deflate_state *s));
|
|
|
269 int ct_tally __P((deflate_state *s, int dist, int lc));
|
|
|
270 ulg ct_flush_block __P((deflate_state *s, char *buf, ulg stored_len, int eof));
|
|
|
271 void ct_stored_block __P((deflate_state *s, char *buf, ulg stored_len,
|
|
|
272 int eof));
|
