1 |
cycrow |
1 |
#include <windows.h>
|
|
|
2 |
#include <stdio.h>
|
|
|
3 |
#include <tchar.h>
|
|
|
4 |
#include "zip.h"
|
|
|
5 |
|
|
|
6 |
|
|
|
7 |
// THIS FILE is almost entirely based upon code by info-zip.
|
|
|
8 |
// It has been modified by Lucian Wischik. The modifications
|
|
|
9 |
// were a complete rewrite of the bit of code that generates the
|
|
|
10 |
// layout of the zipfile, and support for zipping to/from memory
|
|
|
11 |
// or handles or pipes or pagefile or diskfiles, encryption, unicode.
|
|
|
12 |
// The original code may be found at http://www.info-zip.org
|
|
|
13 |
// The original copyright text follows.
|
|
|
14 |
//
|
|
|
15 |
//
|
|
|
16 |
//
|
|
|
17 |
// This is version 1999-Oct-05 of the Info-ZIP copyright and license.
|
|
|
18 |
// The definitive version of this document should be available at
|
|
|
19 |
// ftp://ftp.cdrom.com/pub/infozip/license.html indefinitely.
|
|
|
20 |
//
|
|
|
21 |
// Copyright (c) 1990-1999 Info-ZIP. All rights reserved.
|
|
|
22 |
//
|
|
|
23 |
// For the purposes of this copyright and license, "Info-ZIP" is defined as
|
|
|
24 |
// the following set of individuals:
|
|
|
25 |
//
|
|
|
26 |
// Mark Adler, John Bush, Karl Davis, Harald Denker, Jean-Michel Dubois,
|
|
|
27 |
// Jean-loup Gailly, Hunter Goatley, Ian Gorman, Chris Herborth, Dirk Haase,
|
|
|
28 |
// Greg Hartwig, Robert Heath, Jonathan Hudson, Paul Kienitz, David Kirschbaum,
|
|
|
29 |
// Johnny Lee, Onno van der Linden, Igor Mandrichenko, Steve P. Miller,
|
|
|
30 |
// Sergio Monesi, Keith Owens, George Petrov, Greg Roelofs, Kai Uwe Rommel,
|
|
|
31 |
// Steve Salisbury, Dave Smith, Christian Spieler, Antoine Verheijen,
|
|
|
32 |
// Paul von Behren, Rich Wales, Mike White
|
|
|
33 |
//
|
|
|
34 |
// This software is provided "as is," without warranty of any kind, express
|
|
|
35 |
// or implied. In no event shall Info-ZIP or its contributors be held liable
|
|
|
36 |
// for any direct, indirect, incidental, special or consequential damages
|
|
|
37 |
// arising out of the use of or inability to use this software.
|
|
|
38 |
//
|
|
|
39 |
// Permission is granted to anyone to use this software for any purpose,
|
|
|
40 |
// including commercial applications, and to alter it and redistribute it
|
|
|
41 |
// freely, subject to the following restrictions:
|
|
|
42 |
//
|
|
|
43 |
// 1. Redistributions of source code must retain the above copyright notice,
|
|
|
44 |
// definition, disclaimer, and this list of conditions.
|
|
|
45 |
//
|
|
|
46 |
// 2. Redistributions in binary form must reproduce the above copyright
|
|
|
47 |
// notice, definition, disclaimer, and this list of conditions in
|
|
|
48 |
// documentation and/or other materials provided with the distribution.
|
|
|
49 |
//
|
|
|
50 |
// 3. Altered versions--including, but not limited to, ports to new operating
|
|
|
51 |
// systems, existing ports with new graphical interfaces, and dynamic,
|
|
|
52 |
// shared, or static library versions--must be plainly marked as such
|
|
|
53 |
// and must not be misrepresented as being the original source. Such
|
|
|
54 |
// altered versions also must not be misrepresented as being Info-ZIP
|
|
|
55 |
// releases--including, but not limited to, labeling of the altered
|
|
|
56 |
// versions with the names "Info-ZIP" (or any variation thereof, including,
|
|
|
57 |
// but not limited to, different capitalizations), "Pocket UnZip," "WiZ"
|
|
|
58 |
// or "MacZip" without the explicit permission of Info-ZIP. Such altered
|
|
|
59 |
// versions are further prohibited from misrepresentative use of the
|
|
|
60 |
// Zip-Bugs or Info-ZIP e-mail addresses or of the Info-ZIP URL(s).
|
|
|
61 |
//
|
|
|
62 |
// 4. Info-ZIP retains the right to use the names "Info-ZIP," "Zip," "UnZip,"
|
|
|
63 |
// "WiZ," "Pocket UnZip," "Pocket Zip," and "MacZip" for its own source and
|
|
|
64 |
// binary releases.
|
|
|
65 |
//
|
|
|
66 |
|
|
|
67 |
|
|
|
68 |
typedef unsigned char uch; // unsigned 8-bit value
|
|
|
69 |
typedef unsigned short ush; // unsigned 16-bit value
|
|
|
70 |
typedef unsigned long ulg; // unsigned 32-bit value
|
|
|
71 |
typedef size_t extent; // file size
|
|
|
72 |
typedef unsigned Pos; // must be at least 32 bits
|
|
|
73 |
typedef unsigned IPos; // A Pos is an index in the character window. Pos is used only for parameter passing
|
|
|
74 |
|
|
|
75 |
#ifndef EOF
|
|
|
76 |
#define EOF (-1)
|
|
|
77 |
#endif
|
|
|
78 |
|
|
|
79 |
|
|
|
80 |
// Error return values. The values 0..4 and 12..18 follow the conventions
|
|
|
81 |
// of PKZIP. The values 4..10 are all assigned to "insufficient memory"
|
|
|
82 |
// by PKZIP, so the codes 5..10 are used here for other purposes.
|
|
|
83 |
#define ZE_MISS -1 // used by procname(), zipbare()
|
|
|
84 |
#define ZE_OK 0 // success
|
|
|
85 |
#define ZE_EOF 2 // unexpected end of zip file
|
|
|
86 |
#define ZE_FORM 3 // zip file structure error
|
|
|
87 |
#define ZE_MEM 4 // out of memory
|
|
|
88 |
#define ZE_LOGIC 5 // internal logic error
|
|
|
89 |
#define ZE_BIG 6 // entry too large to split
|
|
|
90 |
#define ZE_NOTE 7 // invalid comment format
|
|
|
91 |
#define ZE_TEST 8 // zip test (-T) failed or out of memory
|
|
|
92 |
#define ZE_ABORT 9 // user interrupt or termination
|
|
|
93 |
#define ZE_TEMP 10 // error using a temp file
|
|
|
94 |
#define ZE_READ 11 // read or seek error
|
|
|
95 |
#define ZE_NONE 12 // nothing to do
|
|
|
96 |
#define ZE_NAME 13 // missing or empty zip file
|
|
|
97 |
#define ZE_WRITE 14 // error writing to a file
|
|
|
98 |
#define ZE_CREAT 15 // couldn't open to write
|
|
|
99 |
#define ZE_PARMS 16 // bad command line
|
|
|
100 |
#define ZE_OPEN 18 // could not open a specified file to read
|
|
|
101 |
#define ZE_MAXERR 18 // the highest error number
|
|
|
102 |
|
|
|
103 |
|
|
|
104 |
// internal file attribute
|
|
|
105 |
#define UNKNOWN (-1)
|
|
|
106 |
#define BINARY 0
|
|
|
107 |
#define ASCII 1
|
|
|
108 |
|
|
|
109 |
#define BEST -1 // Use best method (deflation or store)
|
|
|
110 |
#define STORE 0 // Store method
|
|
|
111 |
#define DEFLATE 8 // Deflation method
|
|
|
112 |
|
|
|
113 |
#define CRCVAL_INITIAL 0L
|
|
|
114 |
|
|
|
115 |
// MSDOS file or directory attributes
|
|
|
116 |
#define MSDOS_HIDDEN_ATTR 0x02
|
|
|
117 |
#define MSDOS_DIR_ATTR 0x10
|
|
|
118 |
|
|
|
119 |
// Lengths of headers after signatures in bytes
|
|
|
120 |
#define LOCHEAD 26
|
|
|
121 |
#define CENHEAD 42
|
|
|
122 |
#define ENDHEAD 18
|
|
|
123 |
|
|
|
124 |
// Definitions for extra field handling:
|
|
|
125 |
#define EB_HEADSIZE 4 /* length of a extra field block header */
|
|
|
126 |
#define EB_LEN 2 /* offset of data length field in header */
|
|
|
127 |
#define EB_UT_MINLEN 1 /* minimal UT field contains Flags byte */
|
|
|
128 |
#define EB_UT_FLAGS 0 /* byte offset of Flags field */
|
|
|
129 |
#define EB_UT_TIME1 1 /* byte offset of 1st time value */
|
|
|
130 |
#define EB_UT_FL_MTIME (1 << 0) /* mtime present */
|
|
|
131 |
#define EB_UT_FL_ATIME (1 << 1) /* atime present */
|
|
|
132 |
#define EB_UT_FL_CTIME (1 << 2) /* ctime present */
|
|
|
133 |
#define EB_UT_LEN(n) (EB_UT_MINLEN + 4 * (n))
|
|
|
134 |
#define EB_L_UT_SIZE (EB_HEADSIZE + EB_UT_LEN(3))
|
|
|
135 |
#define EB_C_UT_SIZE (EB_HEADSIZE + EB_UT_LEN(1))
|
|
|
136 |
|
|
|
137 |
|
|
|
138 |
// Macros for writing machine integers to little-endian format
|
|
|
139 |
#define PUTSH(a,f) {char _putsh_c=(char)((a)&0xff); wfunc(param,&_putsh_c,1); _putsh_c=(char)((a)>>8); wfunc(param,&_putsh_c,1);}
|
|
|
140 |
#define PUTLG(a,f) {PUTSH((a) & 0xffff,(f)) PUTSH((a) >> 16,(f))}
|
|
|
141 |
|
|
|
142 |
|
|
|
143 |
// -- Structure of a ZIP file --
|
|
|
144 |
// Signatures for zip file information headers
|
|
|
145 |
#define LOCSIG 0x04034b50L
|
|
|
146 |
#define CENSIG 0x02014b50L
|
|
|
147 |
#define ENDSIG 0x06054b50L
|
|
|
148 |
#define EXTLOCSIG 0x08074b50L
|
|
|
149 |
|
|
|
150 |
|
|
|
151 |
#define MIN_MATCH 3
|
|
|
152 |
#define MAX_MATCH 258
|
|
|
153 |
// The minimum and maximum match lengths
|
|
|
154 |
|
|
|
155 |
|
|
|
156 |
#define WSIZE (0x8000)
|
|
|
157 |
// Maximum window size = 32K. If you are really short of memory, compile
|
|
|
158 |
// with a smaller WSIZE but this reduces the compression ratio for files
|
|
|
159 |
// of size > WSIZE. WSIZE must be a power of two in the current implementation.
|
|
|
160 |
//
|
|
|
161 |
|
|
|
162 |
#define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1)
|
|
|
163 |
// Minimum amount of lookahead, except at the end of the input file.
|
|
|
164 |
// See deflate.c for comments about the MIN_MATCH+1.
|
|
|
165 |
//
|
|
|
166 |
|
|
|
167 |
#define MAX_DIST (WSIZE-MIN_LOOKAHEAD)
|
|
|
168 |
// In order to simplify the code, particularly on 16 bit machines, match
|
|
|
169 |
// distances are limited to MAX_DIST instead of WSIZE.
|
|
|
170 |
//
|
|
|
171 |
|
|
|
172 |
|
|
|
173 |
#define ZIP_HANDLE 1
|
|
|
174 |
#define ZIP_FILENAME 2
|
|
|
175 |
#define ZIP_MEMORY 3
|
|
|
176 |
#define ZIP_FOLDER 4
|
|
|
177 |
|
|
|
178 |
|
|
|
179 |
|
|
|
180 |
// ===========================================================================
|
|
|
181 |
// Constants
|
|
|
182 |
//
|
|
|
183 |
|
|
|
184 |
#define MAX_BITS 15
|
|
|
185 |
// All codes must not exceed MAX_BITS bits
|
|
|
186 |
|
|
|
187 |
#define MAX_BL_BITS 7
|
|
|
188 |
// Bit length codes must not exceed MAX_BL_BITS bits
|
|
|
189 |
|
|
|
190 |
#define LENGTH_CODES 29
|
|
|
191 |
// number of length codes, not counting the special END_BLOCK code
|
|
|
192 |
|
|
|
193 |
#define LITERALS 256
|
|
|
194 |
// number of literal bytes 0..255
|
|
|
195 |
|
|
|
196 |
#define END_BLOCK 256
|
|
|
197 |
// end of block literal code
|
|
|
198 |
|
|
|
199 |
#define L_CODES (LITERALS+1+LENGTH_CODES)
|
|
|
200 |
// number of Literal or Length codes, including the END_BLOCK code
|
|
|
201 |
|
|
|
202 |
#define D_CODES 30
|
|
|
203 |
// number of distance codes
|
|
|
204 |
|
|
|
205 |
#define BL_CODES 19
|
|
|
206 |
// number of codes used to transfer the bit lengths
|
|
|
207 |
|
|
|
208 |
|
|
|
209 |
#define STORED_BLOCK 0
|
|
|
210 |
#define STATIC_TREES 1
|
|
|
211 |
#define DYN_TREES 2
|
|
|
212 |
// The three kinds of block type
|
|
|
213 |
|
|
|
214 |
#define LIT_BUFSIZE 0x8000
|
|
|
215 |
#define DIST_BUFSIZE LIT_BUFSIZE
|
|
|
216 |
// Sizes of match buffers for literals/lengths and distances. There are
|
|
|
217 |
// 4 reasons for limiting LIT_BUFSIZE to 64K:
|
|
|
218 |
// - frequencies can be kept in 16 bit counters
|
|
|
219 |
// - if compression is not successful for the first block, all input data is
|
|
|
220 |
// still in the window so we can still emit a stored block even when input
|
|
|
221 |
// comes from standard input. (This can also be done for all blocks if
|
|
|
222 |
// LIT_BUFSIZE is not greater than 32K.)
|
|
|
223 |
// - if compression is not successful for a file smaller than 64K, we can
|
|
|
224 |
// even emit a stored file instead of a stored block (saving 5 bytes).
|
|
|
225 |
// - creating new Huffman trees less frequently may not provide fast
|
|
|
226 |
// adaptation to changes in the input data statistics. (Take for
|
|
|
227 |
// example a binary file with poorly compressible code followed by
|
|
|
228 |
// a highly compressible string table.) Smaller buffer sizes give
|
|
|
229 |
// fast adaptation but have of course the overhead of transmitting trees
|
|
|
230 |
// more frequently.
|
|
|
231 |
// - I can't count above 4
|
|
|
232 |
// The current code is general and allows DIST_BUFSIZE < LIT_BUFSIZE (to save
|
|
|
233 |
// memory at the expense of compression). Some optimizations would be possible
|
|
|
234 |
// if we rely on DIST_BUFSIZE == LIT_BUFSIZE.
|
|
|
235 |
//
|
|
|
236 |
|
|
|
237 |
#define REP_3_6 16
|
|
|
238 |
// repeat previous bit length 3-6 times (2 bits of repeat count)
|
|
|
239 |
|
|
|
240 |
#define REPZ_3_10 17
|
|
|
241 |
// repeat a zero length 3-10 times (3 bits of repeat count)
|
|
|
242 |
|
|
|
243 |
#define REPZ_11_138 18
|
|
|
244 |
// repeat a zero length 11-138 times (7 bits of repeat count)
|
|
|
245 |
|
|
|
246 |
#define HEAP_SIZE (2*L_CODES+1)
|
|
|
247 |
// maximum heap size
|
|
|
248 |
|
|
|
249 |
|
|
|
250 |
// ===========================================================================
|
|
|
251 |
// Local data used by the "bit string" routines.
|
|
|
252 |
//
|
|
|
253 |
|
|
|
254 |
#define Buf_size (8 * 2*sizeof(char))
|
|
|
255 |
// Number of bits used within bi_buf. (bi_buf may be implemented on
|
|
|
256 |
// more than 16 bits on some systems.)
|
|
|
257 |
|
|
|
258 |
// Output a 16 bit value to the bit stream, lower (oldest) byte first
|
|
|
259 |
#define PUTSHORT(state,w) \
|
|
|
260 |
{ if (state.bs.out_offset >= state.bs.out_size-1) \
|
|
|
261 |
state.flush_outbuf(state.param,state.bs.out_buf, &state.bs.out_offset); \
|
|
|
262 |
state.bs.out_buf[state.bs.out_offset++] = (char) ((w) & 0xff); \
|
|
|
263 |
state.bs.out_buf[state.bs.out_offset++] = (char) ((ush)(w) >> 8); \
|
|
|
264 |
}
|
|
|
265 |
|
|
|
266 |
#define PUTBYTE(state,b) \
|
|
|
267 |
{ if (state.bs.out_offset >= state.bs.out_size) \
|
|
|
268 |
state.flush_outbuf(state.param,state.bs.out_buf, &state.bs.out_offset); \
|
|
|
269 |
state.bs.out_buf[state.bs.out_offset++] = (char) (b); \
|
|
|
270 |
}
|
|
|
271 |
|
|
|
272 |
// DEFLATE.CPP HEADER
|
|
|
273 |
|
|
|
274 |
#define HASH_BITS 15
|
|
|
275 |
// For portability to 16 bit machines, do not use values above 15.
|
|
|
276 |
|
|
|
277 |
#define HASH_SIZE (unsigned)(1<<HASH_BITS)
|
|
|
278 |
#define HASH_MASK (HASH_SIZE-1)
|
|
|
279 |
#define WMASK (WSIZE-1)
|
|
|
280 |
// HASH_SIZE and WSIZE must be powers of two
|
|
|
281 |
|
|
|
282 |
#define NIL 0
|
|
|
283 |
// Tail of hash chains
|
|
|
284 |
|
|
|
285 |
#define FAST 4
|
|
|
286 |
#define SLOW 2
|
|
|
287 |
// speed options for the general purpose bit flag
|
|
|
288 |
|
|
|
289 |
#define TOO_FAR 4096
|
|
|
290 |
// Matches of length 3 are discarded if their distance exceeds TOO_FAR
|
|
|
291 |
|
|
|
292 |
|
|
|
293 |
|
|
|
294 |
#define EQUAL 0
|
|
|
295 |
// result of memcmp for equal strings
|
|
|
296 |
|
|
|
297 |
|
|
|
298 |
// ===========================================================================
|
|
|
299 |
// Local data used by the "longest match" routines.
|
|
|
300 |
|
|
|
301 |
#define H_SHIFT ((HASH_BITS+MIN_MATCH-1)/MIN_MATCH)
|
|
|
302 |
// Number of bits by which ins_h and del_h must be shifted at each
|
|
|
303 |
// input step. It must be such that after MIN_MATCH steps, the oldest
|
|
|
304 |
// byte no longer takes part in the hash key, that is:
|
|
|
305 |
// H_SHIFT * MIN_MATCH >= HASH_BITS
|
|
|
306 |
|
|
|
307 |
#define max_insert_length max_lazy_match
|
|
|
308 |
// Insert new strings in the hash table only if the match length
|
|
|
309 |
// is not greater than this length. This saves time but degrades compression.
|
|
|
310 |
// max_insert_length is used only for compression levels <= 3.
|
|
|
311 |
|
|
|
312 |
|
|
|
313 |
|
|
|
314 |
const int extra_lbits[LENGTH_CODES] // extra bits for each length code
|
|
|
315 |
= {0,0,0,0,0,0,0,0,1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4,5,5,5,5,0};
|
|
|
316 |
|
|
|
317 |
const int extra_dbits[D_CODES] // extra bits for each distance code
|
|
|
318 |
= {0,0,0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,12,12,13,13};
|
|
|
319 |
|
|
|
320 |
const int extra_blbits[BL_CODES]// extra bits for each bit length code
|
|
|
321 |
= {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,3,7};
|
|
|
322 |
|
|
|
323 |
const uch bl_order[BL_CODES] = {16,17,18,0,8,7,9,6,10,5,11,4,12,3,13,2,14,1,15};
|
|
|
324 |
// The lengths of the bit length codes are sent in order of decreasing
|
|
|
325 |
// probability, to avoid transmitting the lengths for unused bit length codes.
|
|
|
326 |
|
|
|
327 |
|
|
|
328 |
typedef struct config {
|
|
|
329 |
ush good_length; // reduce lazy search above this match length
|
|
|
330 |
ush max_lazy; // do not perform lazy search above this match length
|
|
|
331 |
ush nice_length; // quit search above this match length
|
|
|
332 |
ush max_chain;
|
|
|
333 |
} config;
|
|
|
334 |
|
|
|
335 |
// Values for max_lazy_match, good_match, nice_match and max_chain_length,
|
|
|
336 |
// depending on the desired pack level (0..9). The values given below have
|
|
|
337 |
// been tuned to exclude worst case performance for pathological files.
|
|
|
338 |
// Better values may be found for specific files.
|
|
|
339 |
//
|
|
|
340 |
|
|
|
341 |
const config configuration_table[10] = {
|
|
|
342 |
// good lazy nice chain
|
|
|
343 |
{0, 0, 0, 0}, // 0 store only
|
|
|
344 |
{4, 4, 8, 4}, // 1 maximum speed, no lazy matches
|
|
|
345 |
{4, 5, 16, 8}, // 2
|
|
|
346 |
{4, 6, 32, 32}, // 3
|
|
|
347 |
{4, 4, 16, 16}, // 4 lazy matches */
|
|
|
348 |
{8, 16, 32, 32}, // 5
|
|
|
349 |
{8, 16, 128, 128}, // 6
|
|
|
350 |
{8, 32, 128, 256}, // 7
|
|
|
351 |
{32, 128, 258, 1024}, // 8
|
|
|
352 |
{32, 258, 258, 4096}};// 9 maximum compression */
|
|
|
353 |
|
|
|
354 |
// Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4
|
|
|
355 |
// For deflate_fast() (levels <= 3) good is ignored and lazy has a different meaning.
|
|
|
356 |
|
|
|
357 |
|
|
|
358 |
|
|
|
359 |
|
|
|
360 |
|
|
|
361 |
|
|
|
362 |
|
|
|
363 |
// Data structure describing a single value and its code string.
|
|
|
364 |
typedef struct ct_data {
|
|
|
365 |
union {
|
|
|
366 |
ush freq; // frequency count
|
|
|
367 |
ush code; // bit string
|
|
|
368 |
} fc;
|
|
|
369 |
union {
|
|
|
370 |
ush dad; // father node in Huffman tree
|
|
|
371 |
ush len; // length of bit string
|
|
|
372 |
} dl;
|
|
|
373 |
} ct_data;
|
|
|
374 |
|
|
|
375 |
typedef struct tree_desc {
|
|
|
376 |
ct_data *dyn_tree; // the dynamic tree
|
|
|
377 |
ct_data *static_tree; // corresponding static tree or NULL
|
|
|
378 |
const int *extra_bits; // extra bits for each code or NULL
|
|
|
379 |
int extra_base; // base index for extra_bits
|
|
|
380 |
int elems; // max number of elements in the tree
|
|
|
381 |
int max_length; // max bit length for the codes
|
|
|
382 |
int max_code; // largest code with non zero frequency
|
|
|
383 |
} tree_desc;
|
|
|
384 |
|
|
|
385 |
|
|
|
386 |
|
|
|
387 |
|
|
|
388 |
class TTreeState
|
|
|
389 |
{ public:
|
|
|
390 |
TTreeState();
|
|
|
391 |
|
|
|
392 |
ct_data dyn_ltree[HEAP_SIZE]; // literal and length tree
|
|
|
393 |
ct_data dyn_dtree[2*D_CODES+1]; // distance tree
|
|
|
394 |
ct_data static_ltree[L_CODES+2]; // the static literal tree...
|
|
|
395 |
// ... Since the bit lengths are imposed, there is no need for the L_CODES
|
|
|
396 |
// extra codes used during heap construction. However the codes 286 and 287
|
|
|
397 |
// are needed to build a canonical tree (see ct_init below).
|
|
|
398 |
ct_data static_dtree[D_CODES]; // the static distance tree...
|
|
|
399 |
// ... (Actually a trivial tree since all codes use 5 bits.)
|
|
|
400 |
ct_data bl_tree[2*BL_CODES+1]; // Huffman tree for the bit lengths
|
|
|
401 |
|
|
|
402 |
tree_desc l_desc;
|
|
|
403 |
tree_desc d_desc;
|
|
|
404 |
tree_desc bl_desc;
|
|
|
405 |
|
|
|
406 |
ush bl_count[MAX_BITS+1]; // number of codes at each bit length for an optimal tree
|
|
|
407 |
|
|
|
408 |
int heap[2*L_CODES+1]; // heap used to build the Huffman trees
|
|
|
409 |
int heap_len; // number of elements in the heap
|
|
|
410 |
int heap_max; // element of largest frequency
|
|
|
411 |
// The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
|
|
|
412 |
// The same heap array is used to build all trees.
|
|
|
413 |
|
|
|
414 |
uch depth[2*L_CODES+1];
|
|
|
415 |
// Depth of each subtree used as tie breaker for trees of equal frequency
|
|
|
416 |
|
|
|
417 |
uch length_code[MAX_MATCH-MIN_MATCH+1];
|
|
|
418 |
// length code for each normalized match length (0 == MIN_MATCH)
|
|
|
419 |
|
|
|
420 |
uch dist_code[512];
|
|
|
421 |
// distance codes. The first 256 values correspond to the distances
|
|
|
422 |
// 3 .. 258, the last 256 values correspond to the top 8 bits of
|
|
|
423 |
// the 15 bit distances.
|
|
|
424 |
|
|
|
425 |
int base_length[LENGTH_CODES];
|
|
|
426 |
// First normalized length for each code (0 = MIN_MATCH)
|
|
|
427 |
|
|
|
428 |
int base_dist[D_CODES];
|
|
|
429 |
// First normalized distance for each code (0 = distance of 1)
|
|
|
430 |
|
|
|
431 |
uch far l_buf[LIT_BUFSIZE]; // buffer for literals/lengths
|
|
|
432 |
ush far d_buf[DIST_BUFSIZE]; // buffer for distances
|
|
|
433 |
|
|
|
434 |
uch flag_buf[(LIT_BUFSIZE/8)];
|
|
|
435 |
// flag_buf is a bit array distinguishing literals from lengths in
|
|
|
436 |
// l_buf, and thus indicating the presence or absence of a distance.
|
|
|
437 |
|
|
|
438 |
unsigned last_lit; // running index in l_buf
|
|
|
439 |
unsigned last_dist; // running index in d_buf
|
|
|
440 |
unsigned last_flags; // running index in flag_buf
|
|
|
441 |
uch flags; // current flags not yet saved in flag_buf
|
|
|
442 |
uch flag_bit; // current bit used in flags
|
|
|
443 |
// bits are filled in flags starting at bit 0 (least significant).
|
|
|
444 |
// Note: these flags are overkill in the current code since we don't
|
|
|
445 |
// take advantage of DIST_BUFSIZE == LIT_BUFSIZE.
|
|
|
446 |
|
|
|
447 |
ulg opt_len; // bit length of current block with optimal trees
|
|
|
448 |
ulg static_len; // bit length of current block with static trees
|
|
|
449 |
|
|
|
450 |
ulg cmpr_bytelen; // total byte length of compressed file
|
|
|
451 |
ulg cmpr_len_bits; // number of bits past 'cmpr_bytelen'
|
|
|
452 |
|
|
|
453 |
ulg input_len; // total byte length of input file
|
|
|
454 |
// input_len is for debugging only since we can get it by other means.
|
|
|
455 |
|
|
|
456 |
ush *file_type; // pointer to UNKNOWN, BINARY or ASCII
|
|
|
457 |
// int *file_method; // pointer to DEFLATE or STORE
|
|
|
458 |
};
|
|
|
459 |
|
|
|
460 |
TTreeState::TTreeState()
|
|
|
461 |
{ tree_desc a = {dyn_ltree, static_ltree, extra_lbits, LITERALS+1, L_CODES, MAX_BITS, 0}; l_desc = a;
|
|
|
462 |
tree_desc b = {dyn_dtree, static_dtree, extra_dbits, 0, D_CODES, MAX_BITS, 0}; d_desc = b;
|
|
|
463 |
tree_desc c = {bl_tree, NULL, extra_blbits, 0, BL_CODES, MAX_BL_BITS, 0}; bl_desc = c;
|
|
|
464 |
last_lit=0;
|
|
|
465 |
last_dist=0;
|
|
|
466 |
last_flags=0;
|
|
|
467 |
}
|
|
|
468 |
|
|
|
469 |
|
|
|
470 |
|
|
|
471 |
class TBitState
|
|
|
472 |
{ public:
|
|
|
473 |
|
|
|
474 |
int flush_flg;
|
|
|
475 |
//
|
|
|
476 |
unsigned bi_buf;
|
|
|
477 |
// Output buffer. bits are inserted starting at the bottom (least significant
|
|
|
478 |
// bits). The width of bi_buf must be at least 16 bits.
|
|
|
479 |
int bi_valid;
|
|
|
480 |
// Number of valid bits in bi_buf. All bits above the last valid bit
|
|
|
481 |
// are always zero.
|
|
|
482 |
char *out_buf;
|
|
|
483 |
// Current output buffer.
|
|
|
484 |
unsigned out_offset;
|
|
|
485 |
// Current offset in output buffer.
|
|
|
486 |
// On 16 bit machines, the buffer is limited to 64K.
|
|
|
487 |
unsigned out_size;
|
|
|
488 |
// Size of current output buffer
|
|
|
489 |
ulg bits_sent; // bit length of the compressed data only needed for debugging???
|
|
|
490 |
};
|
|
|
491 |
|
|
|
492 |
|
|
|
493 |
|
|
|
494 |
|
|
|
495 |
|
|
|
496 |
|
|
|
497 |
|
|
|
498 |
class TDeflateState
|
|
|
499 |
{ public:
|
|
|
500 |
TDeflateState() {window_size=0;}
|
|
|
501 |
|
|
|
502 |
uch window[2L*WSIZE];
|
|
|
503 |
// Sliding window. Input bytes are read into the second half of the window,
|
|
|
504 |
// and move to the first half later to keep a dictionary of at least WSIZE
|
|
|
505 |
// bytes. With this organization, matches are limited to a distance of
|
|
|
506 |
// WSIZE-MAX_MATCH bytes, but this ensures that IO is always
|
|
|
507 |
// performed with a length multiple of the block size. Also, it limits
|
|
|
508 |
// the window size to 64K, which is quite useful on MSDOS.
|
|
|
509 |
// To do: limit the window size to WSIZE+CBSZ if SMALL_MEM (the code would
|
|
|
510 |
// be less efficient since the data would have to be copied WSIZE/CBSZ times)
|
|
|
511 |
Pos prev[WSIZE];
|
|
|
512 |
// Link to older string with same hash index. To limit the size of this
|
|
|
513 |
// array to 64K, this link is maintained only for the last 32K strings.
|
|
|
514 |
// An index in this array is thus a window index modulo 32K.
|
|
|
515 |
Pos head[HASH_SIZE];
|
|
|
516 |
// Heads of the hash chains or NIL. If your compiler thinks that
|
|
|
517 |
// HASH_SIZE is a dynamic value, recompile with -DDYN_ALLOC.
|
|
|
518 |
|
|
|
519 |
ulg window_size;
|
|
|
520 |
// window size, 2*WSIZE except for MMAP or BIG_MEM, where it is the
|
|
|
521 |
// input file length plus MIN_LOOKAHEAD.
|
|
|
522 |
|
|
|
523 |
long block_start;
|
|
|
524 |
// window position at the beginning of the current output block. Gets
|
|
|
525 |
// negative when the window is moved backwards.
|
|
|
526 |
|
|
|
527 |
int sliding;
|
|
|
528 |
// Set to false when the input file is already in memory
|
|
|
529 |
|
|
|
530 |
unsigned ins_h; // hash index of string to be inserted
|
|
|
531 |
|
|
|
532 |
unsigned int prev_length;
|
|
|
533 |
// Length of the best match at previous step. Matches not greater than this
|
|
|
534 |
// are discarded. This is used in the lazy match evaluation.
|
|
|
535 |
|
|
|
536 |
unsigned strstart; // start of string to insert
|
|
|
537 |
unsigned match_start; // start of matching string
|
|
|
538 |
int eofile; // flag set at end of input file
|
|
|
539 |
unsigned lookahead; // number of valid bytes ahead in window
|
|
|
540 |
|
|
|
541 |
unsigned max_chain_length;
|
|
|
542 |
// To speed up deflation, hash chains are never searched beyond this length.
|
|
|
543 |
// A higher limit improves compression ratio but degrades the speed.
|
|
|
544 |
|
|
|
545 |
unsigned int max_lazy_match;
|
|
|
546 |
// Attempt to find a better match only when the current match is strictly
|
|
|
547 |
// smaller than this value. This mechanism is used only for compression
|
|
|
548 |
// levels >= 4.
|
|
|
549 |
|
|
|
550 |
unsigned good_match;
|
|
|
551 |
// Use a faster search when the previous match is longer than this
|
|
|
552 |
|
|
|
553 |
int nice_match; // Stop searching when current match exceeds this
|
|
|
554 |
};
|
|
|
555 |
|
|
|
556 |
typedef __int64 lutime_t; // define it ourselves since we don't include time.h
|
|
|
557 |
|
|
|
558 |
typedef struct iztimes {
|
|
|
559 |
lutime_t atime,mtime,ctime;
|
|
|
560 |
} iztimes; // access, modify, create times
|
|
|
561 |
|
|
|
562 |
typedef struct zlist {
|
|
|
563 |
ush vem, ver, flg, how; // See central header in zipfile.c for what vem..off are
|
|
|
564 |
ulg tim, crc, siz, len;
|
|
|
565 |
extent nam, ext, cext, com; // offset of ext must be >= LOCHEAD
|
|
|
566 |
ush dsk, att, lflg; // offset of lflg must be >= LOCHEAD
|
|
|
567 |
ulg atx, off;
|
|
|
568 |
char name[MAX_PATH]; // File name in zip file
|
|
|
569 |
char *extra; // Extra field (set only if ext != 0)
|
|
|
570 |
char *cextra; // Extra in central (set only if cext != 0)
|
|
|
571 |
char *comment; // Comment (set only if com != 0)
|
|
|
572 |
char iname[MAX_PATH]; // Internal file name after cleanup
|
|
|
573 |
char zname[MAX_PATH]; // External version of internal name
|
|
|
574 |
int mark; // Marker for files to operate on
|
|
|
575 |
int trash; // Marker for files to delete
|
|
|
576 |
int dosflag; // Set to force MSDOS file attributes
|
|
|
577 |
struct zlist far *nxt; // Pointer to next header in list
|
|
|
578 |
} TZipFileInfo;
|
|
|
579 |
|
|
|
580 |
|
|
|
581 |
struct TState;
|
|
|
582 |
typedef unsigned (*READFUNC)(TState &state, char *buf,unsigned size);
|
|
|
583 |
typedef unsigned (*FLUSHFUNC)(void *param, const char *buf, unsigned *size);
|
|
|
584 |
typedef unsigned (*WRITEFUNC)(void *param, const char *buf, unsigned size);
|
|
|
585 |
struct TState
|
|
|
586 |
{ void *param;
|
|
|
587 |
int level; bool seekable;
|
|
|
588 |
READFUNC readfunc; FLUSHFUNC flush_outbuf;
|
|
|
589 |
TTreeState ts; TBitState bs; TDeflateState ds;
|
|
|
590 |
const char *err;
|
|
|
591 |
};
|
|
|
592 |
|
|
|
593 |
|
|
|
594 |
|
|
|
595 |
|
|
|
596 |
|
|
|
597 |
|
|
|
598 |
|
|
|
599 |
|
|
|
600 |
|
|
|
601 |
void Assert(TState &state,bool cond, const char *msg)
|
|
|
602 |
{ if (cond) return;
|
|
|
603 |
state.err=msg;
|
|
|
604 |
}
|
|
|
605 |
void __cdecl Trace(const char *x, ...) {va_list paramList; va_start(paramList, x); paramList; va_end(paramList);}
|
|
|
606 |
void __cdecl Tracec(bool ,const char *x, ...) {va_list paramList; va_start(paramList, x); paramList; va_end(paramList);}
|
|
|
607 |
|
|
|
608 |
|
|
|
609 |
|
|
|
610 |
// ===========================================================================
|
|
|
611 |
// Local (static) routines in this file.
|
|
|
612 |
//
|
|
|
613 |
|
|
|
614 |
void init_block (TState &);
|
|
|
615 |
void pqdownheap (TState &,ct_data *tree, int k);
|
|
|
616 |
void gen_bitlen (TState &,tree_desc *desc);
|
|
|
617 |
void gen_codes (TState &state,ct_data *tree, int max_code);
|
|
|
618 |
void build_tree (TState &,tree_desc *desc);
|
|
|
619 |
void scan_tree (TState &,ct_data *tree, int max_code);
|
|
|
620 |
void send_tree (TState &state,ct_data *tree, int max_code);
|
|
|
621 |
int build_bl_tree (TState &);
|
|
|
622 |
void send_all_trees (TState &state,int lcodes, int dcodes, int blcodes);
|
|
|
623 |
void compress_block (TState &state,ct_data *ltree, ct_data *dtree);
|
|
|
624 |
void set_file_type (TState &);
|
|
|
625 |
void send_bits (TState &state, int value, int length);
|
|
|
626 |
unsigned bi_reverse (unsigned code, int len);
|
|
|
627 |
void bi_windup (TState &state);
|
|
|
628 |
void copy_block (TState &state,char *buf, unsigned len, int header);
|
|
|
629 |
|
|
|
630 |
|
|
|
631 |
#define send_code(state, c, tree) send_bits(state, tree[c].fc.code, tree[c].dl.len)
|
|
|
632 |
// Send a code of the given tree. c and tree must not have side effects
|
|
|
633 |
|
|
|
634 |
// alternatively...
|
|
|
635 |
//#define send_code(state, c, tree)
|
|
|
636 |
// { if (state.verbose>1) fprintf(stderr,"\ncd %3d ",(c));
|
|
|
637 |
// send_bits(state, tree[c].fc.code, tree[c].dl.len); }
|
|
|
638 |
|
|
|
639 |
#define d_code(dist) ((dist) < 256 ? state.ts.dist_code[dist] : state.ts.dist_code[256+((dist)>>7)])
|
|
|
640 |
// Mapping from a distance to a distance code. dist is the distance - 1 and
|
|
|
641 |
// must not have side effects. dist_code[256] and dist_code[257] are never used.
|
|
|
642 |
|
|
|
643 |
#define Max(a,b) (a >= b ? a : b)
|
|
|
644 |
/* the arguments must not have side effects */
|
|
|
645 |
|
|
|
646 |
/* ===========================================================================
|
|
|
647 |
* Allocate the match buffer, initialize the various tables and save the
|
|
|
648 |
* location of the internal file attribute (ascii/binary) and method
|
|
|
649 |
* (DEFLATE/STORE).
|
|
|
650 |
*/
|
|
|
651 |
void ct_init(TState &state, ush *attr)
|
|
|
652 |
{
|
|
|
653 |
int n; /* iterates over tree elements */
|
|
|
654 |
int bits; /* bit counter */
|
|
|
655 |
int length; /* length value */
|
|
|
656 |
int code; /* code value */
|
|
|
657 |
int dist; /* distance index */
|
|
|
658 |
|
|
|
659 |
state.ts.file_type = attr;
|
|
|
660 |
//state.ts.file_method = method;
|
|
|
661 |
state.ts.cmpr_bytelen = state.ts.cmpr_len_bits = 0L;
|
|
|
662 |
state.ts.input_len = 0L;
|
|
|
663 |
|
|
|
664 |
if (state.ts.static_dtree[0].dl.len != 0) return; /* ct_init already called */
|
|
|
665 |
|
|
|
666 |
/* Initialize the mapping length (0..255) -> length code (0..28) */
|
|
|
667 |
length = 0;
|
|
|
668 |
for (code = 0; code < LENGTH_CODES-1; code++) {
|
|
|
669 |
state.ts.base_length[code] = length;
|
|
|
670 |
for (n = 0; n < (1<<extra_lbits[code]); n++) {
|
|
|
671 |
state.ts.length_code[length++] = (uch)code;
|
|
|
672 |
}
|
|
|
673 |
}
|
|
|
674 |
Assert(state,length == 256, "ct_init: length != 256");
|
|
|
675 |
/* Note that the length 255 (match length 258) can be represented
|
|
|
676 |
* in two different ways: code 284 + 5 bits or code 285, so we
|
|
|
677 |
* overwrite length_code[255] to use the best encoding:
|
|
|
678 |
*/
|
|
|
679 |
state.ts.length_code[length-1] = (uch)code;
|
|
|
680 |
|
|
|
681 |
/* Initialize the mapping dist (0..32K) -> dist code (0..29) */
|
|
|
682 |
dist = 0;
|
|
|
683 |
for (code = 0 ; code < 16; code++) {
|
|
|
684 |
state.ts.base_dist[code] = dist;
|
|
|
685 |
for (n = 0; n < (1<<extra_dbits[code]); n++) {
|
|
|
686 |
state.ts.dist_code[dist++] = (uch)code;
|
|
|
687 |
}
|
|
|
688 |
}
|
|
|
689 |
Assert(state,dist == 256, "ct_init: dist != 256");
|
|
|
690 |
dist >>= 7; /* from now on, all distances are divided by 128 */
|
|
|
691 |
for ( ; code < D_CODES; code++) {
|
|
|
692 |
state.ts.base_dist[code] = dist << 7;
|
|
|
693 |
for (n = 0; n < (1<<(extra_dbits[code]-7)); n++) {
|
|
|
694 |
state.ts.dist_code[256 + dist++] = (uch)code;
|
|
|
695 |
}
|
|
|
696 |
}
|
|
|
697 |
Assert(state,dist == 256, "ct_init: 256+dist != 512");
|
|
|
698 |
|
|
|
699 |
/* Construct the codes of the static literal tree */
|
|
|
700 |
for (bits = 0; bits <= MAX_BITS; bits++) state.ts.bl_count[bits] = 0;
|
|
|
701 |
n = 0;
|
|
|
702 |
while (n <= 143) state.ts.static_ltree[n++].dl.len = 8, state.ts.bl_count[8]++;
|
|
|
703 |
while (n <= 255) state.ts.static_ltree[n++].dl.len = 9, state.ts.bl_count[9]++;
|
|
|
704 |
while (n <= 279) state.ts.static_ltree[n++].dl.len = 7, state.ts.bl_count[7]++;
|
|
|
705 |
while (n <= 287) state.ts.static_ltree[n++].dl.len = 8, state.ts.bl_count[8]++;
|
|
|
706 |
/* fc.codes 286 and 287 do not exist, but we must include them in the
|
|
|
707 |
* tree construction to get a canonical Huffman tree (longest code
|
|
|
708 |
* all ones)
|
|
|
709 |
*/
|
|
|
710 |
gen_codes(state,(ct_data *)state.ts.static_ltree, L_CODES+1);
|
|
|
711 |
|
|
|
712 |
/* The static distance tree is trivial: */
|
|
|
713 |
for (n = 0; n < D_CODES; n++) {
|
|
|
714 |
state.ts.static_dtree[n].dl.len = 5;
|
|
|
715 |
state.ts.static_dtree[n].fc.code = (ush)bi_reverse(n, 5);
|
|
|
716 |
}
|
|
|
717 |
|
|
|
718 |
/* Initialize the first block of the first file: */
|
|
|
719 |
init_block(state);
|
|
|
720 |
}
|
|
|
721 |
|
|
|
722 |
/* ===========================================================================
|
|
|
723 |
* Initialize a new block.
|
|
|
724 |
*/
|
|
|
725 |
void init_block(TState &state)
|
|
|
726 |
{
|
|
|
727 |
int n; /* iterates over tree elements */
|
|
|
728 |
|
|
|
729 |
/* Initialize the trees. */
|
|
|
730 |
for (n = 0; n < L_CODES; n++) state.ts.dyn_ltree[n].fc.freq = 0;
|
|
|
731 |
for (n = 0; n < D_CODES; n++) state.ts.dyn_dtree[n].fc.freq = 0;
|
|
|
732 |
for (n = 0; n < BL_CODES; n++) state.ts.bl_tree[n].fc.freq = 0;
|
|
|
733 |
|
|
|
734 |
state.ts.dyn_ltree[END_BLOCK].fc.freq = 1;
|
|
|
735 |
state.ts.opt_len = state.ts.static_len = 0L;
|
|
|
736 |
state.ts.last_lit = state.ts.last_dist = state.ts.last_flags = 0;
|
|
|
737 |
state.ts.flags = 0; state.ts.flag_bit = 1;
|
|
|
738 |
}
|
|
|
739 |
|
|
|
740 |
#define SMALLEST 1
|
|
|
741 |
/* Index within the heap array of least frequent node in the Huffman tree */
|
|
|
742 |
|
|
|
743 |
|
|
|
744 |
/* ===========================================================================
|
|
|
745 |
* Remove the smallest element from the heap and recreate the heap with
|
|
|
746 |
* one less element. Updates heap and heap_len.
|
|
|
747 |
*/
|
|
|
748 |
#define pqremove(tree, top) \
|
|
|
749 |
{\
|
|
|
750 |
top = state.ts.heap[SMALLEST]; \
|
|
|
751 |
state.ts.heap[SMALLEST] = state.ts.heap[state.ts.heap_len--]; \
|
|
|
752 |
pqdownheap(state,tree, SMALLEST); \
|
|
|
753 |
}
|
|
|
754 |
|
|
|
755 |
/* ===========================================================================
|
|
|
756 |
* Compares to subtrees, using the tree depth as tie breaker when
|
|
|
757 |
* the subtrees have equal frequency. This minimizes the worst case length.
|
|
|
758 |
*/
|
|
|
759 |
#define smaller(tree, n, m) \
|
|
|
760 |
(tree[n].fc.freq < tree[m].fc.freq || \
|
|
|
761 |
(tree[n].fc.freq == tree[m].fc.freq && state.ts.depth[n] <= state.ts.depth[m]))
|
|
|
762 |
|
|
|
763 |
/* ===========================================================================
|
|
|
764 |
* Restore the heap property by moving down the tree starting at node k,
|
|
|
765 |
* exchanging a node with the smallest of its two sons if necessary, stopping
|
|
|
766 |
* when the heap property is re-established (each father smaller than its
|
|
|
767 |
* two sons).
|
|
|
768 |
*/
|
|
|
769 |
void pqdownheap(TState &state,ct_data *tree, int k)
|
|
|
770 |
{
|
|
|
771 |
int v = state.ts.heap[k];
|
|
|
772 |
int j = k << 1; /* left son of k */
|
|
|
773 |
int htemp; /* required because of bug in SASC compiler */
|
|
|
774 |
|
|
|
775 |
while (j <= state.ts.heap_len) {
|
|
|
776 |
/* Set j to the smallest of the two sons: */
|
|
|
777 |
if (j < state.ts.heap_len && smaller(tree, state.ts.heap[j+1], state.ts.heap[j])) j++;
|
|
|
778 |
|
|
|
779 |
/* Exit if v is smaller than both sons */
|
|
|
780 |
htemp = state.ts.heap[j];
|
|
|
781 |
if (smaller(tree, v, htemp)) break;
|
|
|
782 |
|
|
|
783 |
/* Exchange v with the smallest son */
|
|
|
784 |
state.ts.heap[k] = htemp;
|
|
|
785 |
k = j;
|
|
|
786 |
|
|
|
787 |
/* And continue down the tree, setting j to the left son of k */
|
|
|
788 |
j <<= 1;
|
|
|
789 |
}
|
|
|
790 |
state.ts.heap[k] = v;
|
|
|
791 |
}
|
|
|
792 |
|
|
|
793 |
/* ===========================================================================
|
|
|
794 |
* Compute the optimal bit lengths for a tree and update the total bit length
|
|
|
795 |
* for the current block.
|
|
|
796 |
* IN assertion: the fields freq and dad are set, heap[heap_max] and
|
|
|
797 |
* above are the tree nodes sorted by increasing frequency.
|
|
|
798 |
* OUT assertions: the field len is set to the optimal bit length, the
|
|
|
799 |
* array bl_count contains the frequencies for each bit length.
|
|
|
800 |
* The length opt_len is updated; static_len is also updated if stree is
|
|
|
801 |
* not null.
|
|
|
802 |
*/
|
|
|
803 |
void gen_bitlen(TState &state,tree_desc *desc)
|
|
|
804 |
{
|
|
|
805 |
ct_data *tree = desc->dyn_tree;
|
|
|
806 |
const int *extra = desc->extra_bits;
|
|
|
807 |
int base = desc->extra_base;
|
|
|
808 |
int max_code = desc->max_code;
|
|
|
809 |
int max_length = desc->max_length;
|
|
|
810 |
ct_data *stree = desc->static_tree;
|
|
|
811 |
int h; /* heap index */
|
|
|
812 |
int n, m; /* iterate over the tree elements */
|
|
|
813 |
int bits; /* bit length */
|
|
|
814 |
int xbits; /* extra bits */
|
|
|
815 |
ush f; /* frequency */
|
|
|
816 |
int overflow = 0; /* number of elements with bit length too large */
|
|
|
817 |
|
|
|
818 |
for (bits = 0; bits <= MAX_BITS; bits++) state.ts.bl_count[bits] = 0;
|
|
|
819 |
|
|
|
820 |
/* In a first pass, compute the optimal bit lengths (which may
|
|
|
821 |
* overflow in the case of the bit length tree).
|
|
|
822 |
*/
|
|
|
823 |
tree[state.ts.heap[state.ts.heap_max]].dl.len = 0; /* root of the heap */
|
|
|
824 |
|
|
|
825 |
for (h = state.ts.heap_max+1; h < HEAP_SIZE; h++) {
|
|
|
826 |
n = state.ts.heap[h];
|
|
|
827 |
bits = tree[tree[n].dl.dad].dl.len + 1;
|
|
|
828 |
if (bits > max_length) bits = max_length, overflow++;
|
|
|
829 |
tree[n].dl.len = (ush)bits;
|
|
|
830 |
/* We overwrite tree[n].dl.dad which is no longer needed */
|
|
|
831 |
|
|
|
832 |
if (n > max_code) continue; /* not a leaf node */
|
|
|
833 |
|
|
|
834 |
state.ts.bl_count[bits]++;
|
|
|
835 |
xbits = 0;
|
|
|
836 |
if (n >= base) xbits = extra[n-base];
|
|
|
837 |
f = tree[n].fc.freq;
|
|
|
838 |
state.ts.opt_len += (ulg)f * (bits + xbits);
|
|
|
839 |
if (stree) state.ts.static_len += (ulg)f * (stree[n].dl.len + xbits);
|
|
|
840 |
}
|
|
|
841 |
if (overflow == 0) return;
|
|
|
842 |
|
|
|
843 |
Trace("\nbit length overflow\n");
|
|
|
844 |
/* This happens for example on obj2 and pic of the Calgary corpus */
|
|
|
845 |
|
|
|
846 |
/* Find the first bit length which could increase: */
|
|
|
847 |
do {
|
|
|
848 |
bits = max_length-1;
|
|
|
849 |
while (state.ts.bl_count[bits] == 0) bits--;
|
|
|
850 |
state.ts.bl_count[bits]--; /* move one leaf down the tree */
|
|
|
851 |
state.ts.bl_count[bits+1] += (ush)2; /* move one overflow item as its brother */
|
|
|
852 |
state.ts.bl_count[max_length]--;
|
|
|
853 |
/* The brother of the overflow item also moves one step up,
|
|
|
854 |
* but this does not affect bl_count[max_length]
|
|
|
855 |
*/
|
|
|
856 |
overflow -= 2;
|
|
|
857 |
} while (overflow > 0);
|
|
|
858 |
|
|
|
859 |
/* Now recompute all bit lengths, scanning in increasing frequency.
|
|
|
860 |
* h is still equal to HEAP_SIZE. (It is simpler to reconstruct all
|
|
|
861 |
* lengths instead of fixing only the wrong ones. This idea is taken
|
|
|
862 |
* from 'ar' written by Haruhiko Okumura.)
|
|
|
863 |
*/
|
|
|
864 |
for (bits = max_length; bits != 0; bits--) {
|
|
|
865 |
n = state.ts.bl_count[bits];
|
|
|
866 |
while (n != 0) {
|
|
|
867 |
m = state.ts.heap[--h];
|
|
|
868 |
if (m > max_code) continue;
|
|
|
869 |
if (tree[m].dl.len != (ush)bits) {
|
|
|
870 |
Trace("code %d bits %d->%d\n", m, tree[m].dl.len, bits);
|
|
|
871 |
state.ts.opt_len += ((long)bits-(long)tree[m].dl.len)*(long)tree[m].fc.freq;
|
|
|
872 |
tree[m].dl.len = (ush)bits;
|
|
|
873 |
}
|
|
|
874 |
n--;
|
|
|
875 |
}
|
|
|
876 |
}
|
|
|
877 |
}
|
|
|
878 |
|
|
|
879 |
/* ===========================================================================
|
|
|
880 |
* Generate the codes for a given tree and bit counts (which need not be
|
|
|
881 |
* optimal).
|
|
|
882 |
* IN assertion: the array bl_count contains the bit length statistics for
|
|
|
883 |
* the given tree and the field len is set for all tree elements.
|
|
|
884 |
* OUT assertion: the field code is set for all tree elements of non
|
|
|
885 |
* zero code length.
|
|
|
886 |
*/
|
|
|
887 |
void gen_codes (TState &state, ct_data *tree, int max_code)
|
|
|
888 |
{
|
|
|
889 |
ush next_code[MAX_BITS+1]; /* next code value for each bit length */
|
|
|
890 |
ush code = 0; /* running code value */
|
|
|
891 |
int bits; /* bit index */
|
|
|
892 |
int n; /* code index */
|
|
|
893 |
|
|
|
894 |
/* The distribution counts are first used to generate the code values
|
|
|
895 |
* without bit reversal.
|
|
|
896 |
*/
|
|
|
897 |
for (bits = 1; bits <= MAX_BITS; bits++) {
|
|
|
898 |
next_code[bits] = code = (ush)((code + state.ts.bl_count[bits-1]) << 1);
|
|
|
899 |
}
|
|
|
900 |
/* Check that the bit counts in bl_count are consistent. The last code
|
|
|
901 |
* must be all ones.
|
|
|
902 |
*/
|
|
|
903 |
Assert(state,code + state.ts.bl_count[MAX_BITS]-1 == (1<< ((ush) MAX_BITS)) - 1,
|
|
|
904 |
"inconsistent bit counts");
|
|
|
905 |
Trace("\ngen_codes: max_code %d ", max_code);
|
|
|
906 |
|
|
|
907 |
for (n = 0; n <= max_code; n++) {
|
|
|
908 |
int len = tree[n].dl.len;
|
|
|
909 |
if (len == 0) continue;
|
|
|
910 |
/* Now reverse the bits */
|
|
|
911 |
tree[n].fc.code = (ush)bi_reverse(next_code[len]++, len);
|
|
|
912 |
|
|
|
913 |
//Tracec(tree != state.ts.static_ltree, "\nn %3d %c l %2d c %4x (%x) ", n, (isgraph(n) ? n : ' '), len, tree[n].fc.code, next_code[len]-1);
|
|
|
914 |
}
|
|
|
915 |
}
|
|
|
916 |
|
|
|
917 |
/* ===========================================================================
|
|
|
918 |
* Construct one Huffman tree and assigns the code bit strings and lengths.
|
|
|
919 |
* Update the total bit length for the current block.
|
|
|
920 |
* IN assertion: the field freq is set for all tree elements.
|
|
|
921 |
* OUT assertions: the fields len and code are set to the optimal bit length
|
|
|
922 |
* and corresponding code. The length opt_len is updated; static_len is
|
|
|
923 |
* also updated if stree is not null. The field max_code is set.
|
|
|
924 |
*/
|
|
|
925 |
void build_tree(TState &state,tree_desc *desc)
|
|
|
926 |
{
|
|
|
927 |
ct_data *tree = desc->dyn_tree;
|
|
|
928 |
ct_data *stree = desc->static_tree;
|
|
|
929 |
int elems = desc->elems;
|
|
|
930 |
int n, m; /* iterate over heap elements */
|
|
|
931 |
int max_code = -1; /* largest code with non zero frequency */
|
|
|
932 |
int node = elems; /* next internal node of the tree */
|
|
|
933 |
|
|
|
934 |
/* Construct the initial heap, with least frequent element in
|
|
|
935 |
* heap[SMALLEST]. The sons of heap[n] are heap[2*n] and heap[2*n+1].
|
|
|
936 |
* heap[0] is not used.
|
|
|
937 |
*/
|
|
|
938 |
state.ts.heap_len = 0, state.ts.heap_max = HEAP_SIZE;
|
|
|
939 |
|
|
|
940 |
for (n = 0; n < elems; n++) {
|
|
|
941 |
if (tree[n].fc.freq != 0) {
|
|
|
942 |
state.ts.heap[++state.ts.heap_len] = max_code = n;
|
|
|
943 |
state.ts.depth[n] = 0;
|
|
|
944 |
} else {
|
|
|
945 |
tree[n].dl.len = 0;
|
|
|
946 |
}
|
|
|
947 |
}
|
|
|
948 |
|
|
|
949 |
/* The pkzip format requires that at least one distance code exists,
|
|
|
950 |
* and that at least one bit should be sent even if there is only one
|
|
|
951 |
* possible code. So to avoid special checks later on we force at least
|
|
|
952 |
* two codes of non zero frequency.
|
|
|
953 |
*/
|
|
|
954 |
while (state.ts.heap_len < 2) {
|
|
|
955 |
int newcp = state.ts.heap[++state.ts.heap_len] = (max_code < 2 ? ++max_code : 0);
|
|
|
956 |
tree[newcp].fc.freq = 1;
|
|
|
957 |
state.ts.depth[newcp] = 0;
|
|
|
958 |
state.ts.opt_len--; if (stree) state.ts.static_len -= stree[newcp].dl.len;
|
|
|
959 |
/* new is 0 or 1 so it does not have extra bits */
|
|
|
960 |
}
|
|
|
961 |
desc->max_code = max_code;
|
|
|
962 |
|
|
|
963 |
/* The elements heap[heap_len/2+1 .. heap_len] are leaves of the tree,
|
|
|
964 |
* establish sub-heaps of increasing lengths:
|
|
|
965 |
*/
|
|
|
966 |
for (n = state.ts.heap_len/2; n >= 1; n--) pqdownheap(state,tree, n);
|
|
|
967 |
|
|
|
968 |
/* Construct the Huffman tree by repeatedly combining the least two
|
|
|
969 |
* frequent nodes.
|
|
|
970 |
*/
|
|
|
971 |
do {
|
|
|
972 |
pqremove(tree, n); /* n = node of least frequency */
|
|
|
973 |
m = state.ts.heap[SMALLEST]; /* m = node of next least frequency */
|
|
|
974 |
|
|
|
975 |
state.ts.heap[--state.ts.heap_max] = n; /* keep the nodes sorted by frequency */
|
|
|
976 |
state.ts.heap[--state.ts.heap_max] = m;
|
|
|
977 |
|
|
|
978 |
/* Create a new node father of n and m */
|
|
|
979 |
tree[node].fc.freq = (ush)(tree[n].fc.freq + tree[m].fc.freq);
|
|
|
980 |
state.ts.depth[node] = (uch) (Max(state.ts.depth[n], state.ts.depth[m]) + 1);
|
|
|
981 |
tree[n].dl.dad = tree[m].dl.dad = (ush)node;
|
|
|
982 |
/* and insert the new node in the heap */
|
|
|
983 |
state.ts.heap[SMALLEST] = node++;
|
|
|
984 |
pqdownheap(state,tree, SMALLEST);
|
|
|
985 |
|
|
|
986 |
} while (state.ts.heap_len >= 2);
|
|
|
987 |
|
|
|
988 |
state.ts.heap[--state.ts.heap_max] = state.ts.heap[SMALLEST];
|
|
|
989 |
|
|
|
990 |
/* At this point, the fields freq and dad are set. We can now
|
|
|
991 |
* generate the bit lengths.
|
|
|
992 |
*/
|
|
|
993 |
gen_bitlen(state,(tree_desc *)desc);
|
|
|
994 |
|
|
|
995 |
/* The field len is now set, we can generate the bit codes */
|
|
|
996 |
gen_codes (state,(ct_data *)tree, max_code);
|
|
|
997 |
}
|
|
|
998 |
|
|
|
999 |
/* ===========================================================================
|
|
|
1000 |
* Scan a literal or distance tree to determine the frequencies of the codes
|
|
|
1001 |
* in the bit length tree. Updates opt_len to take into account the repeat
|
|
|
1002 |
* counts. (The contribution of the bit length codes will be added later
|
|
|
1003 |
* during the construction of bl_tree.)
|
|
|
1004 |
*/
|
|
|
1005 |
void scan_tree (TState &state,ct_data *tree, int max_code)
|
|
|
1006 |
{
|
|
|
1007 |
int n; /* iterates over all tree elements */
|
|
|
1008 |
int prevlen = -1; /* last emitted length */
|
|
|
1009 |
int curlen; /* length of current code */
|
|
|
1010 |
int nextlen = tree[0].dl.len; /* length of next code */
|
|
|
1011 |
int count = 0; /* repeat count of the current code */
|
|
|
1012 |
int max_count = 7; /* max repeat count */
|
|
|
1013 |
int min_count = 4; /* min repeat count */
|
|
|
1014 |
|
|
|
1015 |
if (nextlen == 0) max_count = 138, min_count = 3;
|
|
|
1016 |
tree[max_code+1].dl.len = (ush)-1; /* guard */
|
|
|
1017 |
|
|
|
1018 |
for (n = 0; n <= max_code; n++) {
|
|
|
1019 |
curlen = nextlen; nextlen = tree[n+1].dl.len;
|
|
|
1020 |
if (++count < max_count && curlen == nextlen) {
|
|
|
1021 |
continue;
|
|
|
1022 |
} else if (count < min_count) {
|
|
|
1023 |
state.ts.bl_tree[curlen].fc.freq = (ush)(state.ts.bl_tree[curlen].fc.freq + count);
|
|
|
1024 |
} else if (curlen != 0) {
|
|
|
1025 |
if (curlen != prevlen) state.ts.bl_tree[curlen].fc.freq++;
|
|
|
1026 |
state.ts.bl_tree[REP_3_6].fc.freq++;
|
|
|
1027 |
} else if (count <= 10) {
|
|
|
1028 |
state.ts.bl_tree[REPZ_3_10].fc.freq++;
|
|
|
1029 |
} else {
|
|
|
1030 |
state.ts.bl_tree[REPZ_11_138].fc.freq++;
|
|
|
1031 |
}
|
|
|
1032 |
count = 0; prevlen = curlen;
|
|
|
1033 |
if (nextlen == 0) {
|
|
|
1034 |
max_count = 138, min_count = 3;
|
|
|
1035 |
} else if (curlen == nextlen) {
|
|
|
1036 |
max_count = 6, min_count = 3;
|
|
|
1037 |
} else {
|
|
|
1038 |
max_count = 7, min_count = 4;
|
|
|
1039 |
}
|
|
|
1040 |
}
|
|
|
1041 |
}
|
|
|
1042 |
|
|
|
1043 |
/* ===========================================================================
|
|
|
1044 |
* Send a literal or distance tree in compressed form, using the codes in
|
|
|
1045 |
* bl_tree.
|
|
|
1046 |
*/
|
|
|
1047 |
void send_tree (TState &state, ct_data *tree, int max_code)
|
|
|
1048 |
{
|
|
|
1049 |
int n; /* iterates over all tree elements */
|
|
|
1050 |
int prevlen = -1; /* last emitted length */
|
|
|
1051 |
int curlen; /* length of current code */
|
|
|
1052 |
int nextlen = tree[0].dl.len; /* length of next code */
|
|
|
1053 |
int count = 0; /* repeat count of the current code */
|
|
|
1054 |
int max_count = 7; /* max repeat count */
|
|
|
1055 |
int min_count = 4; /* min repeat count */
|
|
|
1056 |
|
|
|
1057 |
/* tree[max_code+1].dl.len = -1; */ /* guard already set */
|
|
|
1058 |
if (nextlen == 0) max_count = 138, min_count = 3;
|
|
|
1059 |
|
|
|
1060 |
for (n = 0; n <= max_code; n++) {
|
|
|
1061 |
curlen = nextlen; nextlen = tree[n+1].dl.len;
|
|
|
1062 |
if (++count < max_count && curlen == nextlen) {
|
|
|
1063 |
continue;
|
|
|
1064 |
} else if (count < min_count) {
|
|
|
1065 |
do { send_code(state, curlen, state.ts.bl_tree); } while (--count != 0);
|
|
|
1066 |
|
|
|
1067 |
} else if (curlen != 0) {
|
|
|
1068 |
if (curlen != prevlen) {
|
|
|
1069 |
send_code(state, curlen, state.ts.bl_tree); count--;
|
|
|
1070 |
}
|
|
|
1071 |
Assert(state,count >= 3 && count <= 6, " 3_6?");
|
|
|
1072 |
send_code(state,REP_3_6, state.ts.bl_tree); send_bits(state,count-3, 2);
|
|
|
1073 |
|
|
|
1074 |
} else if (count <= 10) {
|
|
|
1075 |
send_code(state,REPZ_3_10, state.ts.bl_tree); send_bits(state,count-3, 3);
|
|
|
1076 |
|
|
|
1077 |
} else {
|
|
|
1078 |
send_code(state,REPZ_11_138, state.ts.bl_tree); send_bits(state,count-11, 7);
|
|
|
1079 |
}
|
|
|
1080 |
count = 0; prevlen = curlen;
|
|
|
1081 |
if (nextlen == 0) {
|
|
|
1082 |
max_count = 138, min_count = 3;
|
|
|
1083 |
} else if (curlen == nextlen) {
|
|
|
1084 |
max_count = 6, min_count = 3;
|
|
|
1085 |
} else {
|
|
|
1086 |
max_count = 7, min_count = 4;
|
|
|
1087 |
}
|
|
|
1088 |
}
|
|
|
1089 |
}
|
|
|
1090 |
|
|
|
1091 |
/* ===========================================================================
|
|
|
1092 |
* Construct the Huffman tree for the bit lengths and return the index in
|
|
|
1093 |
* bl_order of the last bit length code to send.
|
|
|
1094 |
*/
|
|
|
1095 |
int build_bl_tree(TState &state)
|
|
|
1096 |
{
|
|
|
1097 |
int max_blindex; /* index of last bit length code of non zero freq */
|
|
|
1098 |
|
|
|
1099 |
/* Determine the bit length frequencies for literal and distance trees */
|
|
|
1100 |
scan_tree(state,(ct_data *)state.ts.dyn_ltree, state.ts.l_desc.max_code);
|
|
|
1101 |
scan_tree(state,(ct_data *)state.ts.dyn_dtree, state.ts.d_desc.max_code);
|
|
|
1102 |
|
|
|
1103 |
/* Build the bit length tree: */
|
|
|
1104 |
build_tree(state,(tree_desc *)(&state.ts.bl_desc));
|
|
|
1105 |
/* opt_len now includes the length of the tree representations, except
|
|
|
1106 |
* the lengths of the bit lengths codes and the 5+5+4 bits for the counts.
|
|
|
1107 |
*/
|
|
|
1108 |
|
|
|
1109 |
/* Determine the number of bit length codes to send. The pkzip format
|
|
|
1110 |
* requires that at least 4 bit length codes be sent. (appnote.txt says
|
|
|
1111 |
* 3 but the actual value used is 4.)
|
|
|
1112 |
*/
|
|
|
1113 |
for (max_blindex = BL_CODES-1; max_blindex >= 3; max_blindex--) {
|
|
|
1114 |
if (state.ts.bl_tree[bl_order[max_blindex]].dl.len != 0) break;
|
|
|
1115 |
}
|
|
|
1116 |
/* Update opt_len to include the bit length tree and counts */
|
|
|
1117 |
state.ts.opt_len += 3*(max_blindex+1) + 5+5+4;
|
|
|
1118 |
Trace("\ndyn trees: dyn %ld, stat %ld", state.ts.opt_len, state.ts.static_len);
|
|
|
1119 |
|
|
|
1120 |
return max_blindex;
|
|
|
1121 |
}
|
|
|
1122 |
|
|
|
1123 |
/* ===========================================================================
|
|
|
1124 |
* Send the header for a block using dynamic Huffman trees: the counts, the
|
|
|
1125 |
* lengths of the bit length codes, the literal tree and the distance tree.
|
|
|
1126 |
* IN assertion: lcodes >= 257, dcodes >= 1, blcodes >= 4.
|
|
|
1127 |
*/
|
|
|
1128 |
void send_all_trees(TState &state,int lcodes, int dcodes, int blcodes)
|
|
|
1129 |
{
|
|
|
1130 |
int rank; /* index in bl_order */
|
|
|
1131 |
|
|
|
1132 |
Assert(state,lcodes >= 257 && dcodes >= 1 && blcodes >= 4, "not enough codes");
|
|
|
1133 |
Assert(state,lcodes <= L_CODES && dcodes <= D_CODES && blcodes <= BL_CODES,
|
|
|
1134 |
"too many codes");
|
|
|
1135 |
Trace("\nbl counts: ");
|
|
|
1136 |
send_bits(state,lcodes-257, 5);
|
|
|
1137 |
/* not +255 as stated in appnote.txt 1.93a or -256 in 2.04c */
|
|
|
1138 |
send_bits(state,dcodes-1, 5);
|
|
|
1139 |
send_bits(state,blcodes-4, 4); /* not -3 as stated in appnote.txt */
|
|
|
1140 |
for (rank = 0; rank < blcodes; rank++) {
|
|
|
1141 |
Trace("\nbl code %2d ", bl_order[rank]);
|
|
|
1142 |
send_bits(state,state.ts.bl_tree[bl_order[rank]].dl.len, 3);
|
|
|
1143 |
}
|
|
|
1144 |
Trace("\nbl tree: sent %ld", state.bs.bits_sent);
|
|
|
1145 |
|
|
|
1146 |
send_tree(state,(ct_data *)state.ts.dyn_ltree, lcodes-1); /* send the literal tree */
|
|
|
1147 |
Trace("\nlit tree: sent %ld", state.bs.bits_sent);
|
|
|
1148 |
|
|
|
1149 |
send_tree(state,(ct_data *)state.ts.dyn_dtree, dcodes-1); /* send the distance tree */
|
|
|
1150 |
Trace("\ndist tree: sent %ld", state.bs.bits_sent);
|
|
|
1151 |
}
|
|
|
1152 |
|
|
|
1153 |
/* ===========================================================================
|
|
|
1154 |
* Determine the best encoding for the current block: dynamic trees, static
|
|
|
1155 |
* trees or store, and output the encoded block to the zip file. This function
|
|
|
1156 |
* returns the total compressed length (in bytes) for the file so far.
|
|
|
1157 |
*/
|
|
|
1158 |
ulg flush_block(TState &state,char *buf, ulg stored_len, int eof)
|
|
|
1159 |
{
|
|
|
1160 |
ulg opt_lenb, static_lenb; /* opt_len and static_len in bytes */
|
|
|
1161 |
int max_blindex; /* index of last bit length code of non zero freq */
|
|
|
1162 |
|
|
|
1163 |
state.ts.flag_buf[state.ts.last_flags] = state.ts.flags; /* Save the flags for the last 8 items */
|
|
|
1164 |
|
|
|
1165 |
/* Check if the file is ascii or binary */
|
|
|
1166 |
if (*state.ts.file_type == (ush)UNKNOWN) set_file_type(state);
|
|
|
1167 |
|
|
|
1168 |
/* Construct the literal and distance trees */
|
|
|
1169 |
build_tree(state,(tree_desc *)(&state.ts.l_desc));
|
|
|
1170 |
Trace("\nlit data: dyn %ld, stat %ld", state.ts.opt_len, state.ts.static_len);
|
|
|
1171 |
|
|
|
1172 |
build_tree(state,(tree_desc *)(&state.ts.d_desc));
|
|
|
1173 |
Trace("\ndist data: dyn %ld, stat %ld", state.ts.opt_len, state.ts.static_len);
|
|
|
1174 |
/* At this point, opt_len and static_len are the total bit lengths of
|
|
|
1175 |
* the compressed block data, excluding the tree representations.
|
|
|
1176 |
*/
|
|
|
1177 |
|
|
|
1178 |
/* Build the bit length tree for the above two trees, and get the index
|
|
|
1179 |
* in bl_order of the last bit length code to send.
|
|
|
1180 |
*/
|
|
|
1181 |
max_blindex = build_bl_tree(state);
|
|
|
1182 |
|
|
|
1183 |
/* Determine the best encoding. Compute first the block length in bytes */
|
|
|
1184 |
opt_lenb = (state.ts.opt_len+3+7)>>3;
|
|
|
1185 |
static_lenb = (state.ts.static_len+3+7)>>3;
|
|
|
1186 |
state.ts.input_len += stored_len; /* for debugging only */
|
|
|
1187 |
|
|
|
1188 |
Trace("\nopt %lu(%lu) stat %lu(%lu) stored %lu lit %u dist %u ",
|
|
|
1189 |
opt_lenb, state.ts.opt_len, static_lenb, state.ts.static_len, stored_len,
|
|
|
1190 |
state.ts.last_lit, state.ts.last_dist);
|
|
|
1191 |
|
|
|
1192 |
if (static_lenb <= opt_lenb) opt_lenb = static_lenb;
|
|
|
1193 |
|
|
|
1194 |
// Originally, zip allowed the file to be transformed from a compressed
|
|
|
1195 |
// into a stored file in the case where compression failed, there
|
|
|
1196 |
// was only one block, and it was allowed to change. I've removed this
|
|
|
1197 |
// possibility since the code's cleaner if no changes are allowed.
|
|
|
1198 |
//if (stored_len <= opt_lenb && eof && state.ts.cmpr_bytelen == 0L
|
|
|
1199 |
// && state.ts.cmpr_len_bits == 0L && state.seekable)
|
|
|
1200 |
//{ // && state.ts.file_method != NULL
|
|
|
1201 |
// // Since LIT_BUFSIZE <= 2*WSIZE, the input data must be there:
|
|
|
1202 |
// Assert(state,buf!=NULL,"block vanished");
|
|
|
1203 |
// copy_block(state,buf, (unsigned)stored_len, 0); // without header
|
|
|
1204 |
// state.ts.cmpr_bytelen = stored_len;
|
|
|
1205 |
// Assert(state,false,"unimplemented *state.ts.file_method = STORE;");
|
|
|
1206 |
// //*state.ts.file_method = STORE;
|
|
|
1207 |
//}
|
|
|
1208 |
//else
|
|
|
1209 |
if (stored_len+4 <= opt_lenb && buf != (char*)NULL) {
|
|
|
1210 |
/* 4: two words for the lengths */
|
|
|
1211 |
/* The test buf != NULL is only necessary if LIT_BUFSIZE > WSIZE.
|
|
|
1212 |
* Otherwise we can't have processed more than WSIZE input bytes since
|
|
|
1213 |
* the last block flush, because compression would have been
|
|
|
1214 |
* successful. If LIT_BUFSIZE <= WSIZE, it is never too late to
|
|
|
1215 |
* transform a block into a stored block.
|
|
|
1216 |
*/
|
|
|
1217 |
send_bits(state,(STORED_BLOCK<<1)+eof, 3); /* send block type */
|
|
|
1218 |
state.ts.cmpr_bytelen += ((state.ts.cmpr_len_bits + 3 + 7) >> 3) + stored_len + 4;
|
|
|
1219 |
state.ts.cmpr_len_bits = 0L;
|
|
|
1220 |
|
|
|
1221 |
copy_block(state,buf, (unsigned)stored_len, 1); /* with header */
|
|
|
1222 |
}
|
|
|
1223 |
else if (static_lenb == opt_lenb) {
|
|
|
1224 |
send_bits(state,(STATIC_TREES<<1)+eof, 3);
|
|
|
1225 |
compress_block(state,(ct_data *)state.ts.static_ltree, (ct_data *)state.ts.static_dtree);
|
|
|
1226 |
state.ts.cmpr_len_bits += 3 + state.ts.static_len;
|
|
|
1227 |
state.ts.cmpr_bytelen += state.ts.cmpr_len_bits >> 3;
|
|
|
1228 |
state.ts.cmpr_len_bits &= 7L;
|
|
|
1229 |
}
|
|
|
1230 |
else {
|
|
|
1231 |
send_bits(state,(DYN_TREES<<1)+eof, 3);
|
|
|
1232 |
send_all_trees(state,state.ts.l_desc.max_code+1, state.ts.d_desc.max_code+1, max_blindex+1);
|
|
|
1233 |
compress_block(state,(ct_data *)state.ts.dyn_ltree, (ct_data *)state.ts.dyn_dtree);
|
|
|
1234 |
state.ts.cmpr_len_bits += 3 + state.ts.opt_len;
|
|
|
1235 |
state.ts.cmpr_bytelen += state.ts.cmpr_len_bits >> 3;
|
|
|
1236 |
state.ts.cmpr_len_bits &= 7L;
|
|
|
1237 |
}
|
|
|
1238 |
Assert(state,((state.ts.cmpr_bytelen << 3) + state.ts.cmpr_len_bits) == state.bs.bits_sent, "bad compressed size");
|
|
|
1239 |
init_block(state);
|
|
|
1240 |
|
|
|
1241 |
if (eof) {
|
|
|
1242 |
// Assert(state,input_len == isize, "bad input size");
|
|
|
1243 |
bi_windup(state);
|
|
|
1244 |
state.ts.cmpr_len_bits += 7; /* align on byte boundary */
|
|
|
1245 |
}
|
|
|
1246 |
Trace("\n");
|
|
|
1247 |
|
|
|
1248 |
return state.ts.cmpr_bytelen + (state.ts.cmpr_len_bits >> 3);
|
|
|
1249 |
}
|
|
|
1250 |
|
|
|
1251 |
/* ===========================================================================
|
|
|
1252 |
* Save the match info and tally the frequency counts. Return true if
|
|
|
1253 |
* the current block must be flushed.
|
|
|
1254 |
*/
|
|
|
1255 |
int ct_tally (TState &state,int dist, int lc)
|
|
|
1256 |
{
|
|
|
1257 |
state.ts.l_buf[state.ts.last_lit++] = (uch)lc;
|
|
|
1258 |
if (dist == 0) {
|
|
|
1259 |
/* lc is the unmatched char */
|
|
|
1260 |
state.ts.dyn_ltree[lc].fc.freq++;
|
|
|
1261 |
} else {
|
|
|
1262 |
/* Here, lc is the match length - MIN_MATCH */
|
|
|
1263 |
dist--; /* dist = match distance - 1 */
|
|
|
1264 |
Assert(state,(ush)dist < (ush)MAX_DIST &&
|
|
|
1265 |
(ush)lc <= (ush)(MAX_MATCH-MIN_MATCH) &&
|
|
|
1266 |
(ush)d_code(dist) < (ush)D_CODES, "ct_tally: bad match");
|
|
|
1267 |
|
|
|
1268 |
state.ts.dyn_ltree[state.ts.length_code[lc]+LITERALS+1].fc.freq++;
|
|
|
1269 |
state.ts.dyn_dtree[d_code(dist)].fc.freq++;
|
|
|
1270 |
|
|
|
1271 |
state.ts.d_buf[state.ts.last_dist++] = (ush)dist;
|
|
|
1272 |
state.ts.flags |= state.ts.flag_bit;
|
|
|
1273 |
}
|
|
|
1274 |
state.ts.flag_bit <<= 1;
|
|
|
1275 |
|
|
|
1276 |
/* Output the flags if they fill a byte: */
|
|
|
1277 |
if ((state.ts.last_lit & 7) == 0) {
|
|
|
1278 |
state.ts.flag_buf[state.ts.last_flags++] = state.ts.flags;
|
|
|
1279 |
state.ts.flags = 0, state.ts.flag_bit = 1;
|
|
|
1280 |
}
|
|
|
1281 |
/* Try to guess if it is profitable to stop the current block here */
|
|
|
1282 |
if (state.level > 2 && (state.ts.last_lit & 0xfff) == 0) {
|
|
|
1283 |
/* Compute an upper bound for the compressed length */
|
|
|
1284 |
ulg out_length = (ulg)state.ts.last_lit*8L;
|
|
|
1285 |
ulg in_length = (ulg)state.ds.strstart-state.ds.block_start;
|
|
|
1286 |
int dcode;
|
|
|
1287 |
for (dcode = 0; dcode < D_CODES; dcode++) {
|
|
|
1288 |
out_length += (ulg)state.ts.dyn_dtree[dcode].fc.freq*(5L+extra_dbits[dcode]);
|
|
|
1289 |
}
|
|
|
1290 |
out_length >>= 3;
|
|
|
1291 |
Trace("\nlast_lit %u, last_dist %u, in %ld, out ~%ld(%ld%%) ",
|
|
|
1292 |
state.ts.last_lit, state.ts.last_dist, in_length, out_length,
|
|
|
1293 |
100L - out_length*100L/in_length);
|
|
|
1294 |
if (state.ts.last_dist < state.ts.last_lit/2 && out_length < in_length/2) return 1;
|
|
|
1295 |
}
|
|
|
1296 |
return (state.ts.last_lit == LIT_BUFSIZE-1 || state.ts.last_dist == DIST_BUFSIZE);
|
|
|
1297 |
/* We avoid equality with LIT_BUFSIZE because of wraparound at 64K
|
|
|
1298 |
* on 16 bit machines and because stored blocks are restricted to
|
|
|
1299 |
* 64K-1 bytes.
|
|
|
1300 |
*/
|
|
|
1301 |
}
|
|
|
1302 |
|
|
|
1303 |
/* ===========================================================================
|
|
|
1304 |
* Send the block data compressed using the given Huffman trees
|
|
|
1305 |
*/
|
|
|
1306 |
void compress_block(TState &state,ct_data *ltree, ct_data *dtree)
|
|
|
1307 |
{
|
|
|
1308 |
unsigned dist; /* distance of matched string */
|
|
|
1309 |
int lc; /* match length or unmatched char (if dist == 0) */
|
|
|
1310 |
unsigned lx = 0; /* running index in l_buf */
|
|
|
1311 |
unsigned dx = 0; /* running index in d_buf */
|
|
|
1312 |
unsigned fx = 0; /* running index in flag_buf */
|
|
|
1313 |
uch flag = 0; /* current flags */
|
|
|
1314 |
unsigned code; /* the code to send */
|
|
|
1315 |
int extra; /* number of extra bits to send */
|
|
|
1316 |
|
|
|
1317 |
if (state.ts.last_lit != 0) do {
|
|
|
1318 |
if ((lx & 7) == 0) flag = state.ts.flag_buf[fx++];
|
|
|
1319 |
lc = state.ts.l_buf[lx++];
|
|
|
1320 |
if ((flag & 1) == 0) {
|
|
|
1321 |
send_code(state,lc, ltree); /* send a literal byte */
|
|
|
1322 |
} else {
|
|
|
1323 |
/* Here, lc is the match length - MIN_MATCH */
|
|
|
1324 |
code = state.ts.length_code[lc];
|
|
|
1325 |
send_code(state,code+LITERALS+1, ltree); /* send the length code */
|
|
|
1326 |
extra = extra_lbits[code];
|
|
|
1327 |
if (extra != 0) {
|
|
|
1328 |
lc -= state.ts.base_length[code];
|
|
|
1329 |
send_bits(state,lc, extra); /* send the extra length bits */
|
|
|
1330 |
}
|
|
|
1331 |
dist = state.ts.d_buf[dx++];
|
|
|
1332 |
/* Here, dist is the match distance - 1 */
|
|
|
1333 |
code = d_code(dist);
|
|
|
1334 |
Assert(state,code < D_CODES, "bad d_code");
|
|
|
1335 |
|
|
|
1336 |
send_code(state,code, dtree); /* send the distance code */
|
|
|
1337 |
extra = extra_dbits[code];
|
|
|
1338 |
if (extra != 0) {
|
|
|
1339 |
dist -= state.ts.base_dist[code];
|
|
|
1340 |
send_bits(state,dist, extra); /* send the extra distance bits */
|
|
|
1341 |
}
|
|
|
1342 |
} /* literal or match pair ? */
|
|
|
1343 |
flag >>= 1;
|
|
|
1344 |
} while (lx < state.ts.last_lit);
|
|
|
1345 |
|
|
|
1346 |
send_code(state,END_BLOCK, ltree);
|
|
|
1347 |
}
|
|
|
1348 |
|
|
|
1349 |
/* ===========================================================================
|
|
|
1350 |
* Set the file type to ASCII or BINARY, using a crude approximation:
|
|
|
1351 |
* binary if more than 20% of the bytes are <= 6 or >= 128, ascii otherwise.
|
|
|
1352 |
* IN assertion: the fields freq of dyn_ltree are set and the total of all
|
|
|
1353 |
* frequencies does not exceed 64K (to fit in an int on 16 bit machines).
|
|
|
1354 |
*/
|
|
|
1355 |
void set_file_type(TState &state)
|
|
|
1356 |
{
|
|
|
1357 |
int n = 0;
|
|
|
1358 |
unsigned ascii_freq = 0;
|
|
|
1359 |
unsigned bin_freq = 0;
|
|
|
1360 |
while (n < 7) bin_freq += state.ts.dyn_ltree[n++].fc.freq;
|
|
|
1361 |
while (n < 128) ascii_freq += state.ts.dyn_ltree[n++].fc.freq;
|
|
|
1362 |
while (n < LITERALS) bin_freq += state.ts.dyn_ltree[n++].fc.freq;
|
|
|
1363 |
*state.ts.file_type = (ush)(bin_freq > (ascii_freq >> 2) ? BINARY : ASCII);
|
|
|
1364 |
}
|
|
|
1365 |
|
|
|
1366 |
|
|
|
1367 |
/* ===========================================================================
|
|
|
1368 |
* Initialize the bit string routines.
|
|
|
1369 |
*/
|
|
|
1370 |
void bi_init (TState &state,char *tgt_buf, unsigned tgt_size, int flsh_allowed)
|
|
|
1371 |
{
|
|
|
1372 |
state.bs.out_buf = tgt_buf;
|
|
|
1373 |
state.bs.out_size = tgt_size;
|
|
|
1374 |
state.bs.out_offset = 0;
|
|
|
1375 |
state.bs.flush_flg = flsh_allowed;
|
|
|
1376 |
|
|
|
1377 |
state.bs.bi_buf = 0;
|
|
|
1378 |
state.bs.bi_valid = 0;
|
|
|
1379 |
state.bs.bits_sent = 0L;
|
|
|
1380 |
}
|
|
|
1381 |
|
|
|
1382 |
/* ===========================================================================
|
|
|
1383 |
* Send a value on a given number of bits.
|
|
|
1384 |
* IN assertion: length <= 16 and value fits in length bits.
|
|
|
1385 |
*/
|
|
|
1386 |
void send_bits(TState &state,int value, int length)
|
|
|
1387 |
{
|
|
|
1388 |
Assert(state,length > 0 && length <= 15, "invalid length");
|
|
|
1389 |
state.bs.bits_sent += (ulg)length;
|
|
|
1390 |
/* If not enough room in bi_buf, use (bi_valid) bits from bi_buf and
|
|
|
1391 |
* (Buf_size - bi_valid) bits from value to flush the filled bi_buf,
|
|
|
1392 |
* then fill in the rest of (value), leaving (length - (Buf_size-bi_valid))
|
|
|
1393 |
* unused bits in bi_buf.
|
|
|
1394 |
*/
|
|
|
1395 |
state.bs.bi_buf |= (value << state.bs.bi_valid);
|
|
|
1396 |
state.bs.bi_valid += length;
|
|
|
1397 |
if (state.bs.bi_valid > (int)Buf_size) {
|
|
|
1398 |
PUTSHORT(state,state.bs.bi_buf);
|
|
|
1399 |
state.bs.bi_valid -= Buf_size;
|
|
|
1400 |
state.bs.bi_buf = (unsigned)value >> (length - state.bs.bi_valid);
|
|
|
1401 |
}
|
|
|
1402 |
}
|
|
|
1403 |
|
|
|
1404 |
/* ===========================================================================
|
|
|
1405 |
* Reverse the first len bits of a code, using straightforward code (a faster
|
|
|
1406 |
* method would use a table)
|
|
|
1407 |
* IN assertion: 1 <= len <= 15
|
|
|
1408 |
*/
|
|
|
1409 |
unsigned bi_reverse(unsigned code, int len)
|
|
|
1410 |
{
|
|
|
1411 |
register unsigned res = 0;
|
|
|
1412 |
do {
|
|
|
1413 |
res |= code & 1;
|
|
|
1414 |
code >>= 1, res <<= 1;
|
|
|
1415 |
} while (--len > 0);
|
|
|
1416 |
return res >> 1;
|
|
|
1417 |
}
|
|
|
1418 |
|
|
|
1419 |
/* ===========================================================================
|
|
|
1420 |
* Write out any remaining bits in an incomplete byte.
|
|
|
1421 |
*/
|
|
|
1422 |
void bi_windup(TState &state)
|
|
|
1423 |
{
|
|
|
1424 |
if (state.bs.bi_valid > 8) {
|
|
|
1425 |
PUTSHORT(state,state.bs.bi_buf);
|
|
|
1426 |
} else if (state.bs.bi_valid > 0) {
|
|
|
1427 |
PUTBYTE(state,state.bs.bi_buf);
|
|
|
1428 |
}
|
|
|
1429 |
if (state.bs.flush_flg) {
|
|
|
1430 |
state.flush_outbuf(state.param,state.bs.out_buf, &state.bs.out_offset);
|
|
|
1431 |
}
|
|
|
1432 |
state.bs.bi_buf = 0;
|
|
|
1433 |
state.bs.bi_valid = 0;
|
|
|
1434 |
state.bs.bits_sent = (state.bs.bits_sent+7) & ~7;
|
|
|
1435 |
}
|
|
|
1436 |
|
|
|
1437 |
/* ===========================================================================
|
|
|
1438 |
* Copy a stored block to the zip file, storing first the length and its
|
|
|
1439 |
* one's complement if requested.
|
|
|
1440 |
*/
|
|
|
1441 |
void copy_block(TState &state, char *block, unsigned len, int header)
|
|
|
1442 |
{
|
|
|
1443 |
bi_windup(state); /* align on byte boundary */
|
|
|
1444 |
|
|
|
1445 |
if (header) {
|
|
|
1446 |
PUTSHORT(state,(ush)len);
|
|
|
1447 |
PUTSHORT(state,(ush)~len);
|
|
|
1448 |
state.bs.bits_sent += 2*16;
|
|
|
1449 |
}
|
|
|
1450 |
if (state.bs.flush_flg) {
|
|
|
1451 |
state.flush_outbuf(state.param,state.bs.out_buf, &state.bs.out_offset);
|
|
|
1452 |
state.bs.out_offset = len;
|
|
|
1453 |
state.flush_outbuf(state.param,block, &state.bs.out_offset);
|
|
|
1454 |
} else if (state.bs.out_offset + len > state.bs.out_size) {
|
|
|
1455 |
Assert(state,false,"output buffer too small for in-memory compression");
|
|
|
1456 |
} else {
|
|
|
1457 |
memcpy(state.bs.out_buf + state.bs.out_offset, block, len);
|
|
|
1458 |
state.bs.out_offset += len;
|
|
|
1459 |
}
|
|
|
1460 |
state.bs.bits_sent += (ulg)len<<3;
|
|
|
1461 |
}
|
|
|
1462 |
|
|
|
1463 |
|
|
|
1464 |
|
|
|
1465 |
|
|
|
1466 |
|
|
|
1467 |
|
|
|
1468 |
|
|
|
1469 |
|
|
|
1470 |
/* ===========================================================================
|
|
|
1471 |
* Prototypes for functions.
|
|
|
1472 |
*/
|
|
|
1473 |
|
|
|
1474 |
void fill_window (TState &state);
|
|
|
1475 |
ulg deflate_fast (TState &state);
|
|
|
1476 |
|
|
|
1477 |
int longest_match (TState &state,IPos cur_match);
|
|
|
1478 |
|
|
|
1479 |
|
|
|
1480 |
/* ===========================================================================
|
|
|
1481 |
* Update a hash value with the given input byte
|
|
|
1482 |
* IN assertion: all calls to to UPDATE_HASH are made with consecutive
|
|
|
1483 |
* input characters, so that a running hash key can be computed from the
|
|
|
1484 |
* previous key instead of complete recalculation each time.
|
|
|
1485 |
*/
|
|
|
1486 |
#define UPDATE_HASH(h,c) (h = (((h)<<H_SHIFT) ^ (c)) & HASH_MASK)
|
|
|
1487 |
|
|
|
1488 |
/* ===========================================================================
|
|
|
1489 |
* Insert string s in the dictionary and set match_head to the previous head
|
|
|
1490 |
* of the hash chain (the most recent string with same hash key). Return
|
|
|
1491 |
* the previous length of the hash chain.
|
|
|
1492 |
* IN assertion: all calls to to INSERT_STRING are made with consecutive
|
|
|
1493 |
* input characters and the first MIN_MATCH bytes of s are valid
|
|
|
1494 |
* (except for the last MIN_MATCH-1 bytes of the input file).
|
|
|
1495 |
*/
|
|
|
1496 |
#define INSERT_STRING(s, match_head) \
|
|
|
1497 |
(UPDATE_HASH(state.ds.ins_h, state.ds.window[(s) + (MIN_MATCH-1)]), \
|
|
|
1498 |
state.ds.prev[(s) & WMASK] = match_head = state.ds.head[state.ds.ins_h], \
|
|
|
1499 |
state.ds.head[state.ds.ins_h] = (s))
|
|
|
1500 |
|
|
|
1501 |
/* ===========================================================================
|
|
|
1502 |
* Initialize the "longest match" routines for a new file
|
|
|
1503 |
*
|
|
|
1504 |
* IN assertion: window_size is > 0 if the input file is already read or
|
|
|
1505 |
* mmap'ed in the window[] array, 0 otherwise. In the first case,
|
|
|
1506 |
* window_size is sufficient to contain the whole input file plus
|
|
|
1507 |
* MIN_LOOKAHEAD bytes (to avoid referencing memory beyond the end
|
|
|
1508 |
* of window[] when looking for matches towards the end).
|
|
|
1509 |
*/
|
|
|
1510 |
void lm_init (TState &state, int pack_level, ush *flags)
|
|
|
1511 |
{
|
|
|
1512 |
register unsigned j;
|
|
|
1513 |
|
|
|
1514 |
Assert(state,pack_level>=1 && pack_level<=8,"bad pack level");
|
|
|
1515 |
|
|
|
1516 |
/* Do not slide the window if the whole input is already in memory
|
|
|
1517 |
* (window_size > 0)
|
|
|
1518 |
*/
|
|
|
1519 |
state.ds.sliding = 0;
|
|
|
1520 |
if (state.ds.window_size == 0L) {
|
|
|
1521 |
state.ds.sliding = 1;
|
|
|
1522 |
state.ds.window_size = (ulg)2L*WSIZE;
|
|
|
1523 |
}
|
|
|
1524 |
|
|
|
1525 |
/* Initialize the hash table (avoiding 64K overflow for 16 bit systems).
|
|
|
1526 |
* prev[] will be initialized on the fly.
|
|
|
1527 |
*/
|
|
|
1528 |
state.ds.head[HASH_SIZE-1] = NIL;
|
|
|
1529 |
memset((char*)state.ds.head, NIL, (unsigned)(HASH_SIZE-1)*sizeof(*state.ds.head));
|
|
|
1530 |
|
|
|
1531 |
/* Set the default configuration parameters:
|
|
|
1532 |
*/
|
|
|
1533 |
state.ds.max_lazy_match = configuration_table[pack_level].max_lazy;
|
|
|
1534 |
state.ds.good_match = configuration_table[pack_level].good_length;
|
|
|
1535 |
state.ds.nice_match = configuration_table[pack_level].nice_length;
|
|
|
1536 |
state.ds.max_chain_length = configuration_table[pack_level].max_chain;
|
|
|
1537 |
if (pack_level <= 2) {
|
|
|
1538 |
*flags |= FAST;
|
|
|
1539 |
} else if (pack_level >= 8) {
|
|
|
1540 |
*flags |= SLOW;
|
|
|
1541 |
}
|
|
|
1542 |
/* ??? reduce max_chain_length for binary files */
|
|
|
1543 |
|
|
|
1544 |
state.ds.strstart = 0;
|
|
|
1545 |
state.ds.block_start = 0L;
|
|
|
1546 |
|
|
|
1547 |
j = WSIZE;
|
|
|
1548 |
j <<= 1; // Can read 64K in one step
|
|
|
1549 |
state.ds.lookahead = state.readfunc(state, (char*)state.ds.window, j);
|
|
|
1550 |
|
|
|
1551 |
if (state.ds.lookahead == 0 || state.ds.lookahead == (unsigned)EOF) {
|
|
|
1552 |
state.ds.eofile = 1, state.ds.lookahead = 0;
|
|
|
1553 |
return;
|
|
|
1554 |
}
|
|
|
1555 |
state.ds.eofile = 0;
|
|
|
1556 |
/* Make sure that we always have enough lookahead. This is important
|
|
|
1557 |
* if input comes from a device such as a tty.
|
|
|
1558 |
*/
|
|
|
1559 |
if (state.ds.lookahead < MIN_LOOKAHEAD) fill_window(state);
|
|
|
1560 |
|
|
|
1561 |
state.ds.ins_h = 0;
|
|
|
1562 |
for (j=0; j<MIN_MATCH-1; j++) UPDATE_HASH(state.ds.ins_h, state.ds.window[j]);
|
|
|
1563 |
/* If lookahead < MIN_MATCH, ins_h is garbage, but this is
|
|
|
1564 |
* not important since only literal bytes will be emitted.
|
|
|
1565 |
*/
|
|
|
1566 |
}
|
|
|
1567 |
|
|
|
1568 |
|
|
|
1569 |
/* ===========================================================================
|
|
|
1570 |
* Set match_start to the longest match starting at the given string and
|
|
|
1571 |
* return its length. Matches shorter or equal to prev_length are discarded,
|
|
|
1572 |
* in which case the result is equal to prev_length and match_start is
|
|
|
1573 |
* garbage.
|
|
|
1574 |
* IN assertions: cur_match is the head of the hash chain for the current
|
|
|
1575 |
* string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1
|
|
|
1576 |
*/
|
|
|
1577 |
// For 80x86 and 680x0 and ARM, an optimized version is in match.asm or
|
|
|
1578 |
// match.S. The code is functionally equivalent, so you can use the C version
|
|
|
1579 |
// if desired. Which I do so desire!
|
|
|
1580 |
int longest_match(TState &state,IPos cur_match)
|
|
|
1581 |
{
|
|
|
1582 |
unsigned chain_length = state.ds.max_chain_length; /* max hash chain length */
|
|
|
1583 |
register uch far *scan = state.ds.window + state.ds.strstart; /* current string */
|
|
|
1584 |
register uch far *match; /* matched string */
|
|
|
1585 |
register int len; /* length of current match */
|
|
|
1586 |
int best_len = state.ds.prev_length; /* best match length so far */
|
|
|
1587 |
IPos limit = state.ds.strstart > (IPos)MAX_DIST ? state.ds.strstart - (IPos)MAX_DIST : NIL;
|
|
|
1588 |
/* Stop when cur_match becomes <= limit. To simplify the code,
|
|
|
1589 |
* we prevent matches with the string of window index 0.
|
|
|
1590 |
*/
|
|
|
1591 |
|
|
|
1592 |
// The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
|
|
|
1593 |
// It is easy to get rid of this optimization if necessary.
|
|
|
1594 |
Assert(state,HASH_BITS>=8 && MAX_MATCH==258,"Code too clever");
|
|
|
1595 |
|
|
|
1596 |
|
|
|
1597 |
|
|
|
1598 |
register uch far *strend = state.ds.window + state.ds.strstart + MAX_MATCH;
|
|
|
1599 |
register uch scan_end1 = scan[best_len-1];
|
|
|
1600 |
register uch scan_end = scan[best_len];
|
|
|
1601 |
|
|
|
1602 |
/* Do not waste too much time if we already have a good match: */
|
|
|
1603 |
if (state.ds.prev_length >= state.ds.good_match) {
|
|
|
1604 |
chain_length >>= 2;
|
|
|
1605 |
}
|
|
|
1606 |
|
|
|
1607 |
Assert(state,state.ds.strstart <= state.ds.window_size-MIN_LOOKAHEAD, "insufficient lookahead");
|
|
|
1608 |
|
|
|
1609 |
do {
|
|
|
1610 |
Assert(state,cur_match < state.ds.strstart, "no future");
|
|
|
1611 |
match = state.ds.window + cur_match;
|
|
|
1612 |
|
|
|
1613 |
/* Skip to next match if the match length cannot increase
|
|
|
1614 |
* or if the match length is less than 2:
|
|
|
1615 |
*/
|
|
|
1616 |
if (match[best_len] != scan_end ||
|
|
|
1617 |
match[best_len-1] != scan_end1 ||
|
|
|
1618 |
*match != *scan ||
|
|
|
1619 |
*++match != scan[1]) continue;
|
|
|
1620 |
|
|
|
1621 |
/* The check at best_len-1 can be removed because it will be made
|
|
|
1622 |
* again later. (This heuristic is not always a win.)
|
|
|
1623 |
* It is not necessary to compare scan[2] and match[2] since they
|
|
|
1624 |
* are always equal when the other bytes match, given that
|
|
|
1625 |
* the hash keys are equal and that HASH_BITS >= 8.
|
|
|
1626 |
*/
|
|
|
1627 |
scan += 2, match++;
|
|
|
1628 |
|
|
|
1629 |
/* We check for insufficient lookahead only every 8th comparison;
|
|
|
1630 |
* the 256th check will be made at strstart+258.
|
|
|
1631 |
*/
|
|
|
1632 |
do {
|
|
|
1633 |
} while (*++scan == *++match && *++scan == *++match &&
|
|
|
1634 |
*++scan == *++match && *++scan == *++match &&
|
|
|
1635 |
*++scan == *++match && *++scan == *++match &&
|
|
|
1636 |
*++scan == *++match && *++scan == *++match &&
|
|
|
1637 |
scan < strend);
|
|
|
1638 |
|
|
|
1639 |
Assert(state,scan <= state.ds.window+(unsigned)(state.ds.window_size-1), "wild scan");
|
|
|
1640 |
|
|
|
1641 |
len = MAX_MATCH - (int)(strend - scan);
|
|
|
1642 |
scan = strend - MAX_MATCH;
|
|
|
1643 |
|
|
|
1644 |
|
|
|
1645 |
if (len > best_len) {
|
|
|
1646 |
state.ds.match_start = cur_match;
|
|
|
1647 |
best_len = len;
|
|
|
1648 |
if (len >= state.ds.nice_match) break;
|
|
|
1649 |
scan_end1 = scan[best_len-1];
|
|
|
1650 |
scan_end = scan[best_len];
|
|
|
1651 |
}
|
|
|
1652 |
} while ((cur_match = state.ds.prev[cur_match & WMASK]) > limit
|
|
|
1653 |
&& --chain_length != 0);
|
|
|
1654 |
|
|
|
1655 |
return best_len;
|
|
|
1656 |
}
|
|
|
1657 |
|
|
|
1658 |
|
|
|
1659 |
|
|
|
1660 |
#define check_match(state,start, match, length)
|
|
|
1661 |
// or alternatively...
|
|
|
1662 |
//void check_match(TState &state,IPos start, IPos match, int length)
|
|
|
1663 |
//{ // check that the match is indeed a match
|
|
|
1664 |
// if (memcmp((char*)state.ds.window + match,
|
|
|
1665 |
// (char*)state.ds.window + start, length) != EQUAL) {
|
|
|
1666 |
// fprintf(stderr,
|
|
|
1667 |
// " start %d, match %d, length %d\n",
|
|
|
1668 |
// start, match, length);
|
|
|
1669 |
// error("invalid match");
|
|
|
1670 |
// }
|
|
|
1671 |
// if (state.verbose > 1) {
|
|
|
1672 |
// fprintf(stderr,"\\[%d,%d]", start-match, length);
|
|
|
1673 |
// do { fprintf(stdout,"%c",state.ds.window[start++]); } while (--length != 0);
|
|
|
1674 |
// }
|
|
|
1675 |
//}
|
|
|
1676 |
|
|
|
1677 |
/* ===========================================================================
|
|
|
1678 |
* Fill the window when the lookahead becomes insufficient.
|
|
|
1679 |
* Updates strstart and lookahead, and sets eofile if end of input file.
|
|
|
1680 |
*
|
|
|
1681 |
* IN assertion: lookahead < MIN_LOOKAHEAD && strstart + lookahead > 0
|
|
|
1682 |
* OUT assertions: strstart <= window_size-MIN_LOOKAHEAD
|
|
|
1683 |
* At least one byte has been read, or eofile is set; file reads are
|
|
|
1684 |
* performed for at least two bytes (required for the translate_eol option).
|
|
|
1685 |
*/
|
|
|
1686 |
void fill_window(TState &state)
|
|
|
1687 |
{
|
|
|
1688 |
register unsigned n, m;
|
|
|
1689 |
unsigned more; /* Amount of free space at the end of the window. */
|
|
|
1690 |
|
|
|
1691 |
do {
|
|
|
1692 |
more = (unsigned)(state.ds.window_size - (ulg)state.ds.lookahead - (ulg)state.ds.strstart);
|
|
|
1693 |
|
|
|
1694 |
/* If the window is almost full and there is insufficient lookahead,
|
|
|
1695 |
* move the upper half to the lower one to make room in the upper half.
|
|
|
1696 |
*/
|
|
|
1697 |
if (more == (unsigned)EOF) {
|
|
|
1698 |
/* Very unlikely, but possible on 16 bit machine if strstart == 0
|
|
|
1699 |
* and lookahead == 1 (input done one byte at time)
|
|
|
1700 |
*/
|
|
|
1701 |
more--;
|
|
|
1702 |
|
|
|
1703 |
/* For MMAP or BIG_MEM, the whole input file is already in memory so
|
|
|
1704 |
* we must not perform sliding. We must however call (*read_buf)() in
|
|
|
1705 |
* order to compute the crc, update lookahead and possibly set eofile.
|
|
|
1706 |
*/
|
|
|
1707 |
} else if (state.ds.strstart >= WSIZE+MAX_DIST && state.ds.sliding) {
|
|
|
1708 |
|
|
|
1709 |
/* By the IN assertion, the window is not empty so we can't confuse
|
|
|
1710 |
* more == 0 with more == 64K on a 16 bit machine.
|
|
|
1711 |
*/
|
|
|
1712 |
memcpy((char*)state.ds.window, (char*)state.ds.window+WSIZE, (unsigned)WSIZE);
|
|
|
1713 |
state.ds.match_start -= WSIZE;
|
|
|
1714 |
state.ds.strstart -= WSIZE; /* we now have strstart >= MAX_DIST: */
|
|
|
1715 |
|
|
|
1716 |
state.ds.block_start -= (long) WSIZE;
|
|
|
1717 |
|
|
|
1718 |
for (n = 0; n < HASH_SIZE; n++) {
|
|
|
1719 |
m = state.ds.head[n];
|
|
|
1720 |
state.ds.head[n] = (Pos)(m >= WSIZE ? m-WSIZE : NIL);
|
|
|
1721 |
}
|
|
|
1722 |
for (n = 0; n < WSIZE; n++) {
|
|
|
1723 |
m = state.ds.prev[n];
|
|
|
1724 |
state.ds.prev[n] = (Pos)(m >= WSIZE ? m-WSIZE : NIL);
|
|
|
1725 |
/* If n is not on any hash chain, prev[n] is garbage but
|
|
|
1726 |
* its value will never be used.
|
|
|
1727 |
*/
|
|
|
1728 |
}
|
|
|
1729 |
more += WSIZE;
|
|
|
1730 |
}
|
|
|
1731 |
if (state.ds.eofile) return;
|
|
|
1732 |
|
|
|
1733 |
/* If there was no sliding:
|
|
|
1734 |
* strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 &&
|
|
|
1735 |
* more == window_size - lookahead - strstart
|
|
|
1736 |
* => more >= window_size - (MIN_LOOKAHEAD-1 + WSIZE + MAX_DIST-1)
|
|
|
1737 |
* => more >= window_size - 2*WSIZE + 2
|
|
|
1738 |
* In the MMAP or BIG_MEM case (not yet supported in gzip),
|
|
|
1739 |
* window_size == input_size + MIN_LOOKAHEAD &&
|
|
|
1740 |
* strstart + lookahead <= input_size => more >= MIN_LOOKAHEAD.
|
|
|
1741 |
* Otherwise, window_size == 2*WSIZE so more >= 2.
|
|
|
1742 |
* If there was sliding, more >= WSIZE. So in all cases, more >= 2.
|
|
|
1743 |
*/
|
|
|
1744 |
Assert(state,more >= 2, "more < 2");
|
|
|
1745 |
|
|
|
1746 |
n = state.readfunc(state, (char*)state.ds.window+state.ds.strstart+state.ds.lookahead, more);
|
|
|
1747 |
|
|
|
1748 |
if (n == 0 || n == (unsigned)EOF) {
|
|
|
1749 |
state.ds.eofile = 1;
|
|
|
1750 |
} else {
|
|
|
1751 |
state.ds.lookahead += n;
|
|
|
1752 |
}
|
|
|
1753 |
} while (state.ds.lookahead < MIN_LOOKAHEAD && !state.ds.eofile);
|
|
|
1754 |
}
|
|
|
1755 |
|
|
|
1756 |
/* ===========================================================================
|
|
|
1757 |
* Flush the current block, with given end-of-file flag.
|
|
|
1758 |
* IN assertion: strstart is set to the end of the current match.
|
|
|
1759 |
*/
|
|
|
1760 |
#define FLUSH_BLOCK(state,eof) \
|
|
|
1761 |
flush_block(state,state.ds.block_start >= 0L ? (char*)&state.ds.window[(unsigned)state.ds.block_start] : \
|
|
|
1762 |
(char*)NULL, (long)state.ds.strstart - state.ds.block_start, (eof))
|
|
|
1763 |
|
|
|
1764 |
/* ===========================================================================
|
|
|
1765 |
* Processes a new input file and return its compressed length. This
|
|
|
1766 |
* function does not perform lazy evaluation of matches and inserts
|
|
|
1767 |
* new strings in the dictionary only for unmatched strings or for short
|
|
|
1768 |
* matches. It is used only for the fast compression options.
|
|
|
1769 |
*/
|
|
|
1770 |
ulg deflate_fast(TState &state)
|
|
|
1771 |
{
|
|
|
1772 |
IPos hash_head = NIL; /* head of the hash chain */
|
|
|
1773 |
int flush; /* set if current block must be flushed */
|
|
|
1774 |
unsigned match_length = 0; /* length of best match */
|
|
|
1775 |
|
|
|
1776 |
state.ds.prev_length = MIN_MATCH-1;
|
|
|
1777 |
while (state.ds.lookahead != 0) {
|
|
|
1778 |
/* Insert the string window[strstart .. strstart+2] in the
|
|
|
1779 |
* dictionary, and set hash_head to the head of the hash chain:
|
|
|
1780 |
*/
|
|
|
1781 |
if (state.ds.lookahead >= MIN_MATCH)
|
|
|
1782 |
INSERT_STRING(state.ds.strstart, hash_head);
|
|
|
1783 |
|
|
|
1784 |
/* Find the longest match, discarding those <= prev_length.
|
|
|
1785 |
* At this point we have always match_length < MIN_MATCH
|
|
|
1786 |
*/
|
|
|
1787 |
if (hash_head != NIL && state.ds.strstart - hash_head <= MAX_DIST) {
|
|
|
1788 |
/* To simplify the code, we prevent matches with the string
|
|
|
1789 |
* of window index 0 (in particular we have to avoid a match
|
|
|
1790 |
* of the string with itself at the start of the input file).
|
|
|
1791 |
*/
|
|
|
1792 |
/* Do not look for matches beyond the end of the input.
|
|
|
1793 |
* This is necessary to make deflate deterministic.
|
|
|
1794 |
*/
|
|
|
1795 |
if ((unsigned)state.ds.nice_match > state.ds.lookahead) state.ds.nice_match = (int)state.ds.lookahead;
|
|
|
1796 |
match_length = longest_match (state,hash_head);
|
|
|
1797 |
/* longest_match() sets match_start */
|
|
|
1798 |
if (match_length > state.ds.lookahead) match_length = state.ds.lookahead;
|
|
|
1799 |
}
|
|
|
1800 |
if (match_length >= MIN_MATCH) {
|
|
|
1801 |
check_match(state,state.ds.strstart, state.ds.match_start, match_length);
|
|
|
1802 |
|
|
|
1803 |
flush = ct_tally(state,state.ds.strstart-state.ds.match_start, match_length - MIN_MATCH);
|
|
|
1804 |
|
|
|
1805 |
state.ds.lookahead -= match_length;
|
|
|
1806 |
|
|
|
1807 |
/* Insert new strings in the hash table only if the match length
|
|
|
1808 |
* is not too large. This saves time but degrades compression.
|
|
|
1809 |
*/
|
|
|
1810 |
if (match_length <= state.ds.max_insert_length
|
|
|
1811 |
&& state.ds.lookahead >= MIN_MATCH) {
|
|
|
1812 |
match_length--; /* string at strstart already in hash table */
|
|
|
1813 |
do {
|
|
|
1814 |
state.ds.strstart++;
|
|
|
1815 |
INSERT_STRING(state.ds.strstart, hash_head);
|
|
|
1816 |
/* strstart never exceeds WSIZE-MAX_MATCH, so there are
|
|
|
1817 |
* always MIN_MATCH bytes ahead.
|
|
|
1818 |
*/
|
|
|
1819 |
} while (--match_length != 0);
|
|
|
1820 |
state.ds.strstart++;
|
|
|
1821 |
} else {
|
|
|
1822 |
state.ds.strstart += match_length;
|
|
|
1823 |
match_length = 0;
|
|
|
1824 |
state.ds.ins_h = state.ds.window[state.ds.strstart];
|
|
|
1825 |
UPDATE_HASH(state.ds.ins_h, state.ds.window[state.ds.strstart+1]);
|
|
|
1826 |
Assert(state,MIN_MATCH==3,"Call UPDATE_HASH() MIN_MATCH-3 more times");
|
|
|
1827 |
}
|
|
|
1828 |
} else {
|
|
|
1829 |
/* No match, output a literal byte */
|
|
|
1830 |
flush = ct_tally (state,0, state.ds.window[state.ds.strstart]);
|
|
|
1831 |
state.ds.lookahead--;
|
|
|
1832 |
state.ds.strstart++;
|
|
|
1833 |
}
|
|
|
1834 |
if (flush) FLUSH_BLOCK(state,0), state.ds.block_start = state.ds.strstart;
|
|
|
1835 |
|
|
|
1836 |
/* Make sure that we always have enough lookahead, except
|
|
|
1837 |
* at the end of the input file. We need MAX_MATCH bytes
|
|
|
1838 |
* for the next match, plus MIN_MATCH bytes to insert the
|
|
|
1839 |
* string following the next match.
|
|
|
1840 |
*/
|
|
|
1841 |
if (state.ds.lookahead < MIN_LOOKAHEAD) fill_window(state);
|
|
|
1842 |
}
|
|
|
1843 |
return FLUSH_BLOCK(state,1); /* eof */
|
|
|
1844 |
}
|
|
|
1845 |
|
|
|
1846 |
/* ===========================================================================
|
|
|
1847 |
* Same as above, but achieves better compression. We use a lazy
|
|
|
1848 |
* evaluation for matches: a match is finally adopted only if there is
|
|
|
1849 |
* no better match at the next window position.
|
|
|
1850 |
*/
|
|
|
1851 |
ulg deflate(TState &state)
|
|
|
1852 |
{
|
|
|
1853 |
IPos hash_head = NIL; /* head of hash chain */
|
|
|
1854 |
IPos prev_match; /* previous match */
|
|
|
1855 |
int flush; /* set if current block must be flushed */
|
|
|
1856 |
int match_available = 0; /* set if previous match exists */
|
|
|
1857 |
register unsigned match_length = MIN_MATCH-1; /* length of best match */
|
|
|
1858 |
|
|
|
1859 |
if (state.level <= 3) return deflate_fast(state); /* optimized for speed */
|
|
|
1860 |
|
|
|
1861 |
/* Process the input block. */
|
|
|
1862 |
while (state.ds.lookahead != 0) {
|
|
|
1863 |
/* Insert the string window[strstart .. strstart+2] in the
|
|
|
1864 |
* dictionary, and set hash_head to the head of the hash chain:
|
|
|
1865 |
*/
|
|
|
1866 |
if (state.ds.lookahead >= MIN_MATCH)
|
|
|
1867 |
INSERT_STRING(state.ds.strstart, hash_head);
|
|
|
1868 |
|
|
|
1869 |
/* Find the longest match, discarding those <= prev_length.
|
|
|
1870 |
*/
|
|
|
1871 |
state.ds.prev_length = match_length, prev_match = state.ds.match_start;
|
|
|
1872 |
match_length = MIN_MATCH-1;
|
|
|
1873 |
|
|
|
1874 |
if (hash_head != NIL && state.ds.prev_length < state.ds.max_lazy_match &&
|
|
|
1875 |
state.ds.strstart - hash_head <= MAX_DIST) {
|
|
|
1876 |
/* To simplify the code, we prevent matches with the string
|
|
|
1877 |
* of window index 0 (in particular we have to avoid a match
|
|
|
1878 |
* of the string with itself at the start of the input file).
|
|
|
1879 |
*/
|
|
|
1880 |
/* Do not look for matches beyond the end of the input.
|
|
|
1881 |
* This is necessary to make deflate deterministic.
|
|
|
1882 |
*/
|
|
|
1883 |
if ((unsigned)state.ds.nice_match > state.ds.lookahead) state.ds.nice_match = (int)state.ds.lookahead;
|
|
|
1884 |
match_length = longest_match (state,hash_head);
|
|
|
1885 |
/* longest_match() sets match_start */
|
|
|
1886 |
if (match_length > state.ds.lookahead) match_length = state.ds.lookahead;
|
|
|
1887 |
|
|
|
1888 |
/* Ignore a length 3 match if it is too distant: */
|
|
|
1889 |
if (match_length == MIN_MATCH && state.ds.strstart-state.ds.match_start > TOO_FAR){
|
|
|
1890 |
/* If prev_match is also MIN_MATCH, match_start is garbage
|
|
|
1891 |
* but we will ignore the current match anyway.
|
|
|
1892 |
*/
|
|
|
1893 |
match_length = MIN_MATCH-1;
|
|
|
1894 |
}
|
|
|
1895 |
}
|
|
|
1896 |
/* If there was a match at the previous step and the current
|
|
|
1897 |
* match is not better, output the previous match:
|
|
|
1898 |
*/
|
|
|
1899 |
if (state.ds.prev_length >= MIN_MATCH && match_length <= state.ds.prev_length) {
|
|
|
1900 |
unsigned max_insert = state.ds.strstart + state.ds.lookahead - MIN_MATCH;
|
|
|
1901 |
check_match(state,state.ds.strstart-1, prev_match, state.ds.prev_length);
|
|
|
1902 |
flush = ct_tally(state,state.ds.strstart-1-prev_match, state.ds.prev_length - MIN_MATCH);
|
|
|
1903 |
|
|
|
1904 |
/* Insert in hash table all strings up to the end of the match.
|
|
|
1905 |
* strstart-1 and strstart are already inserted.
|
|
|
1906 |
*/
|
|
|
1907 |
state.ds.lookahead -= state.ds.prev_length-1;
|
|
|
1908 |
state.ds.prev_length -= 2;
|
|
|
1909 |
do {
|
|
|
1910 |
if (++state.ds.strstart <= max_insert) {
|
|
|
1911 |
INSERT_STRING(state.ds.strstart, hash_head);
|
|
|
1912 |
/* strstart never exceeds WSIZE-MAX_MATCH, so there are
|
|
|
1913 |
* always MIN_MATCH bytes ahead.
|
|
|
1914 |
*/
|
|
|
1915 |
}
|
|
|
1916 |
} while (--state.ds.prev_length != 0);
|
|
|
1917 |
state.ds.strstart++;
|
|
|
1918 |
match_available = 0;
|
|
|
1919 |
match_length = MIN_MATCH-1;
|
|
|
1920 |
|
|
|
1921 |
if (flush) FLUSH_BLOCK(state,0), state.ds.block_start = state.ds.strstart;
|
|
|
1922 |
|
|
|
1923 |
} else if (match_available) {
|
|
|
1924 |
/* If there was no match at the previous position, output a
|
|
|
1925 |
* single literal. If there was a match but the current match
|
|
|
1926 |
* is longer, truncate the previous match to a single literal.
|
|
|
1927 |
*/
|
|
|
1928 |
if (ct_tally (state,0, state.ds.window[state.ds.strstart-1])) {
|
|
|
1929 |
FLUSH_BLOCK(state,0), state.ds.block_start = state.ds.strstart;
|
|
|
1930 |
}
|
|
|
1931 |
state.ds.strstart++;
|
|
|
1932 |
state.ds.lookahead--;
|
|
|
1933 |
} else {
|
|
|
1934 |
/* There is no previous match to compare with, wait for
|
|
|
1935 |
* the next step to decide.
|
|
|
1936 |
*/
|
|
|
1937 |
match_available = 1;
|
|
|
1938 |
state.ds.strstart++;
|
|
|
1939 |
state.ds.lookahead--;
|
|
|
1940 |
}
|
|
|
1941 |
// Assert(state,strstart <= isize && lookahead <= isize, "a bit too far");
|
|
|
1942 |
|
|
|
1943 |
/* Make sure that we always have enough lookahead, except
|
|
|
1944 |
* at the end of the input file. We need MAX_MATCH bytes
|
|
|
1945 |
* for the next match, plus MIN_MATCH bytes to insert the
|
|
|
1946 |
* string following the next match.
|
|
|
1947 |
*/
|
|
|
1948 |
if (state.ds.lookahead < MIN_LOOKAHEAD) fill_window(state);
|
|
|
1949 |
}
|
|
|
1950 |
if (match_available) ct_tally (state,0, state.ds.window[state.ds.strstart-1]);
|
|
|
1951 |
|
|
|
1952 |
return FLUSH_BLOCK(state,1); /* eof */
|
|
|
1953 |
}
|
|
|
1954 |
|
|
|
1955 |
|
|
|
1956 |
|
|
|
1957 |
|
|
|
1958 |
|
|
|
1959 |
|
|
|
1960 |
|
|
|
1961 |
|
|
|
1962 |
|
|
|
1963 |
|
|
|
1964 |
|
|
|
1965 |
|
|
|
1966 |
int putlocal(struct zlist far *z, WRITEFUNC wfunc,void *param)
|
|
|
1967 |
{ // Write a local header described by *z to file *f. Return a ZE_ error code.
|
|
|
1968 |
PUTLG(LOCSIG, f);
|
|
|
1969 |
PUTSH(z->ver, f);
|
|
|
1970 |
PUTSH(z->lflg, f);
|
|
|
1971 |
PUTSH(z->how, f);
|
|
|
1972 |
PUTLG(z->tim, f);
|
|
|
1973 |
PUTLG(z->crc, f);
|
|
|
1974 |
PUTLG(z->siz, f);
|
|
|
1975 |
PUTLG(z->len, f);
|
|
|
1976 |
PUTSH(z->nam, f);
|
|
|
1977 |
PUTSH(z->ext, f);
|
|
|
1978 |
size_t res = (size_t)wfunc(param, z->iname, (unsigned int)z->nam);
|
|
|
1979 |
if (res!=z->nam) return ZE_TEMP;
|
|
|
1980 |
if (z->ext)
|
|
|
1981 |
{ res = (size_t)wfunc(param, z->extra, (unsigned int)z->ext);
|
|
|
1982 |
if (res!=z->ext) return ZE_TEMP;
|
|
|
1983 |
}
|
|
|
1984 |
return ZE_OK;
|
|
|
1985 |
}
|
|
|
1986 |
|
|
|
1987 |
int putextended(struct zlist far *z, WRITEFUNC wfunc, void *param)
|
|
|
1988 |
{ // Write an extended local header described by *z to file *f. Returns a ZE_ code
|
|
|
1989 |
PUTLG(EXTLOCSIG, f);
|
|
|
1990 |
PUTLG(z->crc, f);
|
|
|
1991 |
PUTLG(z->siz, f);
|
|
|
1992 |
PUTLG(z->len, f);
|
|
|
1993 |
return ZE_OK;
|
|
|
1994 |
}
|
|
|
1995 |
|
|
|
1996 |
int putcentral(struct zlist far *z, WRITEFUNC wfunc, void *param)
|
|
|
1997 |
{ // Write a central header entry of *z to file *f. Returns a ZE_ code.
|
|
|
1998 |
PUTLG(CENSIG, f);
|
|
|
1999 |
PUTSH(z->vem, f);
|
|
|
2000 |
PUTSH(z->ver, f);
|
|
|
2001 |
PUTSH(z->flg, f);
|
|
|
2002 |
PUTSH(z->how, f);
|
|
|
2003 |
PUTLG(z->tim, f);
|
|
|
2004 |
PUTLG(z->crc, f);
|
|
|
2005 |
PUTLG(z->siz, f);
|
|
|
2006 |
PUTLG(z->len, f);
|
|
|
2007 |
PUTSH(z->nam, f);
|
|
|
2008 |
PUTSH(z->cext, f);
|
|
|
2009 |
PUTSH(z->com, f);
|
|
|
2010 |
PUTSH(z->dsk, f);
|
|
|
2011 |
PUTSH(z->att, f);
|
|
|
2012 |
PUTLG(z->atx, f);
|
|
|
2013 |
PUTLG(z->off, f);
|
|
|
2014 |
if ((size_t)wfunc(param, z->iname, (unsigned int)z->nam) != z->nam ||
|
|
|
2015 |
(z->cext && (size_t)wfunc(param, z->cextra, (unsigned int)z->cext) != z->cext) ||
|
|
|
2016 |
(z->com && (size_t)wfunc(param, z->comment, (unsigned int)z->com) != z->com))
|
|
|
2017 |
return ZE_TEMP;
|
|
|
2018 |
return ZE_OK;
|
|
|
2019 |
}
|
|
|
2020 |
|
|
|
2021 |
|
|
|
2022 |
int putend(int n, ulg s, ulg c, extent m, char *z, WRITEFUNC wfunc, void *param)
|
|
|
2023 |
{ // write the end of the central-directory-data to file *f.
|
|
|
2024 |
PUTLG(ENDSIG, f);
|
|
|
2025 |
PUTSH(0, f);
|
|
|
2026 |
PUTSH(0, f);
|
|
|
2027 |
PUTSH(n, f);
|
|
|
2028 |
PUTSH(n, f);
|
|
|
2029 |
PUTLG(s, f);
|
|
|
2030 |
PUTLG(c, f);
|
|
|
2031 |
PUTSH(m, f);
|
|
|
2032 |
// Write the comment, if any
|
|
|
2033 |
if (m && wfunc(param, z, (unsigned int)m) != m) return ZE_TEMP;
|
|
|
2034 |
return ZE_OK;
|
|
|
2035 |
}
|
|
|
2036 |
|
|
|
2037 |
|
|
|
2038 |
|
|
|
2039 |
|
|
|
2040 |
|
|
|
2041 |
|
|
|
2042 |
const ulg crc_table[256] = {
|
|
|
2043 |
0x00000000L, 0x77073096L, 0xee0e612cL, 0x990951baL, 0x076dc419L,
|
|
|
2044 |
0x706af48fL, 0xe963a535L, 0x9e6495a3L, 0x0edb8832L, 0x79dcb8a4L,
|
|
|
2045 |
0xe0d5e91eL, 0x97d2d988L, 0x09b64c2bL, 0x7eb17cbdL, 0xe7b82d07L,
|
|
|
2046 |
0x90bf1d91L, 0x1db71064L, 0x6ab020f2L, 0xf3b97148L, 0x84be41deL,
|
|
|
2047 |
0x1adad47dL, 0x6ddde4ebL, 0xf4d4b551L, 0x83d385c7L, 0x136c9856L,
|
|
|
2048 |
0x646ba8c0L, 0xfd62f97aL, 0x8a65c9ecL, 0x14015c4fL, 0x63066cd9L,
|
|
|
2049 |
0xfa0f3d63L, 0x8d080df5L, 0x3b6e20c8L, 0x4c69105eL, 0xd56041e4L,
|
|
|
2050 |
0xa2677172L, 0x3c03e4d1L, 0x4b04d447L, 0xd20d85fdL, 0xa50ab56bL,
|
|
|
2051 |
0x35b5a8faL, 0x42b2986cL, 0xdbbbc9d6L, 0xacbcf940L, 0x32d86ce3L,
|
|
|
2052 |
0x45df5c75L, 0xdcd60dcfL, 0xabd13d59L, 0x26d930acL, 0x51de003aL,
|
|
|
2053 |
0xc8d75180L, 0xbfd06116L, 0x21b4f4b5L, 0x56b3c423L, 0xcfba9599L,
|
|
|
2054 |
0xb8bda50fL, 0x2802b89eL, 0x5f058808L, 0xc60cd9b2L, 0xb10be924L,
|
|
|
2055 |
0x2f6f7c87L, 0x58684c11L, 0xc1611dabL, 0xb6662d3dL, 0x76dc4190L,
|
|
|
2056 |
0x01db7106L, 0x98d220bcL, 0xefd5102aL, 0x71b18589L, 0x06b6b51fL,
|
|
|
2057 |
0x9fbfe4a5L, 0xe8b8d433L, 0x7807c9a2L, 0x0f00f934L, 0x9609a88eL,
|
|
|
2058 |
0xe10e9818L, 0x7f6a0dbbL, 0x086d3d2dL, 0x91646c97L, 0xe6635c01L,
|
|
|
2059 |
0x6b6b51f4L, 0x1c6c6162L, 0x856530d8L, 0xf262004eL, 0x6c0695edL,
|
|
|
2060 |
0x1b01a57bL, 0x8208f4c1L, 0xf50fc457L, 0x65b0d9c6L, 0x12b7e950L,
|
|
|
2061 |
0x8bbeb8eaL, 0xfcb9887cL, 0x62dd1ddfL, 0x15da2d49L, 0x8cd37cf3L,
|
|
|
2062 |
0xfbd44c65L, 0x4db26158L, 0x3ab551ceL, 0xa3bc0074L, 0xd4bb30e2L,
|
|
|
2063 |
0x4adfa541L, 0x3dd895d7L, 0xa4d1c46dL, 0xd3d6f4fbL, 0x4369e96aL,
|
|
|
2064 |
0x346ed9fcL, 0xad678846L, 0xda60b8d0L, 0x44042d73L, 0x33031de5L,
|
|
|
2065 |
0xaa0a4c5fL, 0xdd0d7cc9L, 0x5005713cL, 0x270241aaL, 0xbe0b1010L,
|
|
|
2066 |
0xc90c2086L, 0x5768b525L, 0x206f85b3L, 0xb966d409L, 0xce61e49fL,
|
|
|
2067 |
0x5edef90eL, 0x29d9c998L, 0xb0d09822L, 0xc7d7a8b4L, 0x59b33d17L,
|
|
|
2068 |
0x2eb40d81L, 0xb7bd5c3bL, 0xc0ba6cadL, 0xedb88320L, 0x9abfb3b6L,
|
|
|
2069 |
0x03b6e20cL, 0x74b1d29aL, 0xead54739L, 0x9dd277afL, 0x04db2615L,
|
|
|
2070 |
0x73dc1683L, 0xe3630b12L, 0x94643b84L, 0x0d6d6a3eL, 0x7a6a5aa8L,
|
|
|
2071 |
0xe40ecf0bL, 0x9309ff9dL, 0x0a00ae27L, 0x7d079eb1L, 0xf00f9344L,
|
|
|
2072 |
0x8708a3d2L, 0x1e01f268L, 0x6906c2feL, 0xf762575dL, 0x806567cbL,
|
|
|
2073 |
0x196c3671L, 0x6e6b06e7L, 0xfed41b76L, 0x89d32be0L, 0x10da7a5aL,
|
|
|
2074 |
0x67dd4accL, 0xf9b9df6fL, 0x8ebeeff9L, 0x17b7be43L, 0x60b08ed5L,
|
|
|
2075 |
0xd6d6a3e8L, 0xa1d1937eL, 0x38d8c2c4L, 0x4fdff252L, 0xd1bb67f1L,
|
|
|
2076 |
0xa6bc5767L, 0x3fb506ddL, 0x48b2364bL, 0xd80d2bdaL, 0xaf0a1b4cL,
|
|
|
2077 |
0x36034af6L, 0x41047a60L, 0xdf60efc3L, 0xa867df55L, 0x316e8eefL,
|
|
|
2078 |
0x4669be79L, 0xcb61b38cL, 0xbc66831aL, 0x256fd2a0L, 0x5268e236L,
|
|
|
2079 |
0xcc0c7795L, 0xbb0b4703L, 0x220216b9L, 0x5505262fL, 0xc5ba3bbeL,
|
|
|
2080 |
0xb2bd0b28L, 0x2bb45a92L, 0x5cb36a04L, 0xc2d7ffa7L, 0xb5d0cf31L,
|
|
|
2081 |
0x2cd99e8bL, 0x5bdeae1dL, 0x9b64c2b0L, 0xec63f226L, 0x756aa39cL,
|
|
|
2082 |
0x026d930aL, 0x9c0906a9L, 0xeb0e363fL, 0x72076785L, 0x05005713L,
|
|
|
2083 |
0x95bf4a82L, 0xe2b87a14L, 0x7bb12baeL, 0x0cb61b38L, 0x92d28e9bL,
|
|
|
2084 |
0xe5d5be0dL, 0x7cdcefb7L, 0x0bdbdf21L, 0x86d3d2d4L, 0xf1d4e242L,
|
|
|
2085 |
0x68ddb3f8L, 0x1fda836eL, 0x81be16cdL, 0xf6b9265bL, 0x6fb077e1L,
|
|
|
2086 |
0x18b74777L, 0x88085ae6L, 0xff0f6a70L, 0x66063bcaL, 0x11010b5cL,
|
|
|
2087 |
0x8f659effL, 0xf862ae69L, 0x616bffd3L, 0x166ccf45L, 0xa00ae278L,
|
|
|
2088 |
0xd70dd2eeL, 0x4e048354L, 0x3903b3c2L, 0xa7672661L, 0xd06016f7L,
|
|
|
2089 |
0x4969474dL, 0x3e6e77dbL, 0xaed16a4aL, 0xd9d65adcL, 0x40df0b66L,
|
|
|
2090 |
0x37d83bf0L, 0xa9bcae53L, 0xdebb9ec5L, 0x47b2cf7fL, 0x30b5ffe9L,
|
|
|
2091 |
0xbdbdf21cL, 0xcabac28aL, 0x53b39330L, 0x24b4a3a6L, 0xbad03605L,
|
|
|
2092 |
0xcdd70693L, 0x54de5729L, 0x23d967bfL, 0xb3667a2eL, 0xc4614ab8L,
|
|
|
2093 |
0x5d681b02L, 0x2a6f2b94L, 0xb40bbe37L, 0xc30c8ea1L, 0x5a05df1bL,
|
|
|
2094 |
0x2d02ef8dL
|
|
|
2095 |
};
|
|
|
2096 |
|
|
|
2097 |
#define CRC32(c, b) (crc_table[((int)(c) ^ (b)) & 0xff] ^ ((c) >> 8))
|
|
|
2098 |
#define DO1(buf) crc = CRC32(crc, *buf++)
|
|
|
2099 |
#define DO2(buf) DO1(buf); DO1(buf)
|
|
|
2100 |
#define DO4(buf) DO2(buf); DO2(buf)
|
|
|
2101 |
#define DO8(buf) DO4(buf); DO4(buf)
|
|
|
2102 |
|
|
|
2103 |
ulg crc32(ulg crc, const uch *buf, extent len)
|
|
|
2104 |
{ if (buf==NULL) return 0L;
|
|
|
2105 |
crc = crc ^ 0xffffffffL;
|
|
|
2106 |
while (len >= 8) {DO8(buf); len -= 8;}
|
|
|
2107 |
if (len) do {DO1(buf);} while (--len);
|
|
|
2108 |
return crc ^ 0xffffffffL; // (instead of ~c for 64-bit machines)
|
|
|
2109 |
}
|
|
|
2110 |
|
|
|
2111 |
|
160 |
cycrow |
2112 |
void update_keys(unsigned long *keys, char c)
|
|
|
2113 |
{ keys[0] = CRC32(keys[0],c);
|
|
|
2114 |
keys[1] += keys[0] & 0xFF;
|
|
|
2115 |
keys[1] = keys[1]*134775813L +1;
|
|
|
2116 |
keys[2] = CRC32(keys[2], keys[1] >> 24);
|
|
|
2117 |
}
|
|
|
2118 |
char decrypt_byte(unsigned long *keys)
|
|
|
2119 |
{ unsigned temp = ((unsigned)keys[2] & 0xffff) | 2;
|
|
|
2120 |
return (char)(((temp * (temp ^ 1)) >> 8) & 0xff);
|
|
|
2121 |
}
|
|
|
2122 |
char zencode(unsigned long *keys, char c)
|
|
|
2123 |
{ int t=decrypt_byte(keys);
|
|
|
2124 |
update_keys(keys,c);
|
|
|
2125 |
return (char)(t^c);
|
|
|
2126 |
}
|
1 |
cycrow |
2127 |
|
|
|
2128 |
|
|
|
2129 |
|
|
|
2130 |
|
|
|
2131 |
|
|
|
2132 |
|
|
|
2133 |
|
|
|
2134 |
bool HasZipSuffix(const TCHAR *fn)
|
|
|
2135 |
{ const TCHAR *ext = fn+_tcslen(fn);
|
|
|
2136 |
while (ext>fn && *ext!='.') ext--;
|
|
|
2137 |
if (ext==fn && *ext!='.') return false;
|
|
|
2138 |
if (_tcsicmp(ext,_T(".Z"))==0) return true;
|
|
|
2139 |
if (_tcsicmp(ext,_T(".zip"))==0) return true;
|
|
|
2140 |
if (_tcsicmp(ext,_T(".zoo"))==0) return true;
|
|
|
2141 |
if (_tcsicmp(ext,_T(".arc"))==0) return true;
|
|
|
2142 |
if (_tcsicmp(ext,_T(".lzh"))==0) return true;
|
|
|
2143 |
if (_tcsicmp(ext,_T(".arj"))==0) return true;
|
|
|
2144 |
if (_tcsicmp(ext,_T(".gz"))==0) return true;
|
|
|
2145 |
if (_tcsicmp(ext,_T(".tgz"))==0) return true;
|
|
|
2146 |
return false;
|
|
|
2147 |
}
|
|
|
2148 |
|
|
|
2149 |
|
|
|
2150 |
lutime_t filetime2timet(const FILETIME ft)
|
|
|
2151 |
{ __int64 i = *(__int64*)&ft;
|
|
|
2152 |
return (lutime_t)((i-116444736000000000)/10000000);
|
|
|
2153 |
}
|
|
|
2154 |
|
|
|
2155 |
void filetime2dosdatetime(const FILETIME ft, WORD *dosdate,WORD *dostime)
|
|
|
2156 |
{ // date: bits 0-4 are day of month 1-31. Bits 5-8 are month 1..12. Bits 9-15 are year-1980
|
|
|
2157 |
// time: bits 0-4 are seconds/2, bits 5-10 are minute 0..59. Bits 11-15 are hour 0..23
|
|
|
2158 |
SYSTEMTIME st; FileTimeToSystemTime(&ft,&st);
|
|
|
2159 |
*dosdate = (WORD)(((st.wYear-1980)&0x7f) << 9);
|
|
|
2160 |
*dosdate |= (WORD)((st.wMonth&0xf) << 5);
|
|
|
2161 |
*dosdate |= (WORD)((st.wDay&0x1f));
|
|
|
2162 |
*dostime = (WORD)((st.wHour&0x1f) << 11);
|
|
|
2163 |
*dostime |= (WORD)((st.wMinute&0x3f) << 5);
|
|
|
2164 |
*dostime |= (WORD)((st.wSecond*2)&0x1f);
|
|
|
2165 |
}
|
|
|
2166 |
|
|
|
2167 |
|
|
|
2168 |
ZRESULT GetFileInfo(HANDLE hf, ulg *attr, long *size, iztimes *times, ulg *timestamp)
|
|
|
2169 |
{ // The handle must be a handle to a file
|
|
|
2170 |
// The date and time is returned in a long with the date most significant to allow
|
|
|
2171 |
// unsigned integer comparison of absolute times. The attributes have two
|
|
|
2172 |
// high bytes unix attr, and two low bytes a mapping of that to DOS attr.
|
|
|
2173 |
//struct stat s; int res=stat(fn,&s); if (res!=0) return false;
|
|
|
2174 |
// translate windows file attributes into zip ones.
|
|
|
2175 |
BY_HANDLE_FILE_INFORMATION bhi; BOOL res=GetFileInformationByHandle(hf,&bhi);
|
|
|
2176 |
if (!res) return ZR_NOFILE;
|
|
|
2177 |
DWORD fa=bhi.dwFileAttributes; ulg a=0;
|
|
|
2178 |
// Zip uses the lower word for its interpretation of windows stuff
|
|
|
2179 |
if (fa&FILE_ATTRIBUTE_READONLY) a|=0x01;
|
|
|
2180 |
if (fa&FILE_ATTRIBUTE_HIDDEN) a|=0x02;
|
|
|
2181 |
if (fa&FILE_ATTRIBUTE_SYSTEM) a|=0x04;
|
|
|
2182 |
if (fa&FILE_ATTRIBUTE_DIRECTORY)a|=0x10;
|
|
|
2183 |
if (fa&FILE_ATTRIBUTE_ARCHIVE) a|=0x20;
|
|
|
2184 |
// It uses the upper word for standard unix attr, which we manually construct
|
|
|
2185 |
if (fa&FILE_ATTRIBUTE_DIRECTORY)a|=0x40000000; // directory
|
|
|
2186 |
else a|=0x80000000; // normal file
|
|
|
2187 |
a|=0x01000000; // readable
|
|
|
2188 |
if (fa&FILE_ATTRIBUTE_READONLY) {} else a|=0x00800000; // writeable
|
|
|
2189 |
// now just a small heuristic to check if it's an executable:
|
|
|
2190 |
DWORD red, hsize=GetFileSize(hf,NULL); if (hsize>40)
|
|
|
2191 |
{ SetFilePointer(hf,0,NULL,FILE_BEGIN); unsigned short magic; ReadFile(hf,&magic,sizeof(magic),&red,NULL);
|
|
|
2192 |
SetFilePointer(hf,36,NULL,FILE_BEGIN); unsigned long hpos; ReadFile(hf,&hpos,sizeof(hpos),&red,NULL);
|
|
|
2193 |
if (magic==0x54AD && hsize>hpos+4+20+28)
|
|
|
2194 |
{ SetFilePointer(hf,hpos,NULL,FILE_BEGIN); unsigned long signature; ReadFile(hf,&signature,sizeof(signature),&red,NULL);
|
|
|
2195 |
if (signature==IMAGE_DOS_SIGNATURE || signature==IMAGE_OS2_SIGNATURE
|
|
|
2196 |
|| signature==IMAGE_OS2_SIGNATURE_LE || signature==IMAGE_NT_SIGNATURE)
|
|
|
2197 |
{ a |= 0x00400000; // executable
|
|
|
2198 |
}
|
|
|
2199 |
}
|
|
|
2200 |
}
|
|
|
2201 |
//
|
|
|
2202 |
if (attr!=NULL) *attr = a;
|
|
|
2203 |
if (size!=NULL) *size = hsize;
|
|
|
2204 |
if (times!=NULL)
|
|
|
2205 |
{ // lutime_t is 32bit number of seconds elapsed since 0:0:0GMT, Jan1, 1970.
|
|
|
2206 |
// but FILETIME is 64bit number of 100-nanosecs since Jan1, 1601
|
|
|
2207 |
times->atime = filetime2timet(bhi.ftLastAccessTime);
|
|
|
2208 |
times->mtime = filetime2timet(bhi.ftLastWriteTime);
|
|
|
2209 |
times->ctime = filetime2timet(bhi.ftCreationTime);
|
|
|
2210 |
}
|
|
|
2211 |
if (timestamp!=NULL)
|
|
|
2212 |
{ WORD dosdate,dostime;
|
|
|
2213 |
filetime2dosdatetime(bhi.ftLastWriteTime,&dosdate,&dostime);
|
|
|
2214 |
*timestamp = (WORD)dostime | (((DWORD)dosdate)<<16);
|
|
|
2215 |
}
|
|
|
2216 |
return ZR_OK;
|
|
|
2217 |
}
|
|
|
2218 |
|
|
|
2219 |
|
|
|
2220 |
|
|
|
2221 |
|
|
|
2222 |
|
|
|
2223 |
|
|
|
2224 |
|
|
|
2225 |
|
|
|
2226 |
class TZip
|
|
|
2227 |
{ public:
|
|
|
2228 |
TZip(const char *pwd) : hfout(0),mustclosehfout(false),hmapout(0),zfis(0),obuf(0),hfin(0),writ(0),oerr(false),hasputcen(false),ooffset(0),encwriting(false),encbuf(0),password(0), state(0) {if (pwd!=0 && *pwd!=0) {password=new char[strlen(pwd)+1]; strcpy(password,pwd);}}
|
|
|
2229 |
~TZip() {if (state!=0) delete state; state=0; if (encbuf!=0) delete[] encbuf; encbuf=0; if (password!=0) delete[] password; password=0;}
|
|
|
2230 |
|
|
|
2231 |
// These variables say about the file we're writing into
|
|
|
2232 |
// We can write to pipe, file-by-handle, file-by-name, memory-to-memmapfile
|
|
|
2233 |
char *password; // keep a copy of the password
|
|
|
2234 |
HANDLE hfout; // if valid, we'll write here (for files or pipes)
|
|
|
2235 |
bool mustclosehfout; // if true, we are responsible for closing hfout
|
|
|
2236 |
HANDLE hmapout; // otherwise, we'll write here (for memmap)
|
|
|
2237 |
unsigned ooffset; // for hfout, this is where the pointer was initially
|
|
|
2238 |
ZRESULT oerr; // did a write operation give rise to an error?
|
|
|
2239 |
unsigned writ; // how far have we written. This is maintained by Add, not write(), to avoid confusion over seeks
|
|
|
2240 |
bool ocanseek; // can we seek?
|
|
|
2241 |
char *obuf; // this is where we've locked mmap to view.
|
|
|
2242 |
unsigned int opos; // current pos in the mmap
|
|
|
2243 |
unsigned int mapsize; // the size of the map we created
|
|
|
2244 |
bool hasputcen; // have we yet placed the central directory?
|
|
|
2245 |
bool encwriting; // if true, then we'll encrypt stuff using 'keys' before we write it to disk
|
|
|
2246 |
unsigned long keys[3]; // keys are initialised inside Add()
|
|
|
2247 |
char *encbuf; // if encrypting, then this is a temporary workspace for encrypting the data
|
|
|
2248 |
unsigned int encbufsize; // (to be used and resized inside write(), and deleted in the destructor)
|
|
|
2249 |
//
|
|
|
2250 |
TZipFileInfo *zfis; // each file gets added onto this list, for writing the table at the end
|
|
|
2251 |
TState *state; // we use just one state object per zip, because it's big (500k)
|
|
|
2252 |
|
|
|
2253 |
ZRESULT Create(void *z,unsigned int len,DWORD flags);
|
|
|
2254 |
static unsigned sflush(void *param,const char *buf, unsigned *size);
|
|
|
2255 |
static unsigned swrite(void *param,const char *buf, unsigned size);
|
|
|
2256 |
unsigned int write(const char *buf,unsigned int size);
|
|
|
2257 |
bool oseek(unsigned int pos);
|
|
|
2258 |
ZRESULT GetMemory(void **pbuf, unsigned long *plen);
|
|
|
2259 |
ZRESULT Close();
|
|
|
2260 |
|
|
|
2261 |
// some variables to do with the file currently being read:
|
|
|
2262 |
// I haven't done it object-orientedly here, just put them all
|
|
|
2263 |
// together, since OO didn't seem to make the design any clearer.
|
|
|
2264 |
ulg attr; iztimes times; ulg timestamp; // all open_* methods set these
|
|
|
2265 |
bool iseekable; long isize,ired; // size is not set until close() on pips
|
|
|
2266 |
ulg crc; // crc is not set until close(). iwrit is cumulative
|
|
|
2267 |
HANDLE hfin; bool selfclosehf; // for input files and pipes
|
|
|
2268 |
const char *bufin; unsigned int lenin,posin; // for memory
|
|
|
2269 |
// and a variable for what we've done with the input: (i.e. compressed it!)
|
|
|
2270 |
ulg csize; // compressed size, set by the compression routines
|
|
|
2271 |
// and this is used by some of the compression routines
|
|
|
2272 |
char buf[16384];
|
|
|
2273 |
|
|
|
2274 |
|
|
|
2275 |
ZRESULT open_file(const TCHAR *fn);
|
|
|
2276 |
ZRESULT open_handle(HANDLE hf,unsigned int len);
|
|
|
2277 |
ZRESULT open_mem(void *src,unsigned int len);
|
|
|
2278 |
ZRESULT open_dir();
|
|
|
2279 |
static unsigned sread(TState &s,char *buf,unsigned size);
|
|
|
2280 |
unsigned read(char *buf, unsigned size);
|
|
|
2281 |
ZRESULT iclose();
|
|
|
2282 |
|
|
|
2283 |
ZRESULT ideflate(TZipFileInfo *zfi);
|
|
|
2284 |
ZRESULT istore();
|
|
|
2285 |
|
|
|
2286 |
ZRESULT Add(const TCHAR *odstzn, void *src,unsigned int len, DWORD flags);
|
|
|
2287 |
ZRESULT AddCentral();
|
|
|
2288 |
|
|
|
2289 |
};
|
|
|
2290 |
|
|
|
2291 |
|
|
|
2292 |
|
|
|
2293 |
ZRESULT TZip::Create(void *z,unsigned int len,DWORD flags)
|
|
|
2294 |
{ if (hfout!=0 || hmapout!=0 || obuf!=0 || writ!=0 || oerr!=ZR_OK || hasputcen) return ZR_NOTINITED;
|
|
|
2295 |
//
|
|
|
2296 |
if (flags==ZIP_HANDLE)
|
|
|
2297 |
{ HANDLE hf = (HANDLE)z;
|
|
|
2298 |
hfout=hf; mustclosehfout=false;
|
|
|
2299 |
#ifdef DuplicateHandle
|
|
|
2300 |
BOOL res = DuplicateHandle(GetCurrentProcess(),hf,GetCurrentProcess(),&hfout,0,FALSE,DUPLICATE_SAME_ACCESS);
|
|
|
2301 |
if (res) mustclosehandle=true;
|
|
|
2302 |
#endif
|
|
|
2303 |
// now we have hfout. Either we duplicated the handle and we close it ourselves
|
|
|
2304 |
// (while the caller closes h themselves), or we couldn't duplicate it.
|
|
|
2305 |
DWORD res = SetFilePointer(hfout,0,0,FILE_CURRENT);
|
|
|
2306 |
ocanseek = (res!=0xFFFFFFFF);
|
|
|
2307 |
if (ocanseek) ooffset=res; else ooffset=0;
|
|
|
2308 |
return ZR_OK;
|
|
|
2309 |
}
|
|
|
2310 |
else if (flags==ZIP_FILENAME)
|
|
|
2311 |
{ const TCHAR *fn = (const TCHAR*)z;
|
|
|
2312 |
hfout = CreateFile(fn,GENERIC_WRITE,0,NULL,CREATE_ALWAYS,FILE_ATTRIBUTE_NORMAL,NULL);
|
|
|
2313 |
if (hfout==INVALID_HANDLE_VALUE) {hfout=0; return ZR_NOFILE;}
|
|
|
2314 |
ocanseek=true;
|
|
|
2315 |
ooffset=0;
|
|
|
2316 |
mustclosehfout=true;
|
|
|
2317 |
return ZR_OK;
|
|
|
2318 |
}
|
|
|
2319 |
else if (flags==ZIP_MEMORY)
|
|
|
2320 |
{ unsigned int size = len;
|
|
|
2321 |
if (size==0) return ZR_MEMSIZE;
|
|
|
2322 |
if (z!=0) obuf=(char*)z;
|
|
|
2323 |
else
|
|
|
2324 |
{ hmapout = CreateFileMapping(INVALID_HANDLE_VALUE,NULL,PAGE_READWRITE,0,size,NULL);
|
|
|
2325 |
if (hmapout==NULL) return ZR_NOALLOC;
|
|
|
2326 |
obuf = (char*)MapViewOfFile(hmapout,FILE_MAP_ALL_ACCESS,0,0,size);
|
|
|
2327 |
if (obuf==0) {CloseHandle(hmapout); hmapout=0; return ZR_NOALLOC;}
|
|
|
2328 |
}
|
|
|
2329 |
ocanseek=true;
|
|
|
2330 |
opos=0; mapsize=size;
|
|
|
2331 |
return ZR_OK;
|
|
|
2332 |
}
|
|
|
2333 |
else return ZR_ARGS;
|
|
|
2334 |
}
|
|
|
2335 |
|
|
|
2336 |
unsigned TZip::sflush(void *param,const char *buf, unsigned *size)
|
|
|
2337 |
{ // static
|
|
|
2338 |
if (*size==0) return 0;
|
|
|
2339 |
TZip *zip = (TZip*)param;
|
|
|
2340 |
unsigned int writ = zip->write(buf,*size);
|
|
|
2341 |
if (writ!=0) *size=0;
|
|
|
2342 |
return writ;
|
|
|
2343 |
}
|
|
|
2344 |
unsigned TZip::swrite(void *param,const char *buf, unsigned size)
|
|
|
2345 |
{ // static
|
|
|
2346 |
if (size==0) return 0;
|
|
|
2347 |
TZip *zip=(TZip*)param; return zip->write(buf,size);
|
|
|
2348 |
}
|
|
|
2349 |
unsigned int TZip::write(const char *buf,unsigned int size)
|
|
|
2350 |
{ const char *srcbuf=buf;
|
|
|
2351 |
if (encwriting)
|
|
|
2352 |
{ if (encbuf!=0 && encbufsize<size) {delete[] encbuf; encbuf=0;}
|
|
|
2353 |
if (encbuf==0) {encbuf=new char[size*2]; encbufsize=size;}
|
|
|
2354 |
memcpy(encbuf,buf,size);
|
|
|
2355 |
for (unsigned int i=0; i<size; i++) encbuf[i]=zencode(keys,encbuf[i]);
|
|
|
2356 |
srcbuf=encbuf;
|
|
|
2357 |
}
|
|
|
2358 |
if (obuf!=0)
|
|
|
2359 |
{ if (opos+size>=mapsize) {oerr=ZR_MEMSIZE; return 0;}
|
|
|
2360 |
memcpy(obuf+opos, srcbuf, size);
|
|
|
2361 |
opos+=size;
|
|
|
2362 |
return size;
|
|
|
2363 |
}
|
|
|
2364 |
else if (hfout!=0)
|
|
|
2365 |
{ DWORD writ; WriteFile(hfout,srcbuf,size,&writ,NULL);
|
|
|
2366 |
return writ;
|
|
|
2367 |
}
|
|
|
2368 |
oerr=ZR_NOTINITED; return 0;
|
|
|
2369 |
}
|
|
|
2370 |
|
|
|
2371 |
bool TZip::oseek(unsigned int pos)
|
|
|
2372 |
{ if (!ocanseek) {oerr=ZR_SEEK; return false;}
|
|
|
2373 |
if (obuf!=0)
|
|
|
2374 |
{ if (pos>=mapsize) {oerr=ZR_MEMSIZE; return false;}
|
|
|
2375 |
opos=pos;
|
|
|
2376 |
return true;
|
|
|
2377 |
}
|
|
|
2378 |
else if (hfout!=0)
|
|
|
2379 |
{ SetFilePointer(hfout,pos+ooffset,NULL,FILE_BEGIN);
|
|
|
2380 |
return true;
|
|
|
2381 |
}
|
|
|
2382 |
oerr=ZR_NOTINITED; return 0;
|
|
|
2383 |
}
|
|
|
2384 |
|
|
|
2385 |
ZRESULT TZip::GetMemory(void **pbuf, unsigned long *plen)
|
|
|
2386 |
{ // When the user calls GetMemory, they're presumably at the end
|
|
|
2387 |
// of all their adding. In any case, we have to add the central
|
|
|
2388 |
// directory now, otherwise the memory we tell them won't be complete.
|
|
|
2389 |
if (!hasputcen) AddCentral(); hasputcen=true;
|
|
|
2390 |
if (pbuf!=NULL) *pbuf=(void*)obuf;
|
|
|
2391 |
if (plen!=NULL) *plen=writ;
|
|
|
2392 |
if (obuf==NULL) return ZR_NOTMMAP;
|
|
|
2393 |
return ZR_OK;
|
|
|
2394 |
}
|
|
|
2395 |
|
|
|
2396 |
ZRESULT TZip::Close()
|
|
|
2397 |
{ // if the directory hadn't already been added through a call to GetMemory,
|
|
|
2398 |
// then we do it now
|
|
|
2399 |
ZRESULT res=ZR_OK; if (!hasputcen) res=AddCentral(); hasputcen=true;
|
|
|
2400 |
if (obuf!=0 && hmapout!=0) UnmapViewOfFile(obuf); obuf=0;
|
|
|
2401 |
if (hmapout!=0) CloseHandle(hmapout); hmapout=0;
|
|
|
2402 |
if (hfout!=0 && mustclosehfout) CloseHandle(hfout); hfout=0; mustclosehfout=false;
|
|
|
2403 |
return res;
|
|
|
2404 |
}
|
|
|
2405 |
|
|
|
2406 |
|
|
|
2407 |
|
|
|
2408 |
|
|
|
2409 |
ZRESULT TZip::open_file(const TCHAR *fn)
|
|
|
2410 |
{ hfin=0; bufin=0; selfclosehf=false; crc=CRCVAL_INITIAL; isize=0; csize=0; ired=0;
|
|
|
2411 |
if (fn==0) return ZR_ARGS;
|
|
|
2412 |
HANDLE hf = CreateFile(fn,GENERIC_READ,FILE_SHARE_READ,NULL,OPEN_EXISTING,0,NULL);
|
|
|
2413 |
if (hf==INVALID_HANDLE_VALUE) return ZR_NOFILE;
|
|
|
2414 |
ZRESULT res = open_handle(hf,0);
|
|
|
2415 |
if (res!=ZR_OK) {CloseHandle(hf); return res;}
|
|
|
2416 |
selfclosehf=true;
|
|
|
2417 |
return ZR_OK;
|
|
|
2418 |
}
|
|
|
2419 |
ZRESULT TZip::open_handle(HANDLE hf,unsigned int len)
|
|
|
2420 |
{ hfin=0; bufin=0; selfclosehf=false; crc=CRCVAL_INITIAL; isize=0; csize=0; ired=0;
|
|
|
2421 |
if (hf==0 || hf==INVALID_HANDLE_VALUE) return ZR_ARGS;
|
|
|
2422 |
DWORD res = SetFilePointer(hfout,0,0,FILE_CURRENT);
|
|
|
2423 |
if (res!=0xFFFFFFFF)
|
|
|
2424 |
{ ZRESULT res = GetFileInfo(hf,&attr,&isize,×,×tamp);
|
|
|
2425 |
if (res!=ZR_OK) return res;
|
|
|
2426 |
SetFilePointer(hf,0,NULL,FILE_BEGIN); // because GetFileInfo will have screwed it up
|
|
|
2427 |
iseekable=true; hfin=hf;
|
|
|
2428 |
return ZR_OK;
|
|
|
2429 |
}
|
|
|
2430 |
else
|
|
|
2431 |
{ attr= 0x80000000; // just a normal file
|
|
|
2432 |
isize = -1; // can't know size until at the end
|
|
|
2433 |
if (len!=0) isize=len; // unless we were told explicitly!
|
|
|
2434 |
iseekable=false;
|
|
|
2435 |
SYSTEMTIME st; GetLocalTime(&st);
|
|
|
2436 |
FILETIME ft; SystemTimeToFileTime(&st,&ft);
|
|
|
2437 |
WORD dosdate,dostime; filetime2dosdatetime(ft,&dosdate,&dostime);
|
|
|
2438 |
times.atime = filetime2timet(ft);
|
|
|
2439 |
times.mtime = times.atime;
|
|
|
2440 |
times.ctime = times.atime;
|
|
|
2441 |
timestamp = (WORD)dostime | (((DWORD)dosdate)<<16);
|
|
|
2442 |
hfin=hf;
|
|
|
2443 |
return ZR_OK;
|
|
|
2444 |
}
|
|
|
2445 |
}
|
|
|
2446 |
ZRESULT TZip::open_mem(void *src,unsigned int len)
|
|
|
2447 |
{ hfin=0; bufin=(const char*)src; selfclosehf=false; crc=CRCVAL_INITIAL; ired=0; csize=0; ired=0;
|
|
|
2448 |
lenin=len; posin=0;
|
|
|
2449 |
if (src==0 || len==0) return ZR_ARGS;
|
|
|
2450 |
attr= 0x80000000; // just a normal file
|
|
|
2451 |
isize = len;
|
|
|
2452 |
iseekable=true;
|
|
|
2453 |
SYSTEMTIME st; GetLocalTime(&st);
|
|
|
2454 |
FILETIME ft; SystemTimeToFileTime(&st,&ft);
|
|
|
2455 |
WORD dosdate,dostime; filetime2dosdatetime(ft,&dosdate,&dostime);
|
|
|
2456 |
times.atime = filetime2timet(ft);
|
|
|
2457 |
times.mtime = times.atime;
|
|
|
2458 |
times.ctime = times.atime;
|
|
|
2459 |
timestamp = (WORD)dostime | (((DWORD)dosdate)<<16);
|
|
|
2460 |
return ZR_OK;
|
|
|
2461 |
}
|
|
|
2462 |
ZRESULT TZip::open_dir()
|
|
|
2463 |
{ hfin=0; bufin=0; selfclosehf=false; crc=CRCVAL_INITIAL; isize=0; csize=0; ired=0;
|
|
|
2464 |
attr= 0x41C00010; // a readable writable directory, and again directory
|
|
|
2465 |
isize = 0;
|
|
|
2466 |
iseekable=false;
|
|
|
2467 |
SYSTEMTIME st; GetLocalTime(&st);
|
|
|
2468 |
FILETIME ft; SystemTimeToFileTime(&st,&ft);
|
|
|
2469 |
WORD dosdate,dostime; filetime2dosdatetime(ft,&dosdate,&dostime);
|
|
|
2470 |
times.atime = filetime2timet(ft);
|
|
|
2471 |
times.mtime = times.atime;
|
|
|
2472 |
times.ctime = times.atime;
|
|
|
2473 |
timestamp = (WORD)dostime | (((DWORD)dosdate)<<16);
|
|
|
2474 |
return ZR_OK;
|
|
|
2475 |
}
|
|
|
2476 |
|
|
|
2477 |
unsigned TZip::sread(TState &s,char *buf,unsigned size)
|
|
|
2478 |
{ // static
|
|
|
2479 |
TZip *zip = (TZip*)s.param;
|
|
|
2480 |
return zip->read(buf,size);
|
|
|
2481 |
}
|
|
|
2482 |
|
|
|
2483 |
unsigned TZip::read(char *buf, unsigned size)
|
|
|
2484 |
{ if (bufin!=0)
|
|
|
2485 |
{ if (posin>=lenin) return 0; // end of input
|
|
|
2486 |
ulg red = lenin-posin;
|
|
|
2487 |
if (red>size) red=size;
|
|
|
2488 |
memcpy(buf, bufin+posin, red);
|
|
|
2489 |
posin += red;
|
|
|
2490 |
ired += red;
|
|
|
2491 |
crc = crc32(crc, (uch*)buf, red);
|
|
|
2492 |
return red;
|
|
|
2493 |
}
|
|
|
2494 |
else if (hfin!=0)
|
|
|
2495 |
{ DWORD red;
|
|
|
2496 |
BOOL ok = ReadFile(hfin,buf,size,&red,NULL);
|
|
|
2497 |
if (!ok) return 0;
|
|
|
2498 |
ired += red;
|
|
|
2499 |
crc = crc32(crc, (uch*)buf, red);
|
|
|
2500 |
return red;
|
|
|
2501 |
}
|
|
|
2502 |
else {oerr=ZR_NOTINITED; return 0;}
|
|
|
2503 |
}
|
|
|
2504 |
|
|
|
2505 |
ZRESULT TZip::iclose()
|
|
|
2506 |
{ if (selfclosehf && hfin!=0) CloseHandle(hfin); hfin=0;
|
|
|
2507 |
bool mismatch = (isize!=-1 && isize!=ired);
|
|
|
2508 |
isize=ired; // and crc has been being updated anyway
|
|
|
2509 |
if (mismatch) return ZR_MISSIZE;
|
|
|
2510 |
else return ZR_OK;
|
|
|
2511 |
}
|
|
|
2512 |
|
|
|
2513 |
|
|
|
2514 |
|
|
|
2515 |
ZRESULT TZip::ideflate(TZipFileInfo *zfi)
|
|
|
2516 |
{ if (state==0) state=new TState();
|
|
|
2517 |
// It's a very big object! 500k! We allocate it on the heap, because PocketPC's
|
|
|
2518 |
// stack breaks if we try to put it all on the stack. It will be deleted lazily
|
|
|
2519 |
state->err=0;
|
|
|
2520 |
state->readfunc=sread; state->flush_outbuf=sflush;
|
|
|
2521 |
state->param=this; state->level=8; state->seekable=iseekable; state->err=NULL;
|
|
|
2522 |
// the following line will make ct_init realise it has to perform the init
|
|
|
2523 |
state->ts.static_dtree[0].dl.len = 0;
|
|
|
2524 |
// Thanks to Alvin77 for this crucial fix:
|
|
|
2525 |
state->ds.window_size=0;
|
|
|
2526 |
// I think that covers everything that needs to be initted.
|
|
|
2527 |
//
|
|
|
2528 |
bi_init(*state,buf, sizeof(buf), TRUE); // it used to be just 1024-size, not 16384 as here
|
|
|
2529 |
ct_init(*state,&zfi->att);
|
|
|
2530 |
lm_init(*state,state->level, &zfi->flg);
|
|
|
2531 |
ulg sz = deflate(*state);
|
|
|
2532 |
csize=sz;
|
|
|
2533 |
ZRESULT r=ZR_OK; if (state->err!=NULL) r=ZR_FLATE;
|
|
|
2534 |
return r;
|
|
|
2535 |
}
|
|
|
2536 |
|
|
|
2537 |
ZRESULT TZip::istore()
|
|
|
2538 |
{ ulg size=0;
|
|
|
2539 |
for (;;)
|
|
|
2540 |
{ unsigned int cin=read(buf,16384); if (cin<=0 || cin==(unsigned int)EOF) break;
|
|
|
2541 |
unsigned int cout = write(buf,cin); if (cout!=cin) return ZR_MISSIZE;
|
|
|
2542 |
size += cin;
|
|
|
2543 |
}
|
|
|
2544 |
csize=size;
|
|
|
2545 |
return ZR_OK;
|
|
|
2546 |
}
|
|
|
2547 |
|
|
|
2548 |
|
|
|
2549 |
|
|
|
2550 |
|
|
|
2551 |
|
|
|
2552 |
bool has_seeded=false;
|
|
|
2553 |
ZRESULT TZip::Add(const TCHAR *odstzn, void *src,unsigned int len, DWORD flags)
|
|
|
2554 |
{ if (oerr) return ZR_FAILED;
|
|
|
2555 |
if (hasputcen) return ZR_ENDED;
|
|
|
2556 |
|
|
|
2557 |
// if we use password encryption, then every isize and csize is 12 bytes bigger
|
|
|
2558 |
int passex=0; if (password!=0 && flags!=ZIP_FOLDER) passex=12;
|
|
|
2559 |
|
|
|
2560 |
// zip has its own notion of what its names should look like: i.e. dir/file.stuff
|
|
|
2561 |
TCHAR dstzn[MAX_PATH]; _tcscpy(dstzn,odstzn);
|
|
|
2562 |
if (*dstzn==0) return ZR_ARGS;
|
|
|
2563 |
TCHAR *d=dstzn; while (*d!=0) {if (*d=='\\') *d='/'; d++;}
|
|
|
2564 |
bool isdir = (flags==ZIP_FOLDER);
|
|
|
2565 |
bool needs_trailing_slash = (isdir && dstzn[_tcslen(dstzn)-1]!='/');
|
|
|
2566 |
int method=DEFLATE; if (isdir || HasZipSuffix(dstzn)) method=STORE;
|
|
|
2567 |
|
|
|
2568 |
// now open whatever was our input source:
|
|
|
2569 |
ZRESULT openres;
|
|
|
2570 |
if (flags==ZIP_FILENAME) openres=open_file((const TCHAR*)src);
|
|
|
2571 |
else if (flags==ZIP_HANDLE) openres=open_handle((HANDLE)src,len);
|
|
|
2572 |
else if (flags==ZIP_MEMORY) openres=open_mem(src,len);
|
|
|
2573 |
else if (flags==ZIP_FOLDER) openres=open_dir();
|
|
|
2574 |
else return ZR_ARGS;
|
|
|
2575 |
if (openres!=ZR_OK) return openres;
|
|
|
2576 |
|
|
|
2577 |
// A zip "entry" consists of a local header (which includes the file name),
|
|
|
2578 |
// then the compressed data, and possibly an extended local header.
|
|
|
2579 |
|
|
|
2580 |
// Initialize the local header
|
|
|
2581 |
TZipFileInfo zfi; zfi.nxt=NULL;
|
|
|
2582 |
strcpy(zfi.name,"");
|
|
|
2583 |
#ifdef UNICODE
|
|
|
2584 |
WideCharToMultiByte(CP_UTF8,0,dstzn,-1,zfi.iname,MAX_PATH,0,0);
|
|
|
2585 |
#else
|
|
|
2586 |
strcpy(zfi.iname,dstzn);
|
|
|
2587 |
#endif
|
|
|
2588 |
zfi.nam=strlen(zfi.iname);
|
|
|
2589 |
if (needs_trailing_slash) {strcat(zfi.iname,"/"); zfi.nam++;}
|
|
|
2590 |
strcpy(zfi.zname,"");
|
|
|
2591 |
zfi.extra=NULL; zfi.ext=0; // extra header to go after this compressed data, and its length
|
|
|
2592 |
zfi.cextra=NULL; zfi.cext=0; // extra header to go in the central end-of-zip directory, and its length
|
|
|
2593 |
zfi.comment=NULL; zfi.com=0; // comment, and its length
|
|
|
2594 |
zfi.mark = 1;
|
|
|
2595 |
zfi.dosflag = 0;
|
|
|
2596 |
zfi.att = (ush)BINARY;
|
|
|
2597 |
zfi.vem = (ush)0xB17; // 0xB00 is win32 os-code. 0x17 is 23 in decimal: zip 2.3
|
|
|
2598 |
zfi.ver = (ush)20; // Needs PKUNZIP 2.0 to unzip it
|
|
|
2599 |
zfi.tim = timestamp;
|
|
|
2600 |
// Even though we write the header now, it will have to be rewritten, since we don't know compressed size or crc.
|
|
|
2601 |
zfi.crc = 0; // to be updated later
|
|
|
2602 |
zfi.flg = 8; // 8 means 'there is an extra header'. Assume for the moment that we need it.
|
|
|
2603 |
if (password!=0 && !isdir) zfi.flg=9; // and 1 means 'password-encrypted'
|
|
|
2604 |
zfi.lflg = zfi.flg; // to be updated later
|
|
|
2605 |
zfi.how = (ush)method; // to be updated later
|
|
|
2606 |
zfi.siz = (ulg)(method==STORE && isize>=0 ? isize+passex : 0); // to be updated later
|
|
|
2607 |
zfi.len = (ulg)(isize); // to be updated later
|
|
|
2608 |
zfi.dsk = 0;
|
|
|
2609 |
zfi.atx = attr;
|
|
|
2610 |
zfi.off = writ+ooffset; // offset within file of the start of this local record
|
|
|
2611 |
// stuff the 'times' structure into zfi.extra
|
|
|
2612 |
|
|
|
2613 |
// nb. apparently there's a problem with PocketPC CE(zip)->CE(unzip) fails. And removing the following block fixes it up.
|
|
|
2614 |
char xloc[EB_L_UT_SIZE]; zfi.extra=xloc; zfi.ext=EB_L_UT_SIZE;
|
|
|
2615 |
char xcen[EB_C_UT_SIZE]; zfi.cextra=xcen; zfi.cext=EB_C_UT_SIZE;
|
|
|
2616 |
xloc[0] = 'U';
|
|
|
2617 |
xloc[1] = 'T';
|
|
|
2618 |
xloc[2] = EB_UT_LEN(3); // length of data part of e.f.
|
|
|
2619 |
xloc[3] = 0;
|
|
|
2620 |
xloc[4] = EB_UT_FL_MTIME | EB_UT_FL_ATIME | EB_UT_FL_CTIME;
|
|
|
2621 |
xloc[5] = (char)(times.mtime);
|
|
|
2622 |
xloc[6] = (char)(times.mtime >> 8);
|
|
|
2623 |
xloc[7] = (char)(times.mtime >> 16);
|
|
|
2624 |
xloc[8] = (char)(times.mtime >> 24);
|
|
|
2625 |
xloc[9] = (char)(times.atime);
|
|
|
2626 |
xloc[10] = (char)(times.atime >> 8);
|
|
|
2627 |
xloc[11] = (char)(times.atime >> 16);
|
|
|
2628 |
xloc[12] = (char)(times.atime >> 24);
|
|
|
2629 |
xloc[13] = (char)(times.ctime);
|
|
|
2630 |
xloc[14] = (char)(times.ctime >> 8);
|
|
|
2631 |
xloc[15] = (char)(times.ctime >> 16);
|
|
|
2632 |
xloc[16] = (char)(times.ctime >> 24);
|
|
|
2633 |
memcpy(zfi.cextra,zfi.extra,EB_C_UT_SIZE);
|
|
|
2634 |
zfi.cextra[EB_LEN] = EB_UT_LEN(1);
|
|
|
2635 |
|
|
|
2636 |
|
|
|
2637 |
// (1) Start by writing the local header:
|
|
|
2638 |
int r = putlocal(&zfi,swrite,this);
|
|
|
2639 |
if (r!=ZE_OK) {iclose(); return ZR_WRITE;}
|
|
|
2640 |
writ += 4 + LOCHEAD + (unsigned int)zfi.nam + (unsigned int)zfi.ext;
|
|
|
2641 |
if (oerr!=ZR_OK) {iclose(); return oerr;}
|
|
|
2642 |
|
|
|
2643 |
// (1.5) if necessary, write the encryption header
|
160 |
cycrow |
2644 |
keys[0]=305419896L;
|
|
|
2645 |
keys[1]=591751049L;
|
|
|
2646 |
keys[2]=878082192L;
|
|
|
2647 |
for (const char *cp=password; cp!=0 && *cp!=0; cp++) update_keys(keys,*cp);
|
|
|
2648 |
// generate some random bytes
|
|
|
2649 |
if (!has_seeded) srand(GetTickCount()^(unsigned long)GetDesktopWindow());
|
|
|
2650 |
char encbuf[12]; for (int i=0; i<12; i++) encbuf[i]=(char)((rand()>>7)&0xff);
|
|
|
2651 |
encbuf[11] = (char)((zfi.tim>>8)&0xff);
|
|
|
2652 |
for (int ei=0; ei<12; ei++) encbuf[ei]=zencode(keys,encbuf[ei]);
|
|
|
2653 |
if (password!=0 && !isdir) {swrite(this,encbuf,12); writ+=12;}
|
1 |
cycrow |
2654 |
|
|
|
2655 |
//(2) Write deflated/stored file to zip file
|
|
|
2656 |
ZRESULT writeres=ZR_OK;
|
|
|
2657 |
encwriting = (password!=0 && !isdir); // an object member variable to say whether we write to disk encrypted
|
|
|
2658 |
if (!isdir && method==DEFLATE) writeres=ideflate(&zfi);
|
|
|
2659 |
else if (!isdir && method==STORE) writeres=istore();
|
|
|
2660 |
else if (isdir) csize=0;
|
|
|
2661 |
encwriting = false;
|
|
|
2662 |
iclose();
|
|
|
2663 |
writ += csize;
|
|
|
2664 |
if (oerr!=ZR_OK) return oerr;
|
|
|
2665 |
if (writeres!=ZR_OK) return ZR_WRITE;
|
|
|
2666 |
|
|
|
2667 |
// (3) Either rewrite the local header with correct information...
|
|
|
2668 |
bool first_header_has_size_right = (zfi.siz==csize+passex);
|
|
|
2669 |
zfi.crc = crc;
|
|
|
2670 |
zfi.siz = csize+passex;
|
|
|
2671 |
zfi.len = isize;
|
|
|
2672 |
if (ocanseek && (password==0 || isdir))
|
|
|
2673 |
{ zfi.how = (ush)method;
|
|
|
2674 |
if ((zfi.flg & 1) == 0) zfi.flg &= ~8; // clear the extended local header flag
|
|
|
2675 |
zfi.lflg = zfi.flg;
|
|
|
2676 |
// rewrite the local header:
|
|
|
2677 |
if (!oseek(zfi.off-ooffset)) return ZR_SEEK;
|
|
|
2678 |
if ((r = putlocal(&zfi, swrite,this)) != ZE_OK) return ZR_WRITE;
|
|
|
2679 |
if (!oseek(writ)) return ZR_SEEK;
|
|
|
2680 |
}
|
|
|
2681 |
else
|
|
|
2682 |
{ // (4) ... or put an updated header at the end
|
|
|
2683 |
if (zfi.how != (ush) method) return ZR_NOCHANGE;
|
|
|
2684 |
if (method==STORE && !first_header_has_size_right) return ZR_NOCHANGE;
|
|
|
2685 |
if ((r = putextended(&zfi, swrite,this)) != ZE_OK) return ZR_WRITE;
|
|
|
2686 |
writ += 16L;
|
|
|
2687 |
zfi.flg = zfi.lflg; // if flg modified by inflate, for the central index
|
|
|
2688 |
}
|
|
|
2689 |
if (oerr!=ZR_OK) return oerr;
|
|
|
2690 |
|
|
|
2691 |
// Keep a copy of the zipfileinfo, for our end-of-zip directory
|
|
|
2692 |
char *cextra = new char[zfi.cext]; memcpy(cextra,zfi.cextra,zfi.cext); zfi.cextra=cextra;
|
|
|
2693 |
TZipFileInfo *pzfi = new TZipFileInfo; memcpy(pzfi,&zfi,sizeof(zfi));
|
|
|
2694 |
if (zfis==NULL) zfis=pzfi;
|
|
|
2695 |
else {TZipFileInfo *z=zfis; while (z->nxt!=NULL) z=z->nxt; z->nxt=pzfi;}
|
|
|
2696 |
return ZR_OK;
|
|
|
2697 |
}
|
|
|
2698 |
|
|
|
2699 |
ZRESULT TZip::AddCentral()
|
|
|
2700 |
{ // write central directory
|
|
|
2701 |
int numentries = 0;
|
|
|
2702 |
ulg pos_at_start_of_central = writ;
|
|
|
2703 |
//ulg tot_unc_size=0, tot_compressed_size=0;
|
|
|
2704 |
bool okay=true;
|
|
|
2705 |
for (TZipFileInfo *zfi=zfis; zfi!=NULL; )
|
|
|
2706 |
{ if (okay)
|
|
|
2707 |
{ int res = putcentral(zfi, swrite,this);
|
|
|
2708 |
if (res!=ZE_OK) okay=false;
|
|
|
2709 |
}
|
|
|
2710 |
writ += 4 + CENHEAD + (unsigned int)zfi->nam + (unsigned int)zfi->cext + (unsigned int)zfi->com;
|
|
|
2711 |
//tot_unc_size += zfi->len;
|
|
|
2712 |
//tot_compressed_size += zfi->siz;
|
|
|
2713 |
numentries++;
|
|
|
2714 |
//
|
|
|
2715 |
TZipFileInfo *zfinext = zfi->nxt;
|
|
|
2716 |
if (zfi->cextra!=0) delete[] zfi->cextra;
|
|
|
2717 |
delete zfi;
|
|
|
2718 |
zfi = zfinext;
|
|
|
2719 |
}
|
|
|
2720 |
ulg center_size = writ - pos_at_start_of_central;
|
|
|
2721 |
if (okay)
|
|
|
2722 |
{ int res = putend(numentries, center_size, pos_at_start_of_central+ooffset, 0, NULL, swrite,this);
|
|
|
2723 |
if (res!=ZE_OK) okay=false;
|
|
|
2724 |
writ += 4 + ENDHEAD + 0;
|
|
|
2725 |
}
|
|
|
2726 |
if (!okay) return ZR_WRITE;
|
|
|
2727 |
return ZR_OK;
|
|
|
2728 |
}
|
|
|
2729 |
|
|
|
2730 |
|
|
|
2731 |
|
|
|
2732 |
|
|
|
2733 |
|
|
|
2734 |
ZRESULT lasterrorZ=ZR_OK;
|
|
|
2735 |
|
|
|
2736 |
unsigned int FormatZipMessageZ(ZRESULT code, char *buf,unsigned int len)
|
|
|
2737 |
{ if (code==ZR_RECENT) code=lasterrorZ;
|
|
|
2738 |
const char *msg="unknown zip result code";
|
|
|
2739 |
switch (code)
|
|
|
2740 |
{ case ZR_OK: msg="Success"; break;
|
|
|
2741 |
case ZR_NODUPH: msg="Culdn't duplicate handle"; break;
|
|
|
2742 |
case ZR_NOFILE: msg="Couldn't create/open file"; break;
|
|
|
2743 |
case ZR_NOALLOC: msg="Failed to allocate memory"; break;
|
|
|
2744 |
case ZR_WRITE: msg="Error writing to file"; break;
|
|
|
2745 |
case ZR_NOTFOUND: msg="File not found in the zipfile"; break;
|
|
|
2746 |
case ZR_MORE: msg="Still more data to unzip"; break;
|
|
|
2747 |
case ZR_CORRUPT: msg="Zipfile is corrupt or not a zipfile"; break;
|
|
|
2748 |
case ZR_READ: msg="Error reading file"; break;
|
|
|
2749 |
case ZR_ARGS: msg="Caller: faulty arguments"; break;
|
|
|
2750 |
case ZR_PARTIALUNZ: msg="Caller: the file had already been partially unzipped"; break;
|
|
|
2751 |
case ZR_NOTMMAP: msg="Caller: can only get memory of a memory zipfile"; break;
|
|
|
2752 |
case ZR_MEMSIZE: msg="Caller: not enough space allocated for memory zipfile"; break;
|
|
|
2753 |
case ZR_FAILED: msg="Caller: there was a previous error"; break;
|
|
|
2754 |
case ZR_ENDED: msg="Caller: additions to the zip have already been ended"; break;
|
|
|
2755 |
case ZR_ZMODE: msg="Caller: mixing creation and opening of zip"; break;
|
|
|
2756 |
case ZR_NOTINITED: msg="Zip-bug: internal initialisation not completed"; break;
|
|
|
2757 |
case ZR_SEEK: msg="Zip-bug: trying to seek the unseekable"; break;
|
|
|
2758 |
case ZR_MISSIZE: msg="Zip-bug: the anticipated size turned out wrong"; break;
|
|
|
2759 |
case ZR_NOCHANGE: msg="Zip-bug: tried to change mind, but not allowed"; break;
|
|
|
2760 |
case ZR_FLATE: msg="Zip-bug: an internal error during flation"; break;
|
|
|
2761 |
}
|
|
|
2762 |
unsigned int mlen=(unsigned int)strlen(msg);
|
|
|
2763 |
if (buf==0 || len==0) return mlen;
|
|
|
2764 |
unsigned int n=mlen; if (n+1>len) n=len-1;
|
|
|
2765 |
strncpy(buf,msg,n); buf[n]=0;
|
|
|
2766 |
return mlen;
|
|
|
2767 |
}
|
|
|
2768 |
|
|
|
2769 |
|
|
|
2770 |
|
|
|
2771 |
typedef struct
|
|
|
2772 |
{ DWORD flag;
|
|
|
2773 |
TZip *zip;
|
|
|
2774 |
} TZipHandleData;
|
|
|
2775 |
|
|
|
2776 |
|
|
|
2777 |
HZIP CreateZipInternal(void *z,unsigned int len,DWORD flags, const char *password)
|
|
|
2778 |
{ TZip *zip = new TZip(password);
|
|
|
2779 |
lasterrorZ = zip->Create(z,len,flags);
|
|
|
2780 |
if (lasterrorZ!=ZR_OK) {delete zip; return 0;}
|
|
|
2781 |
TZipHandleData *han = new TZipHandleData;
|
|
|
2782 |
han->flag=2; han->zip=zip; return (HZIP)han;
|
|
|
2783 |
}
|
|
|
2784 |
HZIP CreateZipHandle(HANDLE h, const char *password) {return CreateZipInternal(h,0,ZIP_HANDLE,password);}
|
|
|
2785 |
HZIP CreateZip(const TCHAR *fn, const char *password) {return CreateZipInternal((void*)fn,0,ZIP_FILENAME,password);}
|
|
|
2786 |
HZIP CreateZip(void *z,unsigned int len, const char *password) {return CreateZipInternal(z,len,ZIP_MEMORY,password);}
|
|
|
2787 |
|
|
|
2788 |
|
|
|
2789 |
ZRESULT ZipAddInternal(HZIP hz,const TCHAR *dstzn, void *src,unsigned int len, DWORD flags)
|
|
|
2790 |
{ if (hz==0) {lasterrorZ=ZR_ARGS;return ZR_ARGS;}
|
|
|
2791 |
TZipHandleData *han = (TZipHandleData*)hz;
|
|
|
2792 |
if (han->flag!=2) {lasterrorZ=ZR_ZMODE;return ZR_ZMODE;}
|
|
|
2793 |
TZip *zip = han->zip;
|
|
|
2794 |
lasterrorZ = zip->Add(dstzn,src,len,flags);
|
|
|
2795 |
return lasterrorZ;
|
|
|
2796 |
}
|
|
|
2797 |
ZRESULT ZipAdd(HZIP hz,const TCHAR *dstzn, const TCHAR *fn) {return ZipAddInternal(hz,dstzn,(void*)fn,0,ZIP_FILENAME);}
|
|
|
2798 |
ZRESULT ZipAdd(HZIP hz,const TCHAR *dstzn, void *src,unsigned int len) {return ZipAddInternal(hz,dstzn,src,len,ZIP_MEMORY);}
|
|
|
2799 |
ZRESULT ZipAddHandle(HZIP hz,const TCHAR *dstzn, HANDLE h) {return ZipAddInternal(hz,dstzn,h,0,ZIP_HANDLE);}
|
|
|
2800 |
ZRESULT ZipAddHandle(HZIP hz,const TCHAR *dstzn, HANDLE h, unsigned int len) {return ZipAddInternal(hz,dstzn,h,len,ZIP_HANDLE);}
|
|
|
2801 |
ZRESULT ZipAddFolder(HZIP hz,const TCHAR *dstzn) {return ZipAddInternal(hz,dstzn,0,0,ZIP_FOLDER);}
|
|
|
2802 |
|
|
|
2803 |
|
|
|
2804 |
|
|
|
2805 |
ZRESULT ZipGetMemory(HZIP hz, void **buf, unsigned long *len)
|
|
|
2806 |
{ if (hz==0) {if (buf!=0) *buf=0; if (len!=0) *len=0; lasterrorZ=ZR_ARGS;return ZR_ARGS;}
|
|
|
2807 |
TZipHandleData *han = (TZipHandleData*)hz;
|
|
|
2808 |
if (han->flag!=2) {lasterrorZ=ZR_ZMODE;return ZR_ZMODE;}
|
|
|
2809 |
TZip *zip = han->zip;
|
|
|
2810 |
lasterrorZ = zip->GetMemory(buf,len);
|
|
|
2811 |
return lasterrorZ;
|
|
|
2812 |
}
|
|
|
2813 |
|
|
|
2814 |
ZRESULT CloseZipZ(HZIP hz)
|
|
|
2815 |
{ if (hz==0) {lasterrorZ=ZR_ARGS;return ZR_ARGS;}
|
|
|
2816 |
TZipHandleData *han = (TZipHandleData*)hz;
|
|
|
2817 |
if (han->flag!=2) {lasterrorZ=ZR_ZMODE;return ZR_ZMODE;}
|
|
|
2818 |
TZip *zip = han->zip;
|
|
|
2819 |
lasterrorZ = zip->Close();
|
|
|
2820 |
delete zip;
|
|
|
2821 |
delete han;
|
|
|
2822 |
return lasterrorZ;
|
|
|
2823 |
}
|
|
|
2824 |
|
|
|
2825 |
bool IsZipHandleZ(HZIP hz)
|
|
|
2826 |
{ if (hz==0) return false;
|
|
|
2827 |
TZipHandleData *han = (TZipHandleData*)hz;
|
|
|
2828 |
return (han->flag==2);
|
|
|
2829 |
}
|
|
|
2830 |
|