[821] | 1 | /* vi: set sw=4 ts=4: */
|
---|
| 2 | /*
|
---|
| 3 | * Gzip implementation for busybox
|
---|
| 4 | *
|
---|
| 5 | * Based on GNU gzip Copyright (C) 1992-1993 Jean-loup Gailly.
|
---|
| 6 | *
|
---|
| 7 | * Originally adjusted for busybox by Charles P. Wright <cpw@unix.asb.com>
|
---|
[1765] | 8 | * "this is a stripped down version of gzip I put into busybox, it does
|
---|
| 9 | * only standard in to standard out with -9 compression. It also requires
|
---|
| 10 | * the zcat module for some important functions."
|
---|
[821] | 11 | *
|
---|
| 12 | * Adjusted further by Erik Andersen <andersen@codepoet.org> to support
|
---|
| 13 | * files as well as stdin/stdout, and to generally behave itself wrt
|
---|
| 14 | * command line handling.
|
---|
| 15 | *
|
---|
[2725] | 16 | * Licensed under GPLv2 or later, see file LICENSE in this source tree.
|
---|
[821] | 17 | */
|
---|
[1765] | 18 | /* big objects in bss:
|
---|
| 19 | * 00000020 b bl_count
|
---|
| 20 | * 00000074 b base_length
|
---|
| 21 | * 00000078 b base_dist
|
---|
| 22 | * 00000078 b static_dtree
|
---|
| 23 | * 0000009c b bl_tree
|
---|
| 24 | * 000000f4 b dyn_dtree
|
---|
| 25 | * 00000100 b length_code
|
---|
| 26 | * 00000200 b dist_code
|
---|
| 27 | * 0000023d b depth
|
---|
| 28 | * 00000400 b flag_buf
|
---|
| 29 | * 0000047a b heap
|
---|
| 30 | * 00000480 b static_ltree
|
---|
| 31 | * 000008f4 b dyn_ltree
|
---|
| 32 | */
|
---|
| 33 | /* TODO: full support for -v for DESKTOP
|
---|
| 34 | * "/usr/bin/gzip -v a bogus aa" should say:
|
---|
| 35 | a: 85.1% -- replaced with a.gz
|
---|
| 36 | gzip: bogus: No such file or directory
|
---|
| 37 | aa: 85.1% -- replaced with aa.gz
|
---|
| 38 | */
|
---|
[821] | 39 |
|
---|
[3621] | 40 | //config:config GZIP
|
---|
| 41 | //config: bool "gzip"
|
---|
| 42 | //config: default y
|
---|
| 43 | //config: help
|
---|
| 44 | //config: gzip is used to compress files.
|
---|
| 45 | //config: It's probably the most widely used UNIX compression program.
|
---|
| 46 | //config:
|
---|
| 47 | //config:config FEATURE_GZIP_LONG_OPTIONS
|
---|
| 48 | //config: bool "Enable long options"
|
---|
| 49 | //config: default y
|
---|
| 50 | //config: depends on GZIP && LONG_OPTS
|
---|
| 51 | //config: help
|
---|
| 52 | //config: Enable use of long options, increases size by about 106 Bytes
|
---|
| 53 | //config:
|
---|
| 54 | //config:config GZIP_FAST
|
---|
| 55 | //config: int "Trade memory for gzip speed (0:small,slow - 2:fast,big)"
|
---|
| 56 | //config: default 0
|
---|
| 57 | //config: range 0 2
|
---|
| 58 | //config: depends on GZIP
|
---|
| 59 | //config: help
|
---|
| 60 | //config: Enable big memory options for gzip.
|
---|
| 61 | //config: 0: small buffers, small hash-tables
|
---|
| 62 | //config: 1: larger buffers, larger hash-tables
|
---|
| 63 | //config: 2: larger buffers, largest hash-tables
|
---|
| 64 | //config: Larger models may give slightly better compression
|
---|
| 65 | //config:
|
---|
| 66 | //config:config FEATURE_GZIP_LEVELS
|
---|
| 67 | //config: bool "Enable compression levels"
|
---|
| 68 | //config: default n
|
---|
| 69 | //config: depends on GZIP
|
---|
| 70 | //config: help
|
---|
| 71 | //config: Enable support for compression levels 4-9. The default level
|
---|
| 72 | //config: is 6. If levels 1-3 are specified, 4 is used.
|
---|
| 73 | //config: If this option is not selected, -N options are ignored and -9
|
---|
| 74 | //config: is used.
|
---|
| 75 |
|
---|
| 76 | //applet:IF_GZIP(APPLET(gzip, BB_DIR_BIN, BB_SUID_DROP))
|
---|
| 77 | //kbuild:lib-$(CONFIG_GZIP) += gzip.o
|
---|
| 78 |
|
---|
[3232] | 79 | //usage:#define gzip_trivial_usage
|
---|
[3621] | 80 | //usage: "[-cf" IF_GUNZIP("d") IF_FEATURE_GZIP_LEVELS("123456789") "] [FILE]..."
|
---|
[3232] | 81 | //usage:#define gzip_full_usage "\n\n"
|
---|
| 82 | //usage: "Compress FILEs (or stdin)\n"
|
---|
[3621] | 83 | //usage: IF_FEATURE_GZIP_LEVELS(
|
---|
| 84 | //usage: "\n -1..9 Compression level"
|
---|
| 85 | //usage: )
|
---|
| 86 | //usage: IF_GUNZIP(
|
---|
[3232] | 87 | //usage: "\n -d Decompress"
|
---|
[3621] | 88 | //usage: )
|
---|
[3232] | 89 | //usage: "\n -c Write to stdout"
|
---|
| 90 | //usage: "\n -f Force"
|
---|
| 91 | //usage:
|
---|
| 92 | //usage:#define gzip_example_usage
|
---|
| 93 | //usage: "$ ls -la /tmp/busybox*\n"
|
---|
| 94 | //usage: "-rw-rw-r-- 1 andersen andersen 1761280 Apr 14 17:47 /tmp/busybox.tar\n"
|
---|
| 95 | //usage: "$ gzip /tmp/busybox.tar\n"
|
---|
| 96 | //usage: "$ ls -la /tmp/busybox*\n"
|
---|
| 97 | //usage: "-rw-rw-r-- 1 andersen andersen 554058 Apr 14 17:49 /tmp/busybox.tar.gz\n"
|
---|
| 98 |
|
---|
[1765] | 99 | #include "libbb.h"
|
---|
[3232] | 100 | #include "bb_archive.h"
|
---|
[821] | 101 |
|
---|
| 102 |
|
---|
[1765] | 103 | /* ===========================================================================
|
---|
| 104 | */
|
---|
| 105 | //#define DEBUG 1
|
---|
| 106 | /* Diagnostic functions */
|
---|
| 107 | #ifdef DEBUG
|
---|
| 108 | # define Assert(cond,msg) { if (!(cond)) bb_error_msg(msg); }
|
---|
| 109 | # define Trace(x) fprintf x
|
---|
| 110 | # define Tracev(x) {if (verbose) fprintf x; }
|
---|
| 111 | # define Tracevv(x) {if (verbose > 1) fprintf x; }
|
---|
| 112 | # define Tracec(c,x) {if (verbose && (c)) fprintf x; }
|
---|
| 113 | # define Tracecv(c,x) {if (verbose > 1 && (c)) fprintf x; }
|
---|
| 114 | #else
|
---|
| 115 | # define Assert(cond,msg)
|
---|
| 116 | # define Trace(x)
|
---|
| 117 | # define Tracev(x)
|
---|
| 118 | # define Tracevv(x)
|
---|
| 119 | # define Tracec(c,x)
|
---|
| 120 | # define Tracecv(c,x)
|
---|
| 121 | #endif
|
---|
[821] | 122 |
|
---|
[1765] | 123 |
|
---|
| 124 | /* ===========================================================================
|
---|
[821] | 125 | */
|
---|
[3232] | 126 | #if CONFIG_GZIP_FAST == 0
|
---|
| 127 | # define SMALL_MEM
|
---|
| 128 | #elif CONFIG_GZIP_FAST == 1
|
---|
| 129 | # define MEDIUM_MEM
|
---|
| 130 | #elif CONFIG_GZIP_FAST == 2
|
---|
| 131 | # define BIG_MEM
|
---|
| 132 | #else
|
---|
| 133 | # error "Invalid CONFIG_GZIP_FAST value"
|
---|
| 134 | #endif
|
---|
[821] | 135 |
|
---|
[2725] | 136 | #ifndef INBUFSIZ
|
---|
[821] | 137 | # ifdef SMALL_MEM
|
---|
| 138 | # define INBUFSIZ 0x2000 /* input buffer size */
|
---|
| 139 | # else
|
---|
| 140 | # define INBUFSIZ 0x8000 /* input buffer size */
|
---|
| 141 | # endif
|
---|
| 142 | #endif
|
---|
| 143 |
|
---|
[2725] | 144 | #ifndef OUTBUFSIZ
|
---|
[821] | 145 | # ifdef SMALL_MEM
|
---|
| 146 | # define OUTBUFSIZ 8192 /* output buffer size */
|
---|
| 147 | # else
|
---|
| 148 | # define OUTBUFSIZ 16384 /* output buffer size */
|
---|
| 149 | # endif
|
---|
| 150 | #endif
|
---|
| 151 |
|
---|
| 152 | #ifndef DIST_BUFSIZE
|
---|
| 153 | # ifdef SMALL_MEM
|
---|
| 154 | # define DIST_BUFSIZE 0x2000 /* buffer for distances, see trees.c */
|
---|
| 155 | # else
|
---|
| 156 | # define DIST_BUFSIZE 0x8000 /* buffer for distances, see trees.c */
|
---|
| 157 | # endif
|
---|
| 158 | #endif
|
---|
| 159 |
|
---|
| 160 | /* gzip flag byte */
|
---|
| 161 | #define ASCII_FLAG 0x01 /* bit 0 set: file probably ascii text */
|
---|
| 162 | #define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gzip file */
|
---|
| 163 | #define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */
|
---|
| 164 | #define ORIG_NAME 0x08 /* bit 3 set: original file name present */
|
---|
| 165 | #define COMMENT 0x10 /* bit 4 set: file comment present */
|
---|
| 166 | #define RESERVED 0xC0 /* bit 6,7: reserved */
|
---|
| 167 |
|
---|
| 168 | /* internal file attribute */
|
---|
| 169 | #define UNKNOWN 0xffff
|
---|
| 170 | #define BINARY 0
|
---|
| 171 | #define ASCII 1
|
---|
| 172 |
|
---|
| 173 | #ifndef WSIZE
|
---|
[1765] | 174 | # define WSIZE 0x8000 /* window size--must be a power of two, and */
|
---|
| 175 | #endif /* at least 32K for zip's deflate method */
|
---|
[821] | 176 |
|
---|
| 177 | #define MIN_MATCH 3
|
---|
| 178 | #define MAX_MATCH 258
|
---|
| 179 | /* The minimum and maximum match lengths */
|
---|
| 180 |
|
---|
| 181 | #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1)
|
---|
| 182 | /* Minimum amount of lookahead, except at the end of the input file.
|
---|
| 183 | * See deflate.c for comments about the MIN_MATCH+1.
|
---|
| 184 | */
|
---|
| 185 |
|
---|
| 186 | #define MAX_DIST (WSIZE-MIN_LOOKAHEAD)
|
---|
| 187 | /* In order to simplify the code, particularly on 16 bit machines, match
|
---|
| 188 | * distances are limited to MAX_DIST instead of WSIZE.
|
---|
| 189 | */
|
---|
| 190 |
|
---|
[1765] | 191 | #ifndef MAX_PATH_LEN
|
---|
| 192 | # define MAX_PATH_LEN 1024 /* max pathname length */
|
---|
[821] | 193 | #endif
|
---|
| 194 |
|
---|
| 195 | #define seekable() 0 /* force sequential output */
|
---|
| 196 | #define translate_eol 0 /* no option -a yet */
|
---|
| 197 |
|
---|
| 198 | #ifndef BITS
|
---|
| 199 | # define BITS 16
|
---|
| 200 | #endif
|
---|
| 201 | #define INIT_BITS 9 /* Initial number of bits per code */
|
---|
| 202 |
|
---|
| 203 | #define BIT_MASK 0x1f /* Mask for 'number of compression bits' */
|
---|
| 204 | /* Mask 0x20 is reserved to mean a fourth header byte, and 0x40 is free.
|
---|
| 205 | * It's a pity that old uncompress does not check bit 0x20. That makes
|
---|
| 206 | * extension of the format actually undesirable because old compress
|
---|
| 207 | * would just crash on the new format instead of giving a meaningful
|
---|
| 208 | * error message. It does check the number of bits, but it's more
|
---|
| 209 | * helpful to say "unsupported format, get a new version" than
|
---|
| 210 | * "can only handle 16 bits".
|
---|
| 211 | */
|
---|
| 212 |
|
---|
[1765] | 213 | #ifdef MAX_EXT_CHARS
|
---|
| 214 | # define MAX_SUFFIX MAX_EXT_CHARS
|
---|
| 215 | #else
|
---|
| 216 | # define MAX_SUFFIX 30
|
---|
| 217 | #endif
|
---|
[821] | 218 |
|
---|
[1765] | 219 |
|
---|
| 220 | /* ===========================================================================
|
---|
| 221 | * Compile with MEDIUM_MEM to reduce the memory requirements or
|
---|
| 222 | * with SMALL_MEM to use as little memory as possible. Use BIG_MEM if the
|
---|
| 223 | * entire input file can be held in memory (not possible on 16 bit systems).
|
---|
| 224 | * Warning: defining these symbols affects HASH_BITS (see below) and thus
|
---|
| 225 | * affects the compression ratio. The compressed output
|
---|
| 226 | * is still correct, and might even be smaller in some cases.
|
---|
[821] | 227 | */
|
---|
| 228 |
|
---|
[1765] | 229 | #ifdef SMALL_MEM
|
---|
| 230 | # define HASH_BITS 13 /* Number of bits used to hash strings */
|
---|
[821] | 231 | #endif
|
---|
[1765] | 232 | #ifdef MEDIUM_MEM
|
---|
| 233 | # define HASH_BITS 14
|
---|
[821] | 234 | #endif
|
---|
[1765] | 235 | #ifndef HASH_BITS
|
---|
| 236 | # define HASH_BITS 15
|
---|
| 237 | /* For portability to 16 bit machines, do not use values above 15. */
|
---|
[821] | 238 | #endif
|
---|
| 239 |
|
---|
[1765] | 240 | #define HASH_SIZE (unsigned)(1<<HASH_BITS)
|
---|
| 241 | #define HASH_MASK (HASH_SIZE-1)
|
---|
| 242 | #define WMASK (WSIZE-1)
|
---|
| 243 | /* HASH_SIZE and WSIZE must be powers of two */
|
---|
| 244 | #ifndef TOO_FAR
|
---|
| 245 | # define TOO_FAR 4096
|
---|
[821] | 246 | #endif
|
---|
[1765] | 247 | /* Matches of length 3 are discarded if their distance exceeds TOO_FAR */
|
---|
[821] | 248 |
|
---|
| 249 |
|
---|
[1765] | 250 | /* ===========================================================================
|
---|
| 251 | * These types are not really 'char', 'short' and 'long'
|
---|
| 252 | */
|
---|
| 253 | typedef uint8_t uch;
|
---|
| 254 | typedef uint16_t ush;
|
---|
| 255 | typedef uint32_t ulg;
|
---|
| 256 | typedef int32_t lng;
|
---|
[821] | 257 |
|
---|
[1765] | 258 | typedef ush Pos;
|
---|
| 259 | typedef unsigned IPos;
|
---|
| 260 | /* A Pos is an index in the character window. We use short instead of int to
|
---|
| 261 | * save space in the various tables. IPos is used only for parameter passing.
|
---|
| 262 | */
|
---|
[821] | 263 |
|
---|
[1765] | 264 | enum {
|
---|
| 265 | WINDOW_SIZE = 2 * WSIZE,
|
---|
| 266 | /* window size, 2*WSIZE except for MMAP or BIG_MEM, where it is the
|
---|
| 267 | * input file length plus MIN_LOOKAHEAD.
|
---|
| 268 | */
|
---|
[821] | 269 |
|
---|
[3621] | 270 | #ifndef ENABLE_FEATURE_GZIP_LEVELS
|
---|
| 271 |
|
---|
[1765] | 272 | max_chain_length = 4096,
|
---|
| 273 | /* To speed up deflation, hash chains are never searched beyond this length.
|
---|
| 274 | * A higher limit improves compression ratio but degrades the speed.
|
---|
| 275 | */
|
---|
[821] | 276 |
|
---|
[1765] | 277 | max_lazy_match = 258,
|
---|
| 278 | /* Attempt to find a better match only when the current match is strictly
|
---|
| 279 | * smaller than this value. This mechanism is used only for compression
|
---|
| 280 | * levels >= 4.
|
---|
| 281 | */
|
---|
[821] | 282 |
|
---|
[1765] | 283 | max_insert_length = max_lazy_match,
|
---|
| 284 | /* Insert new strings in the hash table only if the match length
|
---|
| 285 | * is not greater than this length. This saves time but degrades compression.
|
---|
| 286 | * max_insert_length is used only for compression levels <= 3.
|
---|
| 287 | */
|
---|
[821] | 288 |
|
---|
[1765] | 289 | good_match = 32,
|
---|
| 290 | /* Use a faster search when the previous match is longer than this */
|
---|
| 291 |
|
---|
| 292 | /* Values for max_lazy_match, good_match and max_chain_length, depending on
|
---|
| 293 | * the desired pack level (0..9). The values given below have been tuned to
|
---|
| 294 | * exclude worst case performance for pathological files. Better values may be
|
---|
| 295 | * found for specific files.
|
---|
[821] | 296 | */
|
---|
| 297 |
|
---|
[1765] | 298 | nice_match = 258, /* Stop searching when current match exceeds this */
|
---|
| 299 | /* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4
|
---|
| 300 | * For deflate_fast() (levels <= 3) good is ignored and lazy has a different
|
---|
| 301 | * meaning.
|
---|
[821] | 302 | */
|
---|
[3621] | 303 | #endif /* ENABLE_FEATURE_GZIP_LEVELS */
|
---|
[1765] | 304 | };
|
---|
[821] | 305 |
|
---|
[1765] | 306 |
|
---|
| 307 | struct globals {
|
---|
| 308 |
|
---|
[3621] | 309 | #ifdef ENABLE_FEATURE_GZIP_LEVELS
|
---|
| 310 | unsigned max_chain_length;
|
---|
| 311 | unsigned max_lazy_match;
|
---|
| 312 | unsigned good_match;
|
---|
| 313 | unsigned nice_match;
|
---|
| 314 | #define max_chain_length (G1.max_chain_length)
|
---|
| 315 | #define max_lazy_match (G1.max_lazy_match)
|
---|
| 316 | #define good_match (G1.good_match)
|
---|
| 317 | #define nice_match (G1.nice_match)
|
---|
| 318 | #endif
|
---|
| 319 |
|
---|
[1765] | 320 | lng block_start;
|
---|
| 321 |
|
---|
| 322 | /* window position at the beginning of the current output block. Gets
|
---|
| 323 | * negative when the window is moved backwards.
|
---|
[821] | 324 | */
|
---|
[1765] | 325 | unsigned ins_h; /* hash index of string to be inserted */
|
---|
[821] | 326 |
|
---|
[1765] | 327 | #define H_SHIFT ((HASH_BITS+MIN_MATCH-1) / MIN_MATCH)
|
---|
| 328 | /* Number of bits by which ins_h and del_h must be shifted at each
|
---|
| 329 | * input step. It must be such that after MIN_MATCH steps, the oldest
|
---|
| 330 | * byte no longer takes part in the hash key, that is:
|
---|
| 331 | * H_SHIFT * MIN_MATCH >= HASH_BITS
|
---|
| 332 | */
|
---|
[821] | 333 |
|
---|
[1765] | 334 | unsigned prev_length;
|
---|
| 335 |
|
---|
| 336 | /* Length of the best match at previous step. Matches not greater than this
|
---|
| 337 | * are discarded. This is used in the lazy match evaluation.
|
---|
[821] | 338 | */
|
---|
| 339 |
|
---|
[1765] | 340 | unsigned strstart; /* start of string to insert */
|
---|
| 341 | unsigned match_start; /* start of matching string */
|
---|
| 342 | unsigned lookahead; /* number of valid bytes ahead in window */
|
---|
[821] | 343 |
|
---|
[1765] | 344 | /* ===========================================================================
|
---|
[821] | 345 | */
|
---|
[1765] | 346 | #define DECLARE(type, array, size) \
|
---|
| 347 | type * array
|
---|
| 348 | #define ALLOC(type, array, size) \
|
---|
[2725] | 349 | array = xzalloc((size_t)(((size)+1L)/2) * 2*sizeof(type))
|
---|
[1765] | 350 | #define FREE(array) \
|
---|
| 351 | do { free(array); array = NULL; } while (0)
|
---|
[821] | 352 |
|
---|
[1765] | 353 | /* global buffers */
|
---|
[821] | 354 |
|
---|
[1765] | 355 | /* buffer for literals or lengths */
|
---|
| 356 | /* DECLARE(uch, l_buf, LIT_BUFSIZE); */
|
---|
| 357 | DECLARE(uch, l_buf, INBUFSIZ);
|
---|
| 358 |
|
---|
| 359 | DECLARE(ush, d_buf, DIST_BUFSIZE);
|
---|
| 360 | DECLARE(uch, outbuf, OUTBUFSIZ);
|
---|
| 361 |
|
---|
| 362 | /* Sliding window. Input bytes are read into the second half of the window,
|
---|
| 363 | * and move to the first half later to keep a dictionary of at least WSIZE
|
---|
| 364 | * bytes. With this organization, matches are limited to a distance of
|
---|
| 365 | * WSIZE-MAX_MATCH bytes, but this ensures that IO is always
|
---|
| 366 | * performed with a length multiple of the block size. Also, it limits
|
---|
| 367 | * the window size to 64K, which is quite useful on MSDOS.
|
---|
| 368 | * To do: limit the window size to WSIZE+BSZ if SMALL_MEM (the code would
|
---|
| 369 | * be less efficient).
|
---|
[821] | 370 | */
|
---|
[1765] | 371 | DECLARE(uch, window, 2L * WSIZE);
|
---|
[821] | 372 |
|
---|
[1765] | 373 | /* Link to older string with same hash index. To limit the size of this
|
---|
| 374 | * array to 64K, this link is maintained only for the last 32K strings.
|
---|
| 375 | * An index in this array is thus a window index modulo 32K.
|
---|
| 376 | */
|
---|
| 377 | /* DECLARE(Pos, prev, WSIZE); */
|
---|
| 378 | DECLARE(ush, prev, 1L << BITS);
|
---|
| 379 |
|
---|
| 380 | /* Heads of the hash chains or 0. */
|
---|
| 381 | /* DECLARE(Pos, head, 1<<HASH_BITS); */
|
---|
| 382 | #define head (G1.prev + WSIZE) /* hash head (see deflate.c) */
|
---|
| 383 |
|
---|
| 384 | /* number of input bytes */
|
---|
| 385 | ulg isize; /* only 32 bits stored in .gz file */
|
---|
| 386 |
|
---|
| 387 | /* bbox always use stdin/stdout */
|
---|
| 388 | #define ifd STDIN_FILENO /* input file descriptor */
|
---|
| 389 | #define ofd STDOUT_FILENO /* output file descriptor */
|
---|
| 390 |
|
---|
| 391 | #ifdef DEBUG
|
---|
| 392 | unsigned insize; /* valid bytes in l_buf */
|
---|
| 393 | #endif
|
---|
| 394 | unsigned outcnt; /* bytes in output buffer */
|
---|
| 395 |
|
---|
| 396 | smallint eofile; /* flag set at end of input file */
|
---|
| 397 |
|
---|
[821] | 398 | /* ===========================================================================
|
---|
| 399 | * Local data used by the "bit string" routines.
|
---|
| 400 | */
|
---|
| 401 |
|
---|
[1765] | 402 | unsigned short bi_buf;
|
---|
[821] | 403 |
|
---|
| 404 | /* Output buffer. bits are inserted starting at the bottom (least significant
|
---|
| 405 | * bits).
|
---|
| 406 | */
|
---|
| 407 |
|
---|
[1765] | 408 | #undef BUF_SIZE
|
---|
| 409 | #define BUF_SIZE (8 * sizeof(G1.bi_buf))
|
---|
[821] | 410 | /* Number of bits used within bi_buf. (bi_buf might be implemented on
|
---|
| 411 | * more than 16 bits on some systems.)
|
---|
| 412 | */
|
---|
| 413 |
|
---|
[1765] | 414 | int bi_valid;
|
---|
[821] | 415 |
|
---|
| 416 | /* Current input function. Set to mem_read for in-memory compression */
|
---|
| 417 |
|
---|
| 418 | #ifdef DEBUG
|
---|
[1765] | 419 | ulg bits_sent; /* bit length of the compressed data */
|
---|
[821] | 420 | #endif
|
---|
| 421 |
|
---|
[2725] | 422 | /*uint32_t *crc_32_tab;*/
|
---|
[1765] | 423 | uint32_t crc; /* shift register contents */
|
---|
| 424 | };
|
---|
| 425 |
|
---|
| 426 | #define G1 (*(ptr_to_globals - 1))
|
---|
| 427 |
|
---|
| 428 |
|
---|
[821] | 429 | /* ===========================================================================
|
---|
[1765] | 430 | * Write the output buffer outbuf[0..outcnt-1] and update bytes_out.
|
---|
| 431 | * (used for the compressed data only)
|
---|
[821] | 432 | */
|
---|
[1765] | 433 | static void flush_outbuf(void)
|
---|
[821] | 434 | {
|
---|
[1765] | 435 | if (G1.outcnt == 0)
|
---|
| 436 | return;
|
---|
| 437 |
|
---|
| 438 | xwrite(ofd, (char *) G1.outbuf, G1.outcnt);
|
---|
| 439 | G1.outcnt = 0;
|
---|
| 440 | }
|
---|
| 441 |
|
---|
| 442 |
|
---|
| 443 | /* ===========================================================================
|
---|
| 444 | */
|
---|
| 445 | /* put_8bit is used for the compressed output */
|
---|
| 446 | #define put_8bit(c) \
|
---|
| 447 | do { \
|
---|
| 448 | G1.outbuf[G1.outcnt++] = (c); \
|
---|
[3621] | 449 | if (G1.outcnt == OUTBUFSIZ) \
|
---|
| 450 | flush_outbuf(); \
|
---|
[1765] | 451 | } while (0)
|
---|
| 452 |
|
---|
| 453 | /* Output a 16 bit value, lsb first */
|
---|
| 454 | static void put_16bit(ush w)
|
---|
| 455 | {
|
---|
[3621] | 456 | /* GCC 4.2.1 won't optimize out redundant loads of G1.outcnt
|
---|
| 457 | * (probably because of fear of aliasing with G1.outbuf[]
|
---|
| 458 | * stores), do it explicitly:
|
---|
| 459 | */
|
---|
| 460 | unsigned outcnt = G1.outcnt;
|
---|
| 461 | uch *dst = &G1.outbuf[outcnt];
|
---|
| 462 |
|
---|
| 463 | #if BB_UNALIGNED_MEMACCESS_OK && BB_LITTLE_ENDIAN
|
---|
| 464 | if (outcnt < OUTBUFSIZ-2) {
|
---|
| 465 | /* Common case */
|
---|
| 466 | ush *dst16 = (void*) dst;
|
---|
| 467 | *dst16 = w; /* unalinged LSB 16-bit store */
|
---|
| 468 | G1.outcnt = outcnt + 2;
|
---|
| 469 | return;
|
---|
[1765] | 470 | }
|
---|
[3621] | 471 | *dst = (uch)w;
|
---|
| 472 | w >>= 8;
|
---|
| 473 | #else
|
---|
| 474 | *dst = (uch)w;
|
---|
| 475 | w >>= 8;
|
---|
| 476 | if (outcnt < OUTBUFSIZ-2) {
|
---|
| 477 | /* Common case */
|
---|
| 478 | dst[1] = w;
|
---|
| 479 | G1.outcnt = outcnt + 2;
|
---|
| 480 | return;
|
---|
| 481 | }
|
---|
| 482 | #endif
|
---|
| 483 |
|
---|
| 484 | /* Slowpath: we will need to do flush_outbuf() */
|
---|
| 485 | G1.outcnt = ++outcnt;
|
---|
| 486 | if (outcnt == OUTBUFSIZ)
|
---|
| 487 | flush_outbuf();
|
---|
| 488 | put_8bit(w);
|
---|
[1765] | 489 | }
|
---|
| 490 |
|
---|
| 491 | static void put_32bit(ulg n)
|
---|
| 492 | {
|
---|
| 493 | put_16bit(n);
|
---|
| 494 | put_16bit(n >> 16);
|
---|
| 495 | }
|
---|
| 496 |
|
---|
| 497 | /* ===========================================================================
|
---|
| 498 | * Run a set of bytes through the crc shift register. If s is a NULL
|
---|
| 499 | * pointer, then initialize the crc shift register contents instead.
|
---|
| 500 | * Return the current crc in either case.
|
---|
| 501 | */
|
---|
[2725] | 502 | static void updcrc(uch * s, unsigned n)
|
---|
[1765] | 503 | {
|
---|
[2725] | 504 | G1.crc = crc32_block_endian0(G1.crc, s, n, global_crc32_table /*G1.crc_32_tab*/);
|
---|
[821] | 505 | }
|
---|
| 506 |
|
---|
[1765] | 507 |
|
---|
[821] | 508 | /* ===========================================================================
|
---|
[1765] | 509 | * Read a new buffer from the current input file, perform end-of-line
|
---|
| 510 | * translation, and update the crc and input file size.
|
---|
| 511 | * IN assertion: size >= 2 (for end-of-line translation)
|
---|
| 512 | */
|
---|
| 513 | static unsigned file_read(void *buf, unsigned size)
|
---|
| 514 | {
|
---|
| 515 | unsigned len;
|
---|
| 516 |
|
---|
| 517 | Assert(G1.insize == 0, "l_buf not empty");
|
---|
| 518 |
|
---|
| 519 | len = safe_read(ifd, buf, size);
|
---|
| 520 | if (len == (unsigned)(-1) || len == 0)
|
---|
| 521 | return len;
|
---|
| 522 |
|
---|
| 523 | updcrc(buf, len);
|
---|
| 524 | G1.isize += len;
|
---|
| 525 | return len;
|
---|
| 526 | }
|
---|
| 527 |
|
---|
| 528 |
|
---|
| 529 | /* ===========================================================================
|
---|
[821] | 530 | * Send a value on a given number of bits.
|
---|
| 531 | * IN assertion: length <= 16 and value fits in length bits.
|
---|
| 532 | */
|
---|
| 533 | static void send_bits(int value, int length)
|
---|
| 534 | {
|
---|
| 535 | #ifdef DEBUG
|
---|
| 536 | Tracev((stderr, " l %2d v %4x ", length, value));
|
---|
| 537 | Assert(length > 0 && length <= 15, "invalid length");
|
---|
[1765] | 538 | G1.bits_sent += length;
|
---|
[821] | 539 | #endif
|
---|
| 540 | /* If not enough room in bi_buf, use (valid) bits from bi_buf and
|
---|
| 541 | * (16 - bi_valid) bits from value, leaving (width - (16-bi_valid))
|
---|
| 542 | * unused bits in value.
|
---|
| 543 | */
|
---|
[1765] | 544 | if (G1.bi_valid > (int) BUF_SIZE - length) {
|
---|
| 545 | G1.bi_buf |= (value << G1.bi_valid);
|
---|
| 546 | put_16bit(G1.bi_buf);
|
---|
| 547 | G1.bi_buf = (ush) value >> (BUF_SIZE - G1.bi_valid);
|
---|
| 548 | G1.bi_valid += length - BUF_SIZE;
|
---|
[821] | 549 | } else {
|
---|
[1765] | 550 | G1.bi_buf |= value << G1.bi_valid;
|
---|
| 551 | G1.bi_valid += length;
|
---|
[821] | 552 | }
|
---|
| 553 | }
|
---|
| 554 |
|
---|
[1765] | 555 |
|
---|
[821] | 556 | /* ===========================================================================
|
---|
| 557 | * Reverse the first len bits of a code, using straightforward code (a faster
|
---|
| 558 | * method would use a table)
|
---|
| 559 | * IN assertion: 1 <= len <= 15
|
---|
| 560 | */
|
---|
| 561 | static unsigned bi_reverse(unsigned code, int len)
|
---|
| 562 | {
|
---|
[1765] | 563 | unsigned res = 0;
|
---|
[821] | 564 |
|
---|
[1765] | 565 | while (1) {
|
---|
[821] | 566 | res |= code & 1;
|
---|
[1765] | 567 | if (--len <= 0) return res;
|
---|
| 568 | code >>= 1;
|
---|
| 569 | res <<= 1;
|
---|
| 570 | }
|
---|
[821] | 571 | }
|
---|
| 572 |
|
---|
[1765] | 573 |
|
---|
[821] | 574 | /* ===========================================================================
|
---|
| 575 | * Write out any remaining bits in an incomplete byte.
|
---|
| 576 | */
|
---|
| 577 | static void bi_windup(void)
|
---|
| 578 | {
|
---|
[1765] | 579 | if (G1.bi_valid > 8) {
|
---|
| 580 | put_16bit(G1.bi_buf);
|
---|
| 581 | } else if (G1.bi_valid > 0) {
|
---|
| 582 | put_8bit(G1.bi_buf);
|
---|
[821] | 583 | }
|
---|
[1765] | 584 | G1.bi_buf = 0;
|
---|
| 585 | G1.bi_valid = 0;
|
---|
[821] | 586 | #ifdef DEBUG
|
---|
[1765] | 587 | G1.bits_sent = (G1.bits_sent + 7) & ~7;
|
---|
[821] | 588 | #endif
|
---|
| 589 | }
|
---|
| 590 |
|
---|
[1765] | 591 |
|
---|
[821] | 592 | /* ===========================================================================
|
---|
| 593 | * Copy a stored block to the zip file, storing first the length and its
|
---|
| 594 | * one's complement if requested.
|
---|
| 595 | */
|
---|
| 596 | static void copy_block(char *buf, unsigned len, int header)
|
---|
| 597 | {
|
---|
| 598 | bi_windup(); /* align on byte boundary */
|
---|
| 599 |
|
---|
| 600 | if (header) {
|
---|
[1765] | 601 | put_16bit(len);
|
---|
| 602 | put_16bit(~len);
|
---|
[821] | 603 | #ifdef DEBUG
|
---|
[1765] | 604 | G1.bits_sent += 2 * 16;
|
---|
[821] | 605 | #endif
|
---|
| 606 | }
|
---|
| 607 | #ifdef DEBUG
|
---|
[1765] | 608 | G1.bits_sent += (ulg) len << 3;
|
---|
[821] | 609 | #endif
|
---|
| 610 | while (len--) {
|
---|
[1765] | 611 | put_8bit(*buf++);
|
---|
[821] | 612 | }
|
---|
| 613 | }
|
---|
| 614 |
|
---|
| 615 |
|
---|
| 616 | /* ===========================================================================
|
---|
[1765] | 617 | * Fill the window when the lookahead becomes insufficient.
|
---|
| 618 | * Updates strstart and lookahead, and sets eofile if end of input file.
|
---|
| 619 | * IN assertion: lookahead < MIN_LOOKAHEAD && strstart + lookahead > 0
|
---|
| 620 | * OUT assertions: at least one byte has been read, or eofile is set;
|
---|
| 621 | * file reads are performed for at least two bytes (required for the
|
---|
| 622 | * translate_eol option).
|
---|
[821] | 623 | */
|
---|
[1765] | 624 | static void fill_window(void)
|
---|
[821] | 625 | {
|
---|
[1765] | 626 | unsigned n, m;
|
---|
| 627 | unsigned more = WINDOW_SIZE - G1.lookahead - G1.strstart;
|
---|
| 628 | /* Amount of free space at the end of the window. */
|
---|
[821] | 629 |
|
---|
[1765] | 630 | /* If the window is almost full and there is insufficient lookahead,
|
---|
| 631 | * move the upper half to the lower one to make room in the upper half.
|
---|
| 632 | */
|
---|
| 633 | if (more == (unsigned) -1) {
|
---|
| 634 | /* Very unlikely, but possible on 16 bit machine if strstart == 0
|
---|
| 635 | * and lookahead == 1 (input done one byte at time)
|
---|
| 636 | */
|
---|
| 637 | more--;
|
---|
| 638 | } else if (G1.strstart >= WSIZE + MAX_DIST) {
|
---|
| 639 | /* By the IN assertion, the window is not empty so we can't confuse
|
---|
| 640 | * more == 0 with more == 64K on a 16 bit machine.
|
---|
| 641 | */
|
---|
| 642 | Assert(WINDOW_SIZE == 2 * WSIZE, "no sliding with BIG_MEM");
|
---|
[821] | 643 |
|
---|
[1765] | 644 | memcpy(G1.window, G1.window + WSIZE, WSIZE);
|
---|
| 645 | G1.match_start -= WSIZE;
|
---|
| 646 | G1.strstart -= WSIZE; /* we now have strstart >= MAX_DIST: */
|
---|
[821] | 647 |
|
---|
[1765] | 648 | G1.block_start -= WSIZE;
|
---|
[821] | 649 |
|
---|
[1765] | 650 | for (n = 0; n < HASH_SIZE; n++) {
|
---|
| 651 | m = head[n];
|
---|
| 652 | head[n] = (Pos) (m >= WSIZE ? m - WSIZE : 0);
|
---|
| 653 | }
|
---|
| 654 | for (n = 0; n < WSIZE; n++) {
|
---|
| 655 | m = G1.prev[n];
|
---|
| 656 | G1.prev[n] = (Pos) (m >= WSIZE ? m - WSIZE : 0);
|
---|
| 657 | /* If n is not on any hash chain, prev[n] is garbage but
|
---|
| 658 | * its value will never be used.
|
---|
| 659 | */
|
---|
| 660 | }
|
---|
| 661 | more += WSIZE;
|
---|
[821] | 662 | }
|
---|
[1765] | 663 | /* At this point, more >= 2 */
|
---|
| 664 | if (!G1.eofile) {
|
---|
| 665 | n = file_read(G1.window + G1.strstart + G1.lookahead, more);
|
---|
| 666 | if (n == 0 || n == (unsigned) -1) {
|
---|
| 667 | G1.eofile = 1;
|
---|
| 668 | } else {
|
---|
| 669 | G1.lookahead += n;
|
---|
| 670 | }
|
---|
| 671 | }
|
---|
[821] | 672 | }
|
---|
| 673 |
|
---|
[1765] | 674 |
|
---|
[821] | 675 | /* ===========================================================================
|
---|
| 676 | * Set match_start to the longest match starting at the given string and
|
---|
| 677 | * return its length. Matches shorter or equal to prev_length are discarded,
|
---|
| 678 | * in which case the result is equal to prev_length and match_start is
|
---|
| 679 | * garbage.
|
---|
| 680 | * IN assertions: cur_match is the head of the hash chain for the current
|
---|
| 681 | * string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1
|
---|
| 682 | */
|
---|
| 683 |
|
---|
| 684 | /* For MSDOS, OS/2 and 386 Unix, an optimized version is in match.asm or
|
---|
| 685 | * match.s. The code is functionally equivalent, so you can use the C version
|
---|
| 686 | * if desired.
|
---|
| 687 | */
|
---|
| 688 | static int longest_match(IPos cur_match)
|
---|
| 689 | {
|
---|
| 690 | unsigned chain_length = max_chain_length; /* max hash chain length */
|
---|
[1765] | 691 | uch *scan = G1.window + G1.strstart; /* current string */
|
---|
| 692 | uch *match; /* matched string */
|
---|
| 693 | int len; /* length of current match */
|
---|
| 694 | int best_len = G1.prev_length; /* best match length so far */
|
---|
| 695 | IPos limit = G1.strstart > (IPos) MAX_DIST ? G1.strstart - (IPos) MAX_DIST : 0;
|
---|
[821] | 696 | /* Stop when cur_match becomes <= limit. To simplify the code,
|
---|
| 697 | * we prevent matches with the string of window index 0.
|
---|
| 698 | */
|
---|
| 699 |
|
---|
| 700 | /* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16.
|
---|
| 701 | * It is easy to get rid of this optimization if necessary.
|
---|
| 702 | */
|
---|
| 703 | #if HASH_BITS < 8 || MAX_MATCH != 258
|
---|
| 704 | # error Code too clever
|
---|
| 705 | #endif
|
---|
[1765] | 706 | uch *strend = G1.window + G1.strstart + MAX_MATCH;
|
---|
| 707 | uch scan_end1 = scan[best_len - 1];
|
---|
| 708 | uch scan_end = scan[best_len];
|
---|
[821] | 709 |
|
---|
| 710 | /* Do not waste too much time if we already have a good match: */
|
---|
[1765] | 711 | if (G1.prev_length >= good_match) {
|
---|
[821] | 712 | chain_length >>= 2;
|
---|
| 713 | }
|
---|
[1765] | 714 | Assert(G1.strstart <= WINDOW_SIZE - MIN_LOOKAHEAD, "insufficient lookahead");
|
---|
[821] | 715 |
|
---|
| 716 | do {
|
---|
[1765] | 717 | Assert(cur_match < G1.strstart, "no future");
|
---|
| 718 | match = G1.window + cur_match;
|
---|
[821] | 719 |
|
---|
| 720 | /* Skip to next match if the match length cannot increase
|
---|
| 721 | * or if the match length is less than 2:
|
---|
| 722 | */
|
---|
[2725] | 723 | if (match[best_len] != scan_end
|
---|
| 724 | || match[best_len - 1] != scan_end1
|
---|
| 725 | || *match != *scan || *++match != scan[1]
|
---|
| 726 | ) {
|
---|
[821] | 727 | continue;
|
---|
[2725] | 728 | }
|
---|
[821] | 729 |
|
---|
| 730 | /* The check at best_len-1 can be removed because it will be made
|
---|
| 731 | * again later. (This heuristic is not always a win.)
|
---|
| 732 | * It is not necessary to compare scan[2] and match[2] since they
|
---|
| 733 | * are always equal when the other bytes match, given that
|
---|
| 734 | * the hash keys are equal and that HASH_BITS >= 8.
|
---|
| 735 | */
|
---|
| 736 | scan += 2, match++;
|
---|
| 737 |
|
---|
| 738 | /* We check for insufficient lookahead only every 8th comparison;
|
---|
| 739 | * the 256th check will be made at strstart+258.
|
---|
| 740 | */
|
---|
| 741 | do {
|
---|
| 742 | } while (*++scan == *++match && *++scan == *++match &&
|
---|
| 743 | *++scan == *++match && *++scan == *++match &&
|
---|
| 744 | *++scan == *++match && *++scan == *++match &&
|
---|
| 745 | *++scan == *++match && *++scan == *++match && scan < strend);
|
---|
| 746 |
|
---|
| 747 | len = MAX_MATCH - (int) (strend - scan);
|
---|
| 748 | scan = strend - MAX_MATCH;
|
---|
| 749 |
|
---|
| 750 | if (len > best_len) {
|
---|
[1765] | 751 | G1.match_start = cur_match;
|
---|
[821] | 752 | best_len = len;
|
---|
| 753 | if (len >= nice_match)
|
---|
| 754 | break;
|
---|
| 755 | scan_end1 = scan[best_len - 1];
|
---|
| 756 | scan_end = scan[best_len];
|
---|
| 757 | }
|
---|
[1765] | 758 | } while ((cur_match = G1.prev[cur_match & WMASK]) > limit
|
---|
[821] | 759 | && --chain_length != 0);
|
---|
| 760 |
|
---|
| 761 | return best_len;
|
---|
| 762 | }
|
---|
| 763 |
|
---|
[1765] | 764 |
|
---|
[821] | 765 | #ifdef DEBUG
|
---|
| 766 | /* ===========================================================================
|
---|
| 767 | * Check that the match at match_start is indeed a match.
|
---|
| 768 | */
|
---|
| 769 | static void check_match(IPos start, IPos match, int length)
|
---|
| 770 | {
|
---|
| 771 | /* check that the match is indeed a match */
|
---|
[1765] | 772 | if (memcmp(G1.window + match, G1.window + start, length) != 0) {
|
---|
[821] | 773 | bb_error_msg(" start %d, match %d, length %d", start, match, length);
|
---|
| 774 | bb_error_msg("invalid match");
|
---|
| 775 | }
|
---|
| 776 | if (verbose > 1) {
|
---|
| 777 | bb_error_msg("\\[%d,%d]", start - match, length);
|
---|
| 778 | do {
|
---|
[2725] | 779 | bb_putchar_stderr(G1.window[start++]);
|
---|
[821] | 780 | } while (--length != 0);
|
---|
| 781 | }
|
---|
| 782 | }
|
---|
| 783 | #else
|
---|
[1765] | 784 | # define check_match(start, match, length) ((void)0)
|
---|
[821] | 785 | #endif
|
---|
| 786 |
|
---|
| 787 |
|
---|
| 788 | /* trees.c -- output deflated data using Huffman coding
|
---|
| 789 | * Copyright (C) 1992-1993 Jean-loup Gailly
|
---|
| 790 | * This is free software; you can redistribute it and/or modify it under the
|
---|
| 791 | * terms of the GNU General Public License, see the file COPYING.
|
---|
| 792 | */
|
---|
| 793 |
|
---|
[1765] | 794 | /* PURPOSE
|
---|
[821] | 795 | * Encode various sets of source values using variable-length
|
---|
| 796 | * binary code trees.
|
---|
| 797 | *
|
---|
| 798 | * DISCUSSION
|
---|
| 799 | * The PKZIP "deflation" process uses several Huffman trees. The more
|
---|
| 800 | * common source values are represented by shorter bit sequences.
|
---|
| 801 | *
|
---|
| 802 | * Each code tree is stored in the ZIP file in a compressed form
|
---|
| 803 | * which is itself a Huffman encoding of the lengths of
|
---|
| 804 | * all the code strings (in ascending order by source values).
|
---|
| 805 | * The actual code strings are reconstructed from the lengths in
|
---|
| 806 | * the UNZIP process, as described in the "application note"
|
---|
| 807 | * (APPNOTE.TXT) distributed as part of PKWARE's PKZIP program.
|
---|
| 808 | *
|
---|
| 809 | * REFERENCES
|
---|
| 810 | * Lynch, Thomas J.
|
---|
| 811 | * Data Compression: Techniques and Applications, pp. 53-55.
|
---|
| 812 | * Lifetime Learning Publications, 1985. ISBN 0-534-03418-7.
|
---|
| 813 | *
|
---|
| 814 | * Storer, James A.
|
---|
| 815 | * Data Compression: Methods and Theory, pp. 49-50.
|
---|
| 816 | * Computer Science Press, 1988. ISBN 0-7167-8156-5.
|
---|
| 817 | *
|
---|
| 818 | * Sedgewick, R.
|
---|
| 819 | * Algorithms, p290.
|
---|
| 820 | * Addison-Wesley, 1983. ISBN 0-201-06672-6.
|
---|
| 821 | *
|
---|
| 822 | * INTERFACE
|
---|
[1765] | 823 | * void ct_init()
|
---|
| 824 | * Allocate the match buffer, initialize the various tables [and save
|
---|
[821] | 825 | * the location of the internal file attribute (ascii/binary) and
|
---|
[1765] | 826 | * method (DEFLATE/STORE) -- deleted in bbox]
|
---|
[821] | 827 | *
|
---|
[1765] | 828 | * void ct_tally(int dist, int lc);
|
---|
[821] | 829 | * Save the match info and tally the frequency counts.
|
---|
| 830 | *
|
---|
[1765] | 831 | * ulg flush_block(char *buf, ulg stored_len, int eof)
|
---|
[821] | 832 | * Determine the best encoding for the current block: dynamic trees,
|
---|
| 833 | * static trees or store, and output the encoded block to the zip
|
---|
| 834 | * file. Returns the total compressed length for the file so far.
|
---|
| 835 | */
|
---|
| 836 |
|
---|
| 837 | #define MAX_BITS 15
|
---|
| 838 | /* All codes must not exceed MAX_BITS bits */
|
---|
| 839 |
|
---|
| 840 | #define MAX_BL_BITS 7
|
---|
| 841 | /* Bit length codes must not exceed MAX_BL_BITS bits */
|
---|
| 842 |
|
---|
| 843 | #define LENGTH_CODES 29
|
---|
| 844 | /* number of length codes, not counting the special END_BLOCK code */
|
---|
| 845 |
|
---|
| 846 | #define LITERALS 256
|
---|
| 847 | /* number of literal bytes 0..255 */
|
---|
| 848 |
|
---|
| 849 | #define END_BLOCK 256
|
---|
| 850 | /* end of block literal code */
|
---|
| 851 |
|
---|
| 852 | #define L_CODES (LITERALS+1+LENGTH_CODES)
|
---|
| 853 | /* number of Literal or Length codes, including the END_BLOCK code */
|
---|
| 854 |
|
---|
| 855 | #define D_CODES 30
|
---|
| 856 | /* number of distance codes */
|
---|
| 857 |
|
---|
| 858 | #define BL_CODES 19
|
---|
| 859 | /* number of codes used to transfer the bit lengths */
|
---|
| 860 |
|
---|
| 861 | /* extra bits for each length code */
|
---|
[1765] | 862 | static const uint8_t extra_lbits[LENGTH_CODES] ALIGN1 = {
|
---|
| 863 | 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4,
|
---|
[821] | 864 | 4, 4, 5, 5, 5, 5, 0
|
---|
| 865 | };
|
---|
| 866 |
|
---|
| 867 | /* extra bits for each distance code */
|
---|
[1765] | 868 | static const uint8_t extra_dbits[D_CODES] ALIGN1 = {
|
---|
| 869 | 0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9,
|
---|
[821] | 870 | 10, 10, 11, 11, 12, 12, 13, 13
|
---|
| 871 | };
|
---|
| 872 |
|
---|
| 873 | /* extra bits for each bit length code */
|
---|
[1765] | 874 | static const uint8_t extra_blbits[BL_CODES] ALIGN1 = {
|
---|
| 875 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 3, 7 };
|
---|
[821] | 876 |
|
---|
[1765] | 877 | /* number of codes at each bit length for an optimal tree */
|
---|
| 878 | static const uint8_t bl_order[BL_CODES] ALIGN1 = {
|
---|
| 879 | 16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15 };
|
---|
| 880 |
|
---|
[821] | 881 | #define STORED_BLOCK 0
|
---|
| 882 | #define STATIC_TREES 1
|
---|
| 883 | #define DYN_TREES 2
|
---|
| 884 | /* The three kinds of block type */
|
---|
| 885 |
|
---|
| 886 | #ifndef LIT_BUFSIZE
|
---|
| 887 | # ifdef SMALL_MEM
|
---|
| 888 | # define LIT_BUFSIZE 0x2000
|
---|
| 889 | # else
|
---|
| 890 | # ifdef MEDIUM_MEM
|
---|
| 891 | # define LIT_BUFSIZE 0x4000
|
---|
| 892 | # else
|
---|
| 893 | # define LIT_BUFSIZE 0x8000
|
---|
| 894 | # endif
|
---|
| 895 | # endif
|
---|
| 896 | #endif
|
---|
| 897 | #ifndef DIST_BUFSIZE
|
---|
| 898 | # define DIST_BUFSIZE LIT_BUFSIZE
|
---|
| 899 | #endif
|
---|
| 900 | /* Sizes of match buffers for literals/lengths and distances. There are
|
---|
| 901 | * 4 reasons for limiting LIT_BUFSIZE to 64K:
|
---|
| 902 | * - frequencies can be kept in 16 bit counters
|
---|
| 903 | * - if compression is not successful for the first block, all input data is
|
---|
| 904 | * still in the window so we can still emit a stored block even when input
|
---|
| 905 | * comes from standard input. (This can also be done for all blocks if
|
---|
| 906 | * LIT_BUFSIZE is not greater than 32K.)
|
---|
| 907 | * - if compression is not successful for a file smaller than 64K, we can
|
---|
| 908 | * even emit a stored file instead of a stored block (saving 5 bytes).
|
---|
| 909 | * - creating new Huffman trees less frequently may not provide fast
|
---|
| 910 | * adaptation to changes in the input data statistics. (Take for
|
---|
| 911 | * example a binary file with poorly compressible code followed by
|
---|
| 912 | * a highly compressible string table.) Smaller buffer sizes give
|
---|
| 913 | * fast adaptation but have of course the overhead of transmitting trees
|
---|
| 914 | * more frequently.
|
---|
| 915 | * - I can't count above 4
|
---|
| 916 | * The current code is general and allows DIST_BUFSIZE < LIT_BUFSIZE (to save
|
---|
| 917 | * memory at the expense of compression). Some optimizations would be possible
|
---|
| 918 | * if we rely on DIST_BUFSIZE == LIT_BUFSIZE.
|
---|
| 919 | */
|
---|
| 920 | #define REP_3_6 16
|
---|
| 921 | /* repeat previous bit length 3-6 times (2 bits of repeat count) */
|
---|
| 922 | #define REPZ_3_10 17
|
---|
| 923 | /* repeat a zero length 3-10 times (3 bits of repeat count) */
|
---|
| 924 | #define REPZ_11_138 18
|
---|
| 925 | /* repeat a zero length 11-138 times (7 bits of repeat count) */
|
---|
| 926 |
|
---|
| 927 | /* ===========================================================================
|
---|
[1765] | 928 | */
|
---|
[821] | 929 | /* Data structure describing a single value and its code string. */
|
---|
| 930 | typedef struct ct_data {
|
---|
| 931 | union {
|
---|
| 932 | ush freq; /* frequency count */
|
---|
| 933 | ush code; /* bit string */
|
---|
| 934 | } fc;
|
---|
| 935 | union {
|
---|
| 936 | ush dad; /* father node in Huffman tree */
|
---|
| 937 | ush len; /* length of bit string */
|
---|
| 938 | } dl;
|
---|
| 939 | } ct_data;
|
---|
| 940 |
|
---|
| 941 | #define Freq fc.freq
|
---|
| 942 | #define Code fc.code
|
---|
| 943 | #define Dad dl.dad
|
---|
| 944 | #define Len dl.len
|
---|
| 945 |
|
---|
[1765] | 946 | #define HEAP_SIZE (2*L_CODES + 1)
|
---|
[821] | 947 | /* maximum heap size */
|
---|
| 948 |
|
---|
[1765] | 949 | typedef struct tree_desc {
|
---|
| 950 | ct_data *dyn_tree; /* the dynamic tree */
|
---|
| 951 | ct_data *static_tree; /* corresponding static tree or NULL */
|
---|
| 952 | const uint8_t *extra_bits; /* extra bits for each code or NULL */
|
---|
| 953 | int extra_base; /* base index for extra_bits */
|
---|
| 954 | int elems; /* max number of elements in the tree */
|
---|
| 955 | int max_length; /* max bit length for the codes */
|
---|
| 956 | int max_code; /* largest code with non zero frequency */
|
---|
| 957 | } tree_desc;
|
---|
[821] | 958 |
|
---|
[1765] | 959 | struct globals2 {
|
---|
[821] | 960 |
|
---|
[1765] | 961 | ush heap[HEAP_SIZE]; /* heap used to build the Huffman trees */
|
---|
| 962 | int heap_len; /* number of elements in the heap */
|
---|
| 963 | int heap_max; /* element of largest frequency */
|
---|
| 964 |
|
---|
| 965 | /* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used.
|
---|
| 966 | * The same heap array is used to build all trees.
|
---|
| 967 | */
|
---|
| 968 |
|
---|
| 969 | ct_data dyn_ltree[HEAP_SIZE]; /* literal and length tree */
|
---|
| 970 | ct_data dyn_dtree[2 * D_CODES + 1]; /* distance tree */
|
---|
| 971 |
|
---|
| 972 | ct_data static_ltree[L_CODES + 2];
|
---|
| 973 |
|
---|
[821] | 974 | /* The static literal tree. Since the bit lengths are imposed, there is no
|
---|
| 975 | * need for the L_CODES extra codes used during heap construction. However
|
---|
| 976 | * The codes 286 and 287 are needed to build a canonical tree (see ct_init
|
---|
| 977 | * below).
|
---|
| 978 | */
|
---|
| 979 |
|
---|
[1765] | 980 | ct_data static_dtree[D_CODES];
|
---|
[821] | 981 |
|
---|
| 982 | /* The static distance tree. (Actually a trivial tree since all codes use
|
---|
| 983 | * 5 bits.)
|
---|
| 984 | */
|
---|
| 985 |
|
---|
[1765] | 986 | ct_data bl_tree[2 * BL_CODES + 1];
|
---|
[821] | 987 |
|
---|
| 988 | /* Huffman tree for the bit lengths */
|
---|
| 989 |
|
---|
[1765] | 990 | tree_desc l_desc;
|
---|
| 991 | tree_desc d_desc;
|
---|
| 992 | tree_desc bl_desc;
|
---|
[821] | 993 |
|
---|
[1765] | 994 | ush bl_count[MAX_BITS + 1];
|
---|
[821] | 995 |
|
---|
| 996 | /* The lengths of the bit length codes are sent in order of decreasing
|
---|
| 997 | * probability, to avoid transmitting the lengths for unused bit length codes.
|
---|
| 998 | */
|
---|
| 999 |
|
---|
[1765] | 1000 | uch depth[2 * L_CODES + 1];
|
---|
[821] | 1001 |
|
---|
| 1002 | /* Depth of each subtree used as tie breaker for trees of equal frequency */
|
---|
| 1003 |
|
---|
[1765] | 1004 | uch length_code[MAX_MATCH - MIN_MATCH + 1];
|
---|
[821] | 1005 |
|
---|
| 1006 | /* length code for each normalized match length (0 == MIN_MATCH) */
|
---|
| 1007 |
|
---|
[1765] | 1008 | uch dist_code[512];
|
---|
[821] | 1009 |
|
---|
| 1010 | /* distance codes. The first 256 values correspond to the distances
|
---|
| 1011 | * 3 .. 258, the last 256 values correspond to the top 8 bits of
|
---|
| 1012 | * the 15 bit distances.
|
---|
| 1013 | */
|
---|
| 1014 |
|
---|
[1765] | 1015 | int base_length[LENGTH_CODES];
|
---|
[821] | 1016 |
|
---|
| 1017 | /* First normalized length for each code (0 = MIN_MATCH) */
|
---|
| 1018 |
|
---|
[1765] | 1019 | int base_dist[D_CODES];
|
---|
[821] | 1020 |
|
---|
| 1021 | /* First normalized distance for each code (0 = distance of 1) */
|
---|
| 1022 |
|
---|
[1765] | 1023 | uch flag_buf[LIT_BUFSIZE / 8];
|
---|
[821] | 1024 |
|
---|
| 1025 | /* flag_buf is a bit array distinguishing literals from lengths in
|
---|
| 1026 | * l_buf, thus indicating the presence or absence of a distance.
|
---|
| 1027 | */
|
---|
| 1028 |
|
---|
[1765] | 1029 | unsigned last_lit; /* running index in l_buf */
|
---|
| 1030 | unsigned last_dist; /* running index in d_buf */
|
---|
| 1031 | unsigned last_flags; /* running index in flag_buf */
|
---|
| 1032 | uch flags; /* current flags not yet saved in flag_buf */
|
---|
| 1033 | uch flag_bit; /* current bit used in flags */
|
---|
[821] | 1034 |
|
---|
| 1035 | /* bits are filled in flags starting at bit 0 (least significant).
|
---|
| 1036 | * Note: these flags are overkill in the current code since we don't
|
---|
| 1037 | * take advantage of DIST_BUFSIZE == LIT_BUFSIZE.
|
---|
| 1038 | */
|
---|
| 1039 |
|
---|
[1765] | 1040 | ulg opt_len; /* bit length of current block with optimal trees */
|
---|
| 1041 | ulg static_len; /* bit length of current block with static trees */
|
---|
[821] | 1042 |
|
---|
[1765] | 1043 | ulg compressed_len; /* total bit length of compressed file */
|
---|
| 1044 | };
|
---|
[821] | 1045 |
|
---|
[1765] | 1046 | #define G2ptr ((struct globals2*)(ptr_to_globals))
|
---|
| 1047 | #define G2 (*G2ptr)
|
---|
[821] | 1048 |
|
---|
| 1049 |
|
---|
| 1050 | /* ===========================================================================
|
---|
| 1051 | */
|
---|
| 1052 | static void gen_codes(ct_data * tree, int max_code);
|
---|
| 1053 | static void build_tree(tree_desc * desc);
|
---|
| 1054 | static void scan_tree(ct_data * tree, int max_code);
|
---|
| 1055 | static void send_tree(ct_data * tree, int max_code);
|
---|
| 1056 | static int build_bl_tree(void);
|
---|
| 1057 | static void send_all_trees(int lcodes, int dcodes, int blcodes);
|
---|
| 1058 | static void compress_block(ct_data * ltree, ct_data * dtree);
|
---|
| 1059 |
|
---|
| 1060 |
|
---|
| 1061 | #ifndef DEBUG
|
---|
[1765] | 1062 | /* Send a code of the given tree. c and tree must not have side effects */
|
---|
| 1063 | # define SEND_CODE(c, tree) send_bits(tree[c].Code, tree[c].Len)
|
---|
| 1064 | #else
|
---|
| 1065 | # define SEND_CODE(c, tree) \
|
---|
| 1066 | { \
|
---|
[2725] | 1067 | if (verbose > 1) bb_error_msg("\ncd %3d ", (c)); \
|
---|
[1765] | 1068 | send_bits(tree[c].Code, tree[c].Len); \
|
---|
| 1069 | }
|
---|
[821] | 1070 | #endif
|
---|
| 1071 |
|
---|
[1765] | 1072 | #define D_CODE(dist) \
|
---|
| 1073 | ((dist) < 256 ? G2.dist_code[dist] : G2.dist_code[256 + ((dist)>>7)])
|
---|
[821] | 1074 | /* Mapping from a distance to a distance code. dist is the distance - 1 and
|
---|
| 1075 | * must not have side effects. dist_code[256] and dist_code[257] are never
|
---|
| 1076 | * used.
|
---|
[1765] | 1077 | * The arguments must not have side effects.
|
---|
[821] | 1078 | */
|
---|
| 1079 |
|
---|
| 1080 |
|
---|
| 1081 | /* ===========================================================================
|
---|
| 1082 | * Initialize a new block.
|
---|
| 1083 | */
|
---|
| 1084 | static void init_block(void)
|
---|
| 1085 | {
|
---|
[1765] | 1086 | int n; /* iterates over tree elements */
|
---|
[821] | 1087 |
|
---|
| 1088 | /* Initialize the trees. */
|
---|
| 1089 | for (n = 0; n < L_CODES; n++)
|
---|
[1765] | 1090 | G2.dyn_ltree[n].Freq = 0;
|
---|
[821] | 1091 | for (n = 0; n < D_CODES; n++)
|
---|
[1765] | 1092 | G2.dyn_dtree[n].Freq = 0;
|
---|
[821] | 1093 | for (n = 0; n < BL_CODES; n++)
|
---|
[1765] | 1094 | G2.bl_tree[n].Freq = 0;
|
---|
[821] | 1095 |
|
---|
[1765] | 1096 | G2.dyn_ltree[END_BLOCK].Freq = 1;
|
---|
| 1097 | G2.opt_len = G2.static_len = 0;
|
---|
| 1098 | G2.last_lit = G2.last_dist = G2.last_flags = 0;
|
---|
| 1099 | G2.flags = 0;
|
---|
| 1100 | G2.flag_bit = 1;
|
---|
[821] | 1101 | }
|
---|
| 1102 |
|
---|
| 1103 |
|
---|
| 1104 | /* ===========================================================================
|
---|
| 1105 | * Restore the heap property by moving down the tree starting at node k,
|
---|
| 1106 | * exchanging a node with the smallest of its two sons if necessary, stopping
|
---|
| 1107 | * when the heap property is re-established (each father smaller than its
|
---|
| 1108 | * two sons).
|
---|
| 1109 | */
|
---|
[1765] | 1110 |
|
---|
| 1111 | /* Compares to subtrees, using the tree depth as tie breaker when
|
---|
| 1112 | * the subtrees have equal frequency. This minimizes the worst case length. */
|
---|
| 1113 | #define SMALLER(tree, n, m) \
|
---|
| 1114 | (tree[n].Freq < tree[m].Freq \
|
---|
| 1115 | || (tree[n].Freq == tree[m].Freq && G2.depth[n] <= G2.depth[m]))
|
---|
| 1116 |
|
---|
[821] | 1117 | static void pqdownheap(ct_data * tree, int k)
|
---|
| 1118 | {
|
---|
[1765] | 1119 | int v = G2.heap[k];
|
---|
[821] | 1120 | int j = k << 1; /* left son of k */
|
---|
| 1121 |
|
---|
[1765] | 1122 | while (j <= G2.heap_len) {
|
---|
[821] | 1123 | /* Set j to the smallest of the two sons: */
|
---|
[1765] | 1124 | if (j < G2.heap_len && SMALLER(tree, G2.heap[j + 1], G2.heap[j]))
|
---|
[821] | 1125 | j++;
|
---|
| 1126 |
|
---|
| 1127 | /* Exit if v is smaller than both sons */
|
---|
[1765] | 1128 | if (SMALLER(tree, v, G2.heap[j]))
|
---|
[821] | 1129 | break;
|
---|
| 1130 |
|
---|
| 1131 | /* Exchange v with the smallest son */
|
---|
[1765] | 1132 | G2.heap[k] = G2.heap[j];
|
---|
[821] | 1133 | k = j;
|
---|
| 1134 |
|
---|
| 1135 | /* And continue down the tree, setting j to the left son of k */
|
---|
| 1136 | j <<= 1;
|
---|
| 1137 | }
|
---|
[1765] | 1138 | G2.heap[k] = v;
|
---|
[821] | 1139 | }
|
---|
| 1140 |
|
---|
[1765] | 1141 |
|
---|
[821] | 1142 | /* ===========================================================================
|
---|
| 1143 | * Compute the optimal bit lengths for a tree and update the total bit length
|
---|
| 1144 | * for the current block.
|
---|
| 1145 | * IN assertion: the fields freq and dad are set, heap[heap_max] and
|
---|
| 1146 | * above are the tree nodes sorted by increasing frequency.
|
---|
| 1147 | * OUT assertions: the field len is set to the optimal bit length, the
|
---|
| 1148 | * array bl_count contains the frequencies for each bit length.
|
---|
| 1149 | * The length opt_len is updated; static_len is also updated if stree is
|
---|
| 1150 | * not null.
|
---|
| 1151 | */
|
---|
| 1152 | static void gen_bitlen(tree_desc * desc)
|
---|
| 1153 | {
|
---|
| 1154 | ct_data *tree = desc->dyn_tree;
|
---|
[1765] | 1155 | const uint8_t *extra = desc->extra_bits;
|
---|
[821] | 1156 | int base = desc->extra_base;
|
---|
| 1157 | int max_code = desc->max_code;
|
---|
| 1158 | int max_length = desc->max_length;
|
---|
| 1159 | ct_data *stree = desc->static_tree;
|
---|
| 1160 | int h; /* heap index */
|
---|
| 1161 | int n, m; /* iterate over the tree elements */
|
---|
| 1162 | int bits; /* bit length */
|
---|
| 1163 | int xbits; /* extra bits */
|
---|
| 1164 | ush f; /* frequency */
|
---|
| 1165 | int overflow = 0; /* number of elements with bit length too large */
|
---|
| 1166 |
|
---|
| 1167 | for (bits = 0; bits <= MAX_BITS; bits++)
|
---|
[1765] | 1168 | G2.bl_count[bits] = 0;
|
---|
[821] | 1169 |
|
---|
| 1170 | /* In a first pass, compute the optimal bit lengths (which may
|
---|
| 1171 | * overflow in the case of the bit length tree).
|
---|
| 1172 | */
|
---|
[1765] | 1173 | tree[G2.heap[G2.heap_max]].Len = 0; /* root of the heap */
|
---|
[821] | 1174 |
|
---|
[1765] | 1175 | for (h = G2.heap_max + 1; h < HEAP_SIZE; h++) {
|
---|
| 1176 | n = G2.heap[h];
|
---|
[821] | 1177 | bits = tree[tree[n].Dad].Len + 1;
|
---|
[1765] | 1178 | if (bits > max_length) {
|
---|
| 1179 | bits = max_length;
|
---|
| 1180 | overflow++;
|
---|
| 1181 | }
|
---|
[821] | 1182 | tree[n].Len = (ush) bits;
|
---|
| 1183 | /* We overwrite tree[n].Dad which is no longer needed */
|
---|
| 1184 |
|
---|
| 1185 | if (n > max_code)
|
---|
| 1186 | continue; /* not a leaf node */
|
---|
| 1187 |
|
---|
[1765] | 1188 | G2.bl_count[bits]++;
|
---|
[821] | 1189 | xbits = 0;
|
---|
| 1190 | if (n >= base)
|
---|
| 1191 | xbits = extra[n - base];
|
---|
| 1192 | f = tree[n].Freq;
|
---|
[1765] | 1193 | G2.opt_len += (ulg) f *(bits + xbits);
|
---|
[821] | 1194 |
|
---|
| 1195 | if (stree)
|
---|
[1765] | 1196 | G2.static_len += (ulg) f * (stree[n].Len + xbits);
|
---|
[821] | 1197 | }
|
---|
| 1198 | if (overflow == 0)
|
---|
| 1199 | return;
|
---|
| 1200 |
|
---|
| 1201 | Trace((stderr, "\nbit length overflow\n"));
|
---|
| 1202 | /* This happens for example on obj2 and pic of the Calgary corpus */
|
---|
| 1203 |
|
---|
| 1204 | /* Find the first bit length which could increase: */
|
---|
| 1205 | do {
|
---|
| 1206 | bits = max_length - 1;
|
---|
[1765] | 1207 | while (G2.bl_count[bits] == 0)
|
---|
[821] | 1208 | bits--;
|
---|
[1765] | 1209 | G2.bl_count[bits]--; /* move one leaf down the tree */
|
---|
| 1210 | G2.bl_count[bits + 1] += 2; /* move one overflow item as its brother */
|
---|
| 1211 | G2.bl_count[max_length]--;
|
---|
[821] | 1212 | /* The brother of the overflow item also moves one step up,
|
---|
| 1213 | * but this does not affect bl_count[max_length]
|
---|
| 1214 | */
|
---|
| 1215 | overflow -= 2;
|
---|
| 1216 | } while (overflow > 0);
|
---|
| 1217 |
|
---|
| 1218 | /* Now recompute all bit lengths, scanning in increasing frequency.
|
---|
| 1219 | * h is still equal to HEAP_SIZE. (It is simpler to reconstruct all
|
---|
| 1220 | * lengths instead of fixing only the wrong ones. This idea is taken
|
---|
| 1221 | * from 'ar' written by Haruhiko Okumura.)
|
---|
| 1222 | */
|
---|
| 1223 | for (bits = max_length; bits != 0; bits--) {
|
---|
[1765] | 1224 | n = G2.bl_count[bits];
|
---|
[821] | 1225 | while (n != 0) {
|
---|
[1765] | 1226 | m = G2.heap[--h];
|
---|
[821] | 1227 | if (m > max_code)
|
---|
| 1228 | continue;
|
---|
| 1229 | if (tree[m].Len != (unsigned) bits) {
|
---|
[1765] | 1230 | Trace((stderr, "code %d bits %d->%d\n", m, tree[m].Len, bits));
|
---|
| 1231 | G2.opt_len += ((int32_t) bits - tree[m].Len) * tree[m].Freq;
|
---|
| 1232 | tree[m].Len = bits;
|
---|
[821] | 1233 | }
|
---|
| 1234 | n--;
|
---|
| 1235 | }
|
---|
| 1236 | }
|
---|
| 1237 | }
|
---|
| 1238 |
|
---|
[1765] | 1239 |
|
---|
[821] | 1240 | /* ===========================================================================
|
---|
| 1241 | * Generate the codes for a given tree and bit counts (which need not be
|
---|
| 1242 | * optimal).
|
---|
| 1243 | * IN assertion: the array bl_count contains the bit length statistics for
|
---|
| 1244 | * the given tree and the field len is set for all tree elements.
|
---|
| 1245 | * OUT assertion: the field code is set for all tree elements of non
|
---|
| 1246 | * zero code length.
|
---|
| 1247 | */
|
---|
| 1248 | static void gen_codes(ct_data * tree, int max_code)
|
---|
| 1249 | {
|
---|
| 1250 | ush next_code[MAX_BITS + 1]; /* next code value for each bit length */
|
---|
| 1251 | ush code = 0; /* running code value */
|
---|
| 1252 | int bits; /* bit index */
|
---|
| 1253 | int n; /* code index */
|
---|
| 1254 |
|
---|
| 1255 | /* The distribution counts are first used to generate the code values
|
---|
| 1256 | * without bit reversal.
|
---|
| 1257 | */
|
---|
| 1258 | for (bits = 1; bits <= MAX_BITS; bits++) {
|
---|
[1765] | 1259 | next_code[bits] = code = (code + G2.bl_count[bits - 1]) << 1;
|
---|
[821] | 1260 | }
|
---|
| 1261 | /* Check that the bit counts in bl_count are consistent. The last code
|
---|
| 1262 | * must be all ones.
|
---|
| 1263 | */
|
---|
[1765] | 1264 | Assert(code + G2.bl_count[MAX_BITS] - 1 == (1 << MAX_BITS) - 1,
|
---|
[3232] | 1265 | "inconsistent bit counts");
|
---|
[821] | 1266 | Tracev((stderr, "\ngen_codes: max_code %d ", max_code));
|
---|
| 1267 |
|
---|
| 1268 | for (n = 0; n <= max_code; n++) {
|
---|
| 1269 | int len = tree[n].Len;
|
---|
| 1270 |
|
---|
| 1271 | if (len == 0)
|
---|
| 1272 | continue;
|
---|
| 1273 | /* Now reverse the bits */
|
---|
| 1274 | tree[n].Code = bi_reverse(next_code[len]++, len);
|
---|
| 1275 |
|
---|
[1765] | 1276 | Tracec(tree != G2.static_ltree,
|
---|
[821] | 1277 | (stderr, "\nn %3d %c l %2d c %4x (%x) ", n,
|
---|
[2725] | 1278 | (n > ' ' ? n : ' '), len, tree[n].Code,
|
---|
[821] | 1279 | next_code[len] - 1));
|
---|
| 1280 | }
|
---|
| 1281 | }
|
---|
| 1282 |
|
---|
[1765] | 1283 |
|
---|
[821] | 1284 | /* ===========================================================================
|
---|
| 1285 | * Construct one Huffman tree and assigns the code bit strings and lengths.
|
---|
| 1286 | * Update the total bit length for the current block.
|
---|
| 1287 | * IN assertion: the field freq is set for all tree elements.
|
---|
| 1288 | * OUT assertions: the fields len and code are set to the optimal bit length
|
---|
| 1289 | * and corresponding code. The length opt_len is updated; static_len is
|
---|
| 1290 | * also updated if stree is not null. The field max_code is set.
|
---|
| 1291 | */
|
---|
[1765] | 1292 |
|
---|
| 1293 | /* Remove the smallest element from the heap and recreate the heap with
|
---|
| 1294 | * one less element. Updates heap and heap_len. */
|
---|
| 1295 |
|
---|
| 1296 | #define SMALLEST 1
|
---|
| 1297 | /* Index within the heap array of least frequent node in the Huffman tree */
|
---|
| 1298 |
|
---|
| 1299 | #define PQREMOVE(tree, top) \
|
---|
| 1300 | do { \
|
---|
| 1301 | top = G2.heap[SMALLEST]; \
|
---|
| 1302 | G2.heap[SMALLEST] = G2.heap[G2.heap_len--]; \
|
---|
| 1303 | pqdownheap(tree, SMALLEST); \
|
---|
| 1304 | } while (0)
|
---|
| 1305 |
|
---|
[821] | 1306 | static void build_tree(tree_desc * desc)
|
---|
| 1307 | {
|
---|
| 1308 | ct_data *tree = desc->dyn_tree;
|
---|
| 1309 | ct_data *stree = desc->static_tree;
|
---|
| 1310 | int elems = desc->elems;
|
---|
| 1311 | int n, m; /* iterate over heap elements */
|
---|
| 1312 | int max_code = -1; /* largest code with non zero frequency */
|
---|
| 1313 | int node = elems; /* next internal node of the tree */
|
---|
| 1314 |
|
---|
| 1315 | /* Construct the initial heap, with least frequent element in
|
---|
| 1316 | * heap[SMALLEST]. The sons of heap[n] are heap[2*n] and heap[2*n+1].
|
---|
| 1317 | * heap[0] is not used.
|
---|
| 1318 | */
|
---|
[1765] | 1319 | G2.heap_len = 0;
|
---|
| 1320 | G2.heap_max = HEAP_SIZE;
|
---|
[821] | 1321 |
|
---|
| 1322 | for (n = 0; n < elems; n++) {
|
---|
| 1323 | if (tree[n].Freq != 0) {
|
---|
[1765] | 1324 | G2.heap[++G2.heap_len] = max_code = n;
|
---|
| 1325 | G2.depth[n] = 0;
|
---|
[821] | 1326 | } else {
|
---|
| 1327 | tree[n].Len = 0;
|
---|
| 1328 | }
|
---|
| 1329 | }
|
---|
| 1330 |
|
---|
| 1331 | /* The pkzip format requires that at least one distance code exists,
|
---|
| 1332 | * and that at least one bit should be sent even if there is only one
|
---|
| 1333 | * possible code. So to avoid special checks later on we force at least
|
---|
| 1334 | * two codes of non zero frequency.
|
---|
| 1335 | */
|
---|
[1765] | 1336 | while (G2.heap_len < 2) {
|
---|
| 1337 | int new = G2.heap[++G2.heap_len] = (max_code < 2 ? ++max_code : 0);
|
---|
[821] | 1338 |
|
---|
| 1339 | tree[new].Freq = 1;
|
---|
[1765] | 1340 | G2.depth[new] = 0;
|
---|
| 1341 | G2.opt_len--;
|
---|
[821] | 1342 | if (stree)
|
---|
[1765] | 1343 | G2.static_len -= stree[new].Len;
|
---|
[821] | 1344 | /* new is 0 or 1 so it does not have extra bits */
|
---|
| 1345 | }
|
---|
| 1346 | desc->max_code = max_code;
|
---|
| 1347 |
|
---|
| 1348 | /* The elements heap[heap_len/2+1 .. heap_len] are leaves of the tree,
|
---|
| 1349 | * establish sub-heaps of increasing lengths:
|
---|
| 1350 | */
|
---|
[1765] | 1351 | for (n = G2.heap_len / 2; n >= 1; n--)
|
---|
[821] | 1352 | pqdownheap(tree, n);
|
---|
| 1353 |
|
---|
| 1354 | /* Construct the Huffman tree by repeatedly combining the least two
|
---|
| 1355 | * frequent nodes.
|
---|
| 1356 | */
|
---|
| 1357 | do {
|
---|
[1765] | 1358 | PQREMOVE(tree, n); /* n = node of least frequency */
|
---|
| 1359 | m = G2.heap[SMALLEST]; /* m = node of next least frequency */
|
---|
[821] | 1360 |
|
---|
[1765] | 1361 | G2.heap[--G2.heap_max] = n; /* keep the nodes sorted by frequency */
|
---|
| 1362 | G2.heap[--G2.heap_max] = m;
|
---|
[821] | 1363 |
|
---|
| 1364 | /* Create a new node father of n and m */
|
---|
| 1365 | tree[node].Freq = tree[n].Freq + tree[m].Freq;
|
---|
[1765] | 1366 | G2.depth[node] = MAX(G2.depth[n], G2.depth[m]) + 1;
|
---|
[821] | 1367 | tree[n].Dad = tree[m].Dad = (ush) node;
|
---|
| 1368 | #ifdef DUMP_BL_TREE
|
---|
[1765] | 1369 | if (tree == G2.bl_tree) {
|
---|
[821] | 1370 | bb_error_msg("\nnode %d(%d), sons %d(%d) %d(%d)",
|
---|
| 1371 | node, tree[node].Freq, n, tree[n].Freq, m, tree[m].Freq);
|
---|
| 1372 | }
|
---|
| 1373 | #endif
|
---|
| 1374 | /* and insert the new node in the heap */
|
---|
[1765] | 1375 | G2.heap[SMALLEST] = node++;
|
---|
[821] | 1376 | pqdownheap(tree, SMALLEST);
|
---|
[1765] | 1377 | } while (G2.heap_len >= 2);
|
---|
[821] | 1378 |
|
---|
[1765] | 1379 | G2.heap[--G2.heap_max] = G2.heap[SMALLEST];
|
---|
[821] | 1380 |
|
---|
| 1381 | /* At this point, the fields freq and dad are set. We can now
|
---|
| 1382 | * generate the bit lengths.
|
---|
| 1383 | */
|
---|
| 1384 | gen_bitlen((tree_desc *) desc);
|
---|
| 1385 |
|
---|
| 1386 | /* The field len is now set, we can generate the bit codes */
|
---|
| 1387 | gen_codes((ct_data *) tree, max_code);
|
---|
| 1388 | }
|
---|
| 1389 |
|
---|
[1765] | 1390 |
|
---|
[821] | 1391 | /* ===========================================================================
|
---|
| 1392 | * Scan a literal or distance tree to determine the frequencies of the codes
|
---|
| 1393 | * in the bit length tree. Updates opt_len to take into account the repeat
|
---|
| 1394 | * counts. (The contribution of the bit length codes will be added later
|
---|
| 1395 | * during the construction of bl_tree.)
|
---|
| 1396 | */
|
---|
| 1397 | static void scan_tree(ct_data * tree, int max_code)
|
---|
| 1398 | {
|
---|
| 1399 | int n; /* iterates over all tree elements */
|
---|
| 1400 | int prevlen = -1; /* last emitted length */
|
---|
| 1401 | int curlen; /* length of current code */
|
---|
| 1402 | int nextlen = tree[0].Len; /* length of next code */
|
---|
| 1403 | int count = 0; /* repeat count of the current code */
|
---|
| 1404 | int max_count = 7; /* max repeat count */
|
---|
| 1405 | int min_count = 4; /* min repeat count */
|
---|
| 1406 |
|
---|
[1765] | 1407 | if (nextlen == 0) {
|
---|
| 1408 | max_count = 138;
|
---|
| 1409 | min_count = 3;
|
---|
| 1410 | }
|
---|
| 1411 | tree[max_code + 1].Len = 0xffff; /* guard */
|
---|
[821] | 1412 |
|
---|
| 1413 | for (n = 0; n <= max_code; n++) {
|
---|
| 1414 | curlen = nextlen;
|
---|
| 1415 | nextlen = tree[n + 1].Len;
|
---|
[1765] | 1416 | if (++count < max_count && curlen == nextlen)
|
---|
[821] | 1417 | continue;
|
---|
[1765] | 1418 |
|
---|
| 1419 | if (count < min_count) {
|
---|
| 1420 | G2.bl_tree[curlen].Freq += count;
|
---|
[821] | 1421 | } else if (curlen != 0) {
|
---|
| 1422 | if (curlen != prevlen)
|
---|
[1765] | 1423 | G2.bl_tree[curlen].Freq++;
|
---|
| 1424 | G2.bl_tree[REP_3_6].Freq++;
|
---|
[821] | 1425 | } else if (count <= 10) {
|
---|
[1765] | 1426 | G2.bl_tree[REPZ_3_10].Freq++;
|
---|
[821] | 1427 | } else {
|
---|
[1765] | 1428 | G2.bl_tree[REPZ_11_138].Freq++;
|
---|
[821] | 1429 | }
|
---|
| 1430 | count = 0;
|
---|
| 1431 | prevlen = curlen;
|
---|
[1765] | 1432 |
|
---|
| 1433 | max_count = 7;
|
---|
| 1434 | min_count = 4;
|
---|
[821] | 1435 | if (nextlen == 0) {
|
---|
[1765] | 1436 | max_count = 138;
|
---|
| 1437 | min_count = 3;
|
---|
[821] | 1438 | } else if (curlen == nextlen) {
|
---|
[1765] | 1439 | max_count = 6;
|
---|
| 1440 | min_count = 3;
|
---|
[821] | 1441 | }
|
---|
| 1442 | }
|
---|
| 1443 | }
|
---|
| 1444 |
|
---|
[1765] | 1445 |
|
---|
[821] | 1446 | /* ===========================================================================
|
---|
| 1447 | * Send a literal or distance tree in compressed form, using the codes in
|
---|
| 1448 | * bl_tree.
|
---|
| 1449 | */
|
---|
| 1450 | static void send_tree(ct_data * tree, int max_code)
|
---|
| 1451 | {
|
---|
| 1452 | int n; /* iterates over all tree elements */
|
---|
| 1453 | int prevlen = -1; /* last emitted length */
|
---|
| 1454 | int curlen; /* length of current code */
|
---|
| 1455 | int nextlen = tree[0].Len; /* length of next code */
|
---|
| 1456 | int count = 0; /* repeat count of the current code */
|
---|
| 1457 | int max_count = 7; /* max repeat count */
|
---|
| 1458 | int min_count = 4; /* min repeat count */
|
---|
| 1459 |
|
---|
| 1460 | /* tree[max_code+1].Len = -1; *//* guard already set */
|
---|
| 1461 | if (nextlen == 0)
|
---|
| 1462 | max_count = 138, min_count = 3;
|
---|
| 1463 |
|
---|
| 1464 | for (n = 0; n <= max_code; n++) {
|
---|
| 1465 | curlen = nextlen;
|
---|
| 1466 | nextlen = tree[n + 1].Len;
|
---|
| 1467 | if (++count < max_count && curlen == nextlen) {
|
---|
| 1468 | continue;
|
---|
| 1469 | } else if (count < min_count) {
|
---|
| 1470 | do {
|
---|
[1765] | 1471 | SEND_CODE(curlen, G2.bl_tree);
|
---|
| 1472 | } while (--count);
|
---|
[821] | 1473 | } else if (curlen != 0) {
|
---|
| 1474 | if (curlen != prevlen) {
|
---|
[1765] | 1475 | SEND_CODE(curlen, G2.bl_tree);
|
---|
[821] | 1476 | count--;
|
---|
| 1477 | }
|
---|
| 1478 | Assert(count >= 3 && count <= 6, " 3_6?");
|
---|
[1765] | 1479 | SEND_CODE(REP_3_6, G2.bl_tree);
|
---|
[821] | 1480 | send_bits(count - 3, 2);
|
---|
| 1481 | } else if (count <= 10) {
|
---|
[1765] | 1482 | SEND_CODE(REPZ_3_10, G2.bl_tree);
|
---|
[821] | 1483 | send_bits(count - 3, 3);
|
---|
| 1484 | } else {
|
---|
[1765] | 1485 | SEND_CODE(REPZ_11_138, G2.bl_tree);
|
---|
[821] | 1486 | send_bits(count - 11, 7);
|
---|
| 1487 | }
|
---|
| 1488 | count = 0;
|
---|
| 1489 | prevlen = curlen;
|
---|
| 1490 | if (nextlen == 0) {
|
---|
[1765] | 1491 | max_count = 138;
|
---|
| 1492 | min_count = 3;
|
---|
[821] | 1493 | } else if (curlen == nextlen) {
|
---|
[1765] | 1494 | max_count = 6;
|
---|
| 1495 | min_count = 3;
|
---|
[821] | 1496 | } else {
|
---|
[1765] | 1497 | max_count = 7;
|
---|
| 1498 | min_count = 4;
|
---|
[821] | 1499 | }
|
---|
| 1500 | }
|
---|
| 1501 | }
|
---|
| 1502 |
|
---|
[1765] | 1503 |
|
---|
[821] | 1504 | /* ===========================================================================
|
---|
| 1505 | * Construct the Huffman tree for the bit lengths and return the index in
|
---|
| 1506 | * bl_order of the last bit length code to send.
|
---|
| 1507 | */
|
---|
| 1508 | static int build_bl_tree(void)
|
---|
| 1509 | {
|
---|
| 1510 | int max_blindex; /* index of last bit length code of non zero freq */
|
---|
| 1511 |
|
---|
| 1512 | /* Determine the bit length frequencies for literal and distance trees */
|
---|
[1765] | 1513 | scan_tree(G2.dyn_ltree, G2.l_desc.max_code);
|
---|
| 1514 | scan_tree(G2.dyn_dtree, G2.d_desc.max_code);
|
---|
[821] | 1515 |
|
---|
| 1516 | /* Build the bit length tree: */
|
---|
[1765] | 1517 | build_tree(&G2.bl_desc);
|
---|
[821] | 1518 | /* opt_len now includes the length of the tree representations, except
|
---|
| 1519 | * the lengths of the bit lengths codes and the 5+5+4 bits for the counts.
|
---|
| 1520 | */
|
---|
| 1521 |
|
---|
| 1522 | /* Determine the number of bit length codes to send. The pkzip format
|
---|
| 1523 | * requires that at least 4 bit length codes be sent. (appnote.txt says
|
---|
| 1524 | * 3 but the actual value used is 4.)
|
---|
| 1525 | */
|
---|
| 1526 | for (max_blindex = BL_CODES - 1; max_blindex >= 3; max_blindex--) {
|
---|
[1765] | 1527 | if (G2.bl_tree[bl_order[max_blindex]].Len != 0)
|
---|
[821] | 1528 | break;
|
---|
| 1529 | }
|
---|
| 1530 | /* Update opt_len to include the bit length tree and counts */
|
---|
[1765] | 1531 | G2.opt_len += 3 * (max_blindex + 1) + 5 + 5 + 4;
|
---|
| 1532 | Tracev((stderr, "\ndyn trees: dyn %ld, stat %ld", G2.opt_len, G2.static_len));
|
---|
[821] | 1533 |
|
---|
| 1534 | return max_blindex;
|
---|
| 1535 | }
|
---|
| 1536 |
|
---|
[1765] | 1537 |
|
---|
[821] | 1538 | /* ===========================================================================
|
---|
| 1539 | * Send the header for a block using dynamic Huffman trees: the counts, the
|
---|
| 1540 | * lengths of the bit length codes, the literal tree and the distance tree.
|
---|
| 1541 | * IN assertion: lcodes >= 257, dcodes >= 1, blcodes >= 4.
|
---|
| 1542 | */
|
---|
| 1543 | static void send_all_trees(int lcodes, int dcodes, int blcodes)
|
---|
| 1544 | {
|
---|
| 1545 | int rank; /* index in bl_order */
|
---|
| 1546 |
|
---|
| 1547 | Assert(lcodes >= 257 && dcodes >= 1 && blcodes >= 4, "not enough codes");
|
---|
| 1548 | Assert(lcodes <= L_CODES && dcodes <= D_CODES
|
---|
| 1549 | && blcodes <= BL_CODES, "too many codes");
|
---|
| 1550 | Tracev((stderr, "\nbl counts: "));
|
---|
| 1551 | send_bits(lcodes - 257, 5); /* not +255 as stated in appnote.txt */
|
---|
| 1552 | send_bits(dcodes - 1, 5);
|
---|
| 1553 | send_bits(blcodes - 4, 4); /* not -3 as stated in appnote.txt */
|
---|
| 1554 | for (rank = 0; rank < blcodes; rank++) {
|
---|
| 1555 | Tracev((stderr, "\nbl code %2d ", bl_order[rank]));
|
---|
[1765] | 1556 | send_bits(G2.bl_tree[bl_order[rank]].Len, 3);
|
---|
[821] | 1557 | }
|
---|
[1765] | 1558 | Tracev((stderr, "\nbl tree: sent %ld", G1.bits_sent));
|
---|
[821] | 1559 |
|
---|
[1765] | 1560 | send_tree((ct_data *) G2.dyn_ltree, lcodes - 1); /* send the literal tree */
|
---|
| 1561 | Tracev((stderr, "\nlit tree: sent %ld", G1.bits_sent));
|
---|
[821] | 1562 |
|
---|
[1765] | 1563 | send_tree((ct_data *) G2.dyn_dtree, dcodes - 1); /* send the distance tree */
|
---|
| 1564 | Tracev((stderr, "\ndist tree: sent %ld", G1.bits_sent));
|
---|
[821] | 1565 | }
|
---|
| 1566 |
|
---|
[1765] | 1567 |
|
---|
[821] | 1568 | /* ===========================================================================
|
---|
[1765] | 1569 | * Save the match info and tally the frequency counts. Return true if
|
---|
| 1570 | * the current block must be flushed.
|
---|
| 1571 | */
|
---|
| 1572 | static int ct_tally(int dist, int lc)
|
---|
| 1573 | {
|
---|
| 1574 | G1.l_buf[G2.last_lit++] = lc;
|
---|
| 1575 | if (dist == 0) {
|
---|
| 1576 | /* lc is the unmatched char */
|
---|
| 1577 | G2.dyn_ltree[lc].Freq++;
|
---|
| 1578 | } else {
|
---|
| 1579 | /* Here, lc is the match length - MIN_MATCH */
|
---|
| 1580 | dist--; /* dist = match distance - 1 */
|
---|
| 1581 | Assert((ush) dist < (ush) MAX_DIST
|
---|
| 1582 | && (ush) lc <= (ush) (MAX_MATCH - MIN_MATCH)
|
---|
| 1583 | && (ush) D_CODE(dist) < (ush) D_CODES, "ct_tally: bad match"
|
---|
| 1584 | );
|
---|
| 1585 |
|
---|
| 1586 | G2.dyn_ltree[G2.length_code[lc] + LITERALS + 1].Freq++;
|
---|
| 1587 | G2.dyn_dtree[D_CODE(dist)].Freq++;
|
---|
| 1588 |
|
---|
| 1589 | G1.d_buf[G2.last_dist++] = dist;
|
---|
| 1590 | G2.flags |= G2.flag_bit;
|
---|
| 1591 | }
|
---|
| 1592 | G2.flag_bit <<= 1;
|
---|
| 1593 |
|
---|
| 1594 | /* Output the flags if they fill a byte: */
|
---|
| 1595 | if ((G2.last_lit & 7) == 0) {
|
---|
| 1596 | G2.flag_buf[G2.last_flags++] = G2.flags;
|
---|
| 1597 | G2.flags = 0;
|
---|
| 1598 | G2.flag_bit = 1;
|
---|
| 1599 | }
|
---|
| 1600 | /* Try to guess if it is profitable to stop the current block here */
|
---|
| 1601 | if ((G2.last_lit & 0xfff) == 0) {
|
---|
| 1602 | /* Compute an upper bound for the compressed length */
|
---|
| 1603 | ulg out_length = G2.last_lit * 8L;
|
---|
| 1604 | ulg in_length = (ulg) G1.strstart - G1.block_start;
|
---|
| 1605 | int dcode;
|
---|
| 1606 |
|
---|
| 1607 | for (dcode = 0; dcode < D_CODES; dcode++) {
|
---|
| 1608 | out_length += G2.dyn_dtree[dcode].Freq * (5L + extra_dbits[dcode]);
|
---|
| 1609 | }
|
---|
| 1610 | out_length >>= 3;
|
---|
| 1611 | Trace((stderr,
|
---|
[3232] | 1612 | "\nlast_lit %u, last_dist %u, in %ld, out ~%ld(%ld%%) ",
|
---|
| 1613 | G2.last_lit, G2.last_dist, in_length, out_length,
|
---|
| 1614 | 100L - out_length * 100L / in_length));
|
---|
[1765] | 1615 | if (G2.last_dist < G2.last_lit / 2 && out_length < in_length / 2)
|
---|
| 1616 | return 1;
|
---|
| 1617 | }
|
---|
| 1618 | return (G2.last_lit == LIT_BUFSIZE - 1 || G2.last_dist == DIST_BUFSIZE);
|
---|
| 1619 | /* We avoid equality with LIT_BUFSIZE because of wraparound at 64K
|
---|
| 1620 | * on 16 bit machines and because stored blocks are restricted to
|
---|
| 1621 | * 64K-1 bytes.
|
---|
| 1622 | */
|
---|
| 1623 | }
|
---|
| 1624 |
|
---|
| 1625 | /* ===========================================================================
|
---|
| 1626 | * Send the block data compressed using the given Huffman trees
|
---|
| 1627 | */
|
---|
| 1628 | static void compress_block(ct_data * ltree, ct_data * dtree)
|
---|
| 1629 | {
|
---|
| 1630 | unsigned dist; /* distance of matched string */
|
---|
| 1631 | int lc; /* match length or unmatched char (if dist == 0) */
|
---|
| 1632 | unsigned lx = 0; /* running index in l_buf */
|
---|
| 1633 | unsigned dx = 0; /* running index in d_buf */
|
---|
| 1634 | unsigned fx = 0; /* running index in flag_buf */
|
---|
| 1635 | uch flag = 0; /* current flags */
|
---|
| 1636 | unsigned code; /* the code to send */
|
---|
| 1637 | int extra; /* number of extra bits to send */
|
---|
| 1638 |
|
---|
| 1639 | if (G2.last_lit != 0) do {
|
---|
| 1640 | if ((lx & 7) == 0)
|
---|
| 1641 | flag = G2.flag_buf[fx++];
|
---|
| 1642 | lc = G1.l_buf[lx++];
|
---|
| 1643 | if ((flag & 1) == 0) {
|
---|
| 1644 | SEND_CODE(lc, ltree); /* send a literal byte */
|
---|
[2725] | 1645 | Tracecv(lc > ' ', (stderr, " '%c' ", lc));
|
---|
[1765] | 1646 | } else {
|
---|
| 1647 | /* Here, lc is the match length - MIN_MATCH */
|
---|
| 1648 | code = G2.length_code[lc];
|
---|
| 1649 | SEND_CODE(code + LITERALS + 1, ltree); /* send the length code */
|
---|
| 1650 | extra = extra_lbits[code];
|
---|
| 1651 | if (extra != 0) {
|
---|
| 1652 | lc -= G2.base_length[code];
|
---|
| 1653 | send_bits(lc, extra); /* send the extra length bits */
|
---|
| 1654 | }
|
---|
| 1655 | dist = G1.d_buf[dx++];
|
---|
| 1656 | /* Here, dist is the match distance - 1 */
|
---|
| 1657 | code = D_CODE(dist);
|
---|
| 1658 | Assert(code < D_CODES, "bad d_code");
|
---|
| 1659 |
|
---|
| 1660 | SEND_CODE(code, dtree); /* send the distance code */
|
---|
| 1661 | extra = extra_dbits[code];
|
---|
| 1662 | if (extra != 0) {
|
---|
| 1663 | dist -= G2.base_dist[code];
|
---|
| 1664 | send_bits(dist, extra); /* send the extra distance bits */
|
---|
| 1665 | }
|
---|
| 1666 | } /* literal or match pair ? */
|
---|
| 1667 | flag >>= 1;
|
---|
| 1668 | } while (lx < G2.last_lit);
|
---|
| 1669 |
|
---|
| 1670 | SEND_CODE(END_BLOCK, ltree);
|
---|
| 1671 | }
|
---|
| 1672 |
|
---|
| 1673 |
|
---|
| 1674 | /* ===========================================================================
|
---|
[821] | 1675 | * Determine the best encoding for the current block: dynamic trees, static
|
---|
| 1676 | * trees or store, and output the encoded block to the zip file. This function
|
---|
| 1677 | * returns the total compressed length for the file so far.
|
---|
| 1678 | */
|
---|
| 1679 | static ulg flush_block(char *buf, ulg stored_len, int eof)
|
---|
| 1680 | {
|
---|
[1765] | 1681 | ulg opt_lenb, static_lenb; /* opt_len and static_len in bytes */
|
---|
| 1682 | int max_blindex; /* index of last bit length code of non zero freq */
|
---|
[821] | 1683 |
|
---|
[1765] | 1684 | G2.flag_buf[G2.last_flags] = G2.flags; /* Save the flags for the last 8 items */
|
---|
[821] | 1685 |
|
---|
| 1686 | /* Construct the literal and distance trees */
|
---|
[1765] | 1687 | build_tree(&G2.l_desc);
|
---|
| 1688 | Tracev((stderr, "\nlit data: dyn %ld, stat %ld", G2.opt_len, G2.static_len));
|
---|
[821] | 1689 |
|
---|
[1765] | 1690 | build_tree(&G2.d_desc);
|
---|
| 1691 | Tracev((stderr, "\ndist data: dyn %ld, stat %ld", G2.opt_len, G2.static_len));
|
---|
[821] | 1692 | /* At this point, opt_len and static_len are the total bit lengths of
|
---|
| 1693 | * the compressed block data, excluding the tree representations.
|
---|
| 1694 | */
|
---|
| 1695 |
|
---|
| 1696 | /* Build the bit length tree for the above two trees, and get the index
|
---|
| 1697 | * in bl_order of the last bit length code to send.
|
---|
| 1698 | */
|
---|
| 1699 | max_blindex = build_bl_tree();
|
---|
| 1700 |
|
---|
| 1701 | /* Determine the best encoding. Compute first the block length in bytes */
|
---|
[1765] | 1702 | opt_lenb = (G2.opt_len + 3 + 7) >> 3;
|
---|
| 1703 | static_lenb = (G2.static_len + 3 + 7) >> 3;
|
---|
[821] | 1704 |
|
---|
| 1705 | Trace((stderr,
|
---|
[3232] | 1706 | "\nopt %lu(%lu) stat %lu(%lu) stored %lu lit %u dist %u ",
|
---|
| 1707 | opt_lenb, G2.opt_len, static_lenb, G2.static_len, stored_len,
|
---|
| 1708 | G2.last_lit, G2.last_dist));
|
---|
[821] | 1709 |
|
---|
| 1710 | if (static_lenb <= opt_lenb)
|
---|
| 1711 | opt_lenb = static_lenb;
|
---|
| 1712 |
|
---|
| 1713 | /* If compression failed and this is the first and last block,
|
---|
| 1714 | * and if the zip file can be seeked (to rewrite the local header),
|
---|
| 1715 | * the whole file is transformed into a stored file:
|
---|
| 1716 | */
|
---|
[1765] | 1717 | if (stored_len <= opt_lenb && eof && G2.compressed_len == 0L && seekable()) {
|
---|
[821] | 1718 | /* Since LIT_BUFSIZE <= 2*WSIZE, the input data must be there: */
|
---|
[1765] | 1719 | if (buf == NULL)
|
---|
[821] | 1720 | bb_error_msg("block vanished");
|
---|
| 1721 |
|
---|
| 1722 | copy_block(buf, (unsigned) stored_len, 0); /* without header */
|
---|
[1765] | 1723 | G2.compressed_len = stored_len << 3;
|
---|
| 1724 | } else if (stored_len + 4 <= opt_lenb && buf != NULL) {
|
---|
[821] | 1725 | /* 4: two words for the lengths */
|
---|
| 1726 | /* The test buf != NULL is only necessary if LIT_BUFSIZE > WSIZE.
|
---|
| 1727 | * Otherwise we can't have processed more than WSIZE input bytes since
|
---|
| 1728 | * the last block flush, because compression would have been
|
---|
| 1729 | * successful. If LIT_BUFSIZE <= WSIZE, it is never too late to
|
---|
| 1730 | * transform a block into a stored block.
|
---|
| 1731 | */
|
---|
| 1732 | send_bits((STORED_BLOCK << 1) + eof, 3); /* send block type */
|
---|
[1765] | 1733 | G2.compressed_len = (G2.compressed_len + 3 + 7) & ~7L;
|
---|
| 1734 | G2.compressed_len += (stored_len + 4) << 3;
|
---|
[821] | 1735 |
|
---|
| 1736 | copy_block(buf, (unsigned) stored_len, 1); /* with header */
|
---|
| 1737 | } else if (static_lenb == opt_lenb) {
|
---|
| 1738 | send_bits((STATIC_TREES << 1) + eof, 3);
|
---|
[1765] | 1739 | compress_block((ct_data *) G2.static_ltree, (ct_data *) G2.static_dtree);
|
---|
| 1740 | G2.compressed_len += 3 + G2.static_len;
|
---|
[821] | 1741 | } else {
|
---|
| 1742 | send_bits((DYN_TREES << 1) + eof, 3);
|
---|
[1765] | 1743 | send_all_trees(G2.l_desc.max_code + 1, G2.d_desc.max_code + 1,
|
---|
[3232] | 1744 | max_blindex + 1);
|
---|
[1765] | 1745 | compress_block((ct_data *) G2.dyn_ltree, (ct_data *) G2.dyn_dtree);
|
---|
| 1746 | G2.compressed_len += 3 + G2.opt_len;
|
---|
[821] | 1747 | }
|
---|
[1765] | 1748 | Assert(G2.compressed_len == G1.bits_sent, "bad compressed size");
|
---|
[821] | 1749 | init_block();
|
---|
| 1750 |
|
---|
| 1751 | if (eof) {
|
---|
| 1752 | bi_windup();
|
---|
[1765] | 1753 | G2.compressed_len += 7; /* align on byte boundary */
|
---|
[821] | 1754 | }
|
---|
[1765] | 1755 | Tracev((stderr, "\ncomprlen %lu(%lu) ", G2.compressed_len >> 3,
|
---|
| 1756 | G2.compressed_len - 7 * eof));
|
---|
[821] | 1757 |
|
---|
[1765] | 1758 | return G2.compressed_len >> 3;
|
---|
[821] | 1759 | }
|
---|
| 1760 |
|
---|
[1765] | 1761 |
|
---|
[821] | 1762 | /* ===========================================================================
|
---|
[1765] | 1763 | * Update a hash value with the given input byte
|
---|
[3232] | 1764 | * IN assertion: all calls to UPDATE_HASH are made with consecutive
|
---|
[1765] | 1765 | * input characters, so that a running hash key can be computed from the
|
---|
| 1766 | * previous key instead of complete recalculation each time.
|
---|
[821] | 1767 | */
|
---|
[1765] | 1768 | #define UPDATE_HASH(h, c) (h = (((h)<<H_SHIFT) ^ (c)) & HASH_MASK)
|
---|
| 1769 |
|
---|
| 1770 |
|
---|
| 1771 | /* ===========================================================================
|
---|
| 1772 | * Same as above, but achieves better compression. We use a lazy
|
---|
| 1773 | * evaluation for matches: a match is finally adopted only if there is
|
---|
| 1774 | * no better match at the next window position.
|
---|
| 1775 | *
|
---|
| 1776 | * Processes a new input file and return its compressed length. Sets
|
---|
| 1777 | * the compressed length, crc, deflate flags and internal file
|
---|
| 1778 | * attributes.
|
---|
| 1779 | */
|
---|
| 1780 |
|
---|
| 1781 | /* Flush the current block, with given end-of-file flag.
|
---|
| 1782 | * IN assertion: strstart is set to the end of the current match. */
|
---|
| 1783 | #define FLUSH_BLOCK(eof) \
|
---|
| 1784 | flush_block( \
|
---|
| 1785 | G1.block_start >= 0L \
|
---|
| 1786 | ? (char*)&G1.window[(unsigned)G1.block_start] \
|
---|
| 1787 | : (char*)NULL, \
|
---|
| 1788 | (ulg)G1.strstart - G1.block_start, \
|
---|
| 1789 | (eof) \
|
---|
| 1790 | )
|
---|
| 1791 |
|
---|
| 1792 | /* Insert string s in the dictionary and set match_head to the previous head
|
---|
| 1793 | * of the hash chain (the most recent string with same hash key). Return
|
---|
| 1794 | * the previous length of the hash chain.
|
---|
[3232] | 1795 | * IN assertion: all calls to INSERT_STRING are made with consecutive
|
---|
[1765] | 1796 | * input characters and the first MIN_MATCH bytes of s are valid
|
---|
| 1797 | * (except for the last MIN_MATCH-1 bytes of the input file). */
|
---|
| 1798 | #define INSERT_STRING(s, match_head) \
|
---|
| 1799 | do { \
|
---|
| 1800 | UPDATE_HASH(G1.ins_h, G1.window[(s) + MIN_MATCH-1]); \
|
---|
| 1801 | G1.prev[(s) & WMASK] = match_head = head[G1.ins_h]; \
|
---|
| 1802 | head[G1.ins_h] = (s); \
|
---|
| 1803 | } while (0)
|
---|
| 1804 |
|
---|
| 1805 | static ulg deflate(void)
|
---|
[821] | 1806 | {
|
---|
[1765] | 1807 | IPos hash_head; /* head of hash chain */
|
---|
| 1808 | IPos prev_match; /* previous match */
|
---|
| 1809 | int flush; /* set if current block must be flushed */
|
---|
| 1810 | int match_available = 0; /* set if previous match exists */
|
---|
| 1811 | unsigned match_length = MIN_MATCH - 1; /* length of best match */
|
---|
[821] | 1812 |
|
---|
[1765] | 1813 | /* Process the input block. */
|
---|
| 1814 | while (G1.lookahead != 0) {
|
---|
| 1815 | /* Insert the string window[strstart .. strstart+2] in the
|
---|
| 1816 | * dictionary, and set hash_head to the head of the hash chain:
|
---|
| 1817 | */
|
---|
| 1818 | INSERT_STRING(G1.strstart, hash_head);
|
---|
[821] | 1819 |
|
---|
[1765] | 1820 | /* Find the longest match, discarding those <= prev_length.
|
---|
| 1821 | */
|
---|
| 1822 | G1.prev_length = match_length;
|
---|
| 1823 | prev_match = G1.match_start;
|
---|
| 1824 | match_length = MIN_MATCH - 1;
|
---|
[821] | 1825 |
|
---|
[1765] | 1826 | if (hash_head != 0 && G1.prev_length < max_lazy_match
|
---|
| 1827 | && G1.strstart - hash_head <= MAX_DIST
|
---|
| 1828 | ) {
|
---|
| 1829 | /* To simplify the code, we prevent matches with the string
|
---|
| 1830 | * of window index 0 (in particular we have to avoid a match
|
---|
| 1831 | * of the string with itself at the start of the input file).
|
---|
| 1832 | */
|
---|
| 1833 | match_length = longest_match(hash_head);
|
---|
| 1834 | /* longest_match() sets match_start */
|
---|
| 1835 | if (match_length > G1.lookahead)
|
---|
| 1836 | match_length = G1.lookahead;
|
---|
[821] | 1837 |
|
---|
[1765] | 1838 | /* Ignore a length 3 match if it is too distant: */
|
---|
| 1839 | if (match_length == MIN_MATCH && G1.strstart - G1.match_start > TOO_FAR) {
|
---|
| 1840 | /* If prev_match is also MIN_MATCH, G1.match_start is garbage
|
---|
| 1841 | * but we will ignore the current match anyway.
|
---|
| 1842 | */
|
---|
| 1843 | match_length--;
|
---|
| 1844 | }
|
---|
[821] | 1845 | }
|
---|
[1765] | 1846 | /* If there was a match at the previous step and the current
|
---|
| 1847 | * match is not better, output the previous match:
|
---|
| 1848 | */
|
---|
| 1849 | if (G1.prev_length >= MIN_MATCH && match_length <= G1.prev_length) {
|
---|
| 1850 | check_match(G1.strstart - 1, prev_match, G1.prev_length);
|
---|
| 1851 | flush = ct_tally(G1.strstart - 1 - prev_match, G1.prev_length - MIN_MATCH);
|
---|
| 1852 |
|
---|
| 1853 | /* Insert in hash table all strings up to the end of the match.
|
---|
| 1854 | * strstart-1 and strstart are already inserted.
|
---|
| 1855 | */
|
---|
| 1856 | G1.lookahead -= G1.prev_length - 1;
|
---|
| 1857 | G1.prev_length -= 2;
|
---|
| 1858 | do {
|
---|
| 1859 | G1.strstart++;
|
---|
| 1860 | INSERT_STRING(G1.strstart, hash_head);
|
---|
| 1861 | /* strstart never exceeds WSIZE-MAX_MATCH, so there are
|
---|
| 1862 | * always MIN_MATCH bytes ahead. If lookahead < MIN_MATCH
|
---|
| 1863 | * these bytes are garbage, but it does not matter since the
|
---|
| 1864 | * next lookahead bytes will always be emitted as literals.
|
---|
| 1865 | */
|
---|
| 1866 | } while (--G1.prev_length != 0);
|
---|
| 1867 | match_available = 0;
|
---|
| 1868 | match_length = MIN_MATCH - 1;
|
---|
| 1869 | G1.strstart++;
|
---|
| 1870 | if (flush) {
|
---|
| 1871 | FLUSH_BLOCK(0);
|
---|
| 1872 | G1.block_start = G1.strstart;
|
---|
| 1873 | }
|
---|
| 1874 | } else if (match_available) {
|
---|
| 1875 | /* If there was no match at the previous position, output a
|
---|
| 1876 | * single literal. If there was a match but the current match
|
---|
| 1877 | * is longer, truncate the previous match to a single literal.
|
---|
| 1878 | */
|
---|
| 1879 | Tracevv((stderr, "%c", G1.window[G1.strstart - 1]));
|
---|
| 1880 | if (ct_tally(0, G1.window[G1.strstart - 1])) {
|
---|
| 1881 | FLUSH_BLOCK(0);
|
---|
| 1882 | G1.block_start = G1.strstart;
|
---|
| 1883 | }
|
---|
| 1884 | G1.strstart++;
|
---|
| 1885 | G1.lookahead--;
|
---|
| 1886 | } else {
|
---|
| 1887 | /* There is no previous match to compare with, wait for
|
---|
| 1888 | * the next step to decide.
|
---|
| 1889 | */
|
---|
| 1890 | match_available = 1;
|
---|
| 1891 | G1.strstart++;
|
---|
| 1892 | G1.lookahead--;
|
---|
| 1893 | }
|
---|
| 1894 | Assert(G1.strstart <= G1.isize && lookahead <= G1.isize, "a bit too far");
|
---|
| 1895 |
|
---|
| 1896 | /* Make sure that we always have enough lookahead, except
|
---|
| 1897 | * at the end of the input file. We need MAX_MATCH bytes
|
---|
| 1898 | * for the next match, plus MIN_MATCH bytes to insert the
|
---|
| 1899 | * string following the next match.
|
---|
| 1900 | */
|
---|
| 1901 | while (G1.lookahead < MIN_LOOKAHEAD && !G1.eofile)
|
---|
| 1902 | fill_window();
|
---|
[821] | 1903 | }
|
---|
[1765] | 1904 | if (match_available)
|
---|
| 1905 | ct_tally(0, G1.window[G1.strstart - 1]);
|
---|
| 1906 |
|
---|
| 1907 | return FLUSH_BLOCK(1); /* eof */
|
---|
[821] | 1908 | }
|
---|
| 1909 |
|
---|
[1765] | 1910 |
|
---|
[821] | 1911 | /* ===========================================================================
|
---|
[1765] | 1912 | * Initialize the bit string routines.
|
---|
[821] | 1913 | */
|
---|
[1765] | 1914 | static void bi_init(void)
|
---|
[821] | 1915 | {
|
---|
[1765] | 1916 | G1.bi_buf = 0;
|
---|
| 1917 | G1.bi_valid = 0;
|
---|
| 1918 | #ifdef DEBUG
|
---|
| 1919 | G1.bits_sent = 0L;
|
---|
| 1920 | #endif
|
---|
| 1921 | }
|
---|
[821] | 1922 |
|
---|
| 1923 |
|
---|
[1765] | 1924 | /* ===========================================================================
|
---|
| 1925 | * Initialize the "longest match" routines for a new file
|
---|
| 1926 | */
|
---|
| 1927 | static void lm_init(ush * flagsp)
|
---|
| 1928 | {
|
---|
| 1929 | unsigned j;
|
---|
[821] | 1930 |
|
---|
[1765] | 1931 | /* Initialize the hash table. */
|
---|
| 1932 | memset(head, 0, HASH_SIZE * sizeof(*head));
|
---|
| 1933 | /* prev will be initialized on the fly */
|
---|
| 1934 |
|
---|
| 1935 | /* speed options for the general purpose bit flag */
|
---|
| 1936 | *flagsp |= 2; /* FAST 4, SLOW 2 */
|
---|
| 1937 | /* ??? reduce max_chain_length for binary files */
|
---|
| 1938 |
|
---|
| 1939 | G1.strstart = 0;
|
---|
| 1940 | G1.block_start = 0L;
|
---|
| 1941 |
|
---|
| 1942 | G1.lookahead = file_read(G1.window,
|
---|
| 1943 | sizeof(int) <= 2 ? (unsigned) WSIZE : 2 * WSIZE);
|
---|
| 1944 |
|
---|
| 1945 | if (G1.lookahead == 0 || G1.lookahead == (unsigned) -1) {
|
---|
| 1946 | G1.eofile = 1;
|
---|
| 1947 | G1.lookahead = 0;
|
---|
| 1948 | return;
|
---|
| 1949 | }
|
---|
| 1950 | G1.eofile = 0;
|
---|
| 1951 | /* Make sure that we always have enough lookahead. This is important
|
---|
| 1952 | * if input comes from a device such as a tty.
|
---|
| 1953 | */
|
---|
| 1954 | while (G1.lookahead < MIN_LOOKAHEAD && !G1.eofile)
|
---|
| 1955 | fill_window();
|
---|
| 1956 |
|
---|
| 1957 | G1.ins_h = 0;
|
---|
| 1958 | for (j = 0; j < MIN_MATCH - 1; j++)
|
---|
| 1959 | UPDATE_HASH(G1.ins_h, G1.window[j]);
|
---|
| 1960 | /* If lookahead < MIN_MATCH, ins_h is garbage, but this is
|
---|
| 1961 | * not important since only literal bytes will be emitted.
|
---|
| 1962 | */
|
---|
[821] | 1963 | }
|
---|
| 1964 |
|
---|
[1765] | 1965 |
|
---|
[821] | 1966 | /* ===========================================================================
|
---|
[1765] | 1967 | * Allocate the match buffer, initialize the various tables and save the
|
---|
| 1968 | * location of the internal file attribute (ascii/binary) and method
|
---|
| 1969 | * (DEFLATE/STORE).
|
---|
| 1970 | * One callsite in zip()
|
---|
[821] | 1971 | */
|
---|
[1765] | 1972 | static void ct_init(void)
|
---|
[821] | 1973 | {
|
---|
[1765] | 1974 | int n; /* iterates over tree elements */
|
---|
| 1975 | int length; /* length value */
|
---|
| 1976 | int code; /* code value */
|
---|
| 1977 | int dist; /* distance index */
|
---|
[821] | 1978 |
|
---|
[1765] | 1979 | G2.compressed_len = 0L;
|
---|
| 1980 |
|
---|
| 1981 | #ifdef NOT_NEEDED
|
---|
| 1982 | if (G2.static_dtree[0].Len != 0)
|
---|
| 1983 | return; /* ct_init already called */
|
---|
| 1984 | #endif
|
---|
| 1985 |
|
---|
| 1986 | /* Initialize the mapping length (0..255) -> length code (0..28) */
|
---|
| 1987 | length = 0;
|
---|
| 1988 | for (code = 0; code < LENGTH_CODES - 1; code++) {
|
---|
| 1989 | G2.base_length[code] = length;
|
---|
| 1990 | for (n = 0; n < (1 << extra_lbits[code]); n++) {
|
---|
| 1991 | G2.length_code[length++] = code;
|
---|
| 1992 | }
|
---|
[821] | 1993 | }
|
---|
[1765] | 1994 | Assert(length == 256, "ct_init: length != 256");
|
---|
| 1995 | /* Note that the length 255 (match length 258) can be represented
|
---|
| 1996 | * in two different ways: code 284 + 5 bits or code 285, so we
|
---|
| 1997 | * overwrite length_code[255] to use the best encoding:
|
---|
| 1998 | */
|
---|
| 1999 | G2.length_code[length - 1] = code;
|
---|
[821] | 2000 |
|
---|
[1765] | 2001 | /* Initialize the mapping dist (0..32K) -> dist code (0..29) */
|
---|
| 2002 | dist = 0;
|
---|
| 2003 | for (code = 0; code < 16; code++) {
|
---|
| 2004 | G2.base_dist[code] = dist;
|
---|
| 2005 | for (n = 0; n < (1 << extra_dbits[code]); n++) {
|
---|
| 2006 | G2.dist_code[dist++] = code;
|
---|
| 2007 | }
|
---|
| 2008 | }
|
---|
| 2009 | Assert(dist == 256, "ct_init: dist != 256");
|
---|
| 2010 | dist >>= 7; /* from now on, all distances are divided by 128 */
|
---|
| 2011 | for (; code < D_CODES; code++) {
|
---|
| 2012 | G2.base_dist[code] = dist << 7;
|
---|
| 2013 | for (n = 0; n < (1 << (extra_dbits[code] - 7)); n++) {
|
---|
| 2014 | G2.dist_code[256 + dist++] = code;
|
---|
| 2015 | }
|
---|
| 2016 | }
|
---|
| 2017 | Assert(dist == 256, "ct_init: 256+dist != 512");
|
---|
[821] | 2018 |
|
---|
[1765] | 2019 | /* Construct the codes of the static literal tree */
|
---|
| 2020 | /* already zeroed - it's in bss
|
---|
| 2021 | for (n = 0; n <= MAX_BITS; n++)
|
---|
| 2022 | G2.bl_count[n] = 0; */
|
---|
[821] | 2023 |
|
---|
[1765] | 2024 | n = 0;
|
---|
| 2025 | while (n <= 143) {
|
---|
| 2026 | G2.static_ltree[n++].Len = 8;
|
---|
| 2027 | G2.bl_count[8]++;
|
---|
| 2028 | }
|
---|
| 2029 | while (n <= 255) {
|
---|
| 2030 | G2.static_ltree[n++].Len = 9;
|
---|
| 2031 | G2.bl_count[9]++;
|
---|
| 2032 | }
|
---|
| 2033 | while (n <= 279) {
|
---|
| 2034 | G2.static_ltree[n++].Len = 7;
|
---|
| 2035 | G2.bl_count[7]++;
|
---|
| 2036 | }
|
---|
| 2037 | while (n <= 287) {
|
---|
| 2038 | G2.static_ltree[n++].Len = 8;
|
---|
| 2039 | G2.bl_count[8]++;
|
---|
| 2040 | }
|
---|
| 2041 | /* Codes 286 and 287 do not exist, but we must include them in the
|
---|
| 2042 | * tree construction to get a canonical Huffman tree (longest code
|
---|
| 2043 | * all ones)
|
---|
| 2044 | */
|
---|
| 2045 | gen_codes((ct_data *) G2.static_ltree, L_CODES + 1);
|
---|
[821] | 2046 |
|
---|
[1765] | 2047 | /* The static distance tree is trivial: */
|
---|
| 2048 | for (n = 0; n < D_CODES; n++) {
|
---|
| 2049 | G2.static_dtree[n].Len = 5;
|
---|
| 2050 | G2.static_dtree[n].Code = bi_reverse(n, 5);
|
---|
| 2051 | }
|
---|
| 2052 |
|
---|
| 2053 | /* Initialize the first block of the first file: */
|
---|
| 2054 | init_block();
|
---|
[821] | 2055 | }
|
---|
| 2056 |
|
---|
| 2057 |
|
---|
| 2058 | /* ===========================================================================
|
---|
| 2059 | * Deflate in to out.
|
---|
| 2060 | * IN assertions: the input and output buffers are cleared.
|
---|
| 2061 | */
|
---|
[1765] | 2062 |
|
---|
[3621] | 2063 | static void zip(void)
|
---|
[821] | 2064 | {
|
---|
[1765] | 2065 | ush deflate_flags = 0; /* pkzip -es, -en or -ex equivalent */
|
---|
[821] | 2066 |
|
---|
[1765] | 2067 | G1.outcnt = 0;
|
---|
[821] | 2068 |
|
---|
| 2069 | /* Write the header to the gzip file. See algorithm.doc for the format */
|
---|
[1765] | 2070 | /* magic header for gzip files: 1F 8B */
|
---|
| 2071 | /* compression method: 8 (DEFLATED) */
|
---|
| 2072 | /* general flags: 0 */
|
---|
| 2073 | put_32bit(0x00088b1f);
|
---|
[3621] | 2074 | put_32bit(0); /* Unix timestamp */
|
---|
[821] | 2075 |
|
---|
| 2076 | /* Write deflated file to zip file */
|
---|
[1765] | 2077 | G1.crc = ~0;
|
---|
[821] | 2078 |
|
---|
[1765] | 2079 | bi_init();
|
---|
| 2080 | ct_init();
|
---|
[821] | 2081 | lm_init(&deflate_flags);
|
---|
| 2082 |
|
---|
[1765] | 2083 | put_8bit(deflate_flags); /* extra flags */
|
---|
| 2084 | put_8bit(3); /* OS identifier = 3 (Unix) */
|
---|
[821] | 2085 |
|
---|
[1765] | 2086 | deflate();
|
---|
[821] | 2087 |
|
---|
| 2088 | /* Write the crc and uncompressed size */
|
---|
[1765] | 2089 | put_32bit(~G1.crc);
|
---|
| 2090 | put_32bit(G1.isize);
|
---|
[821] | 2091 |
|
---|
| 2092 | flush_outbuf();
|
---|
| 2093 | }
|
---|
| 2094 |
|
---|
| 2095 |
|
---|
[1765] | 2096 | /* ======================================================================== */
|
---|
| 2097 | static
|
---|
[3621] | 2098 | IF_DESKTOP(long long) int FAST_FUNC pack_gzip(transformer_state_t *xstate UNUSED_PARAM)
|
---|
[821] | 2099 | {
|
---|
[2725] | 2100 | /* Clear input and output buffers */
|
---|
| 2101 | G1.outcnt = 0;
|
---|
| 2102 | #ifdef DEBUG
|
---|
| 2103 | G1.insize = 0;
|
---|
| 2104 | #endif
|
---|
| 2105 | G1.isize = 0;
|
---|
| 2106 |
|
---|
| 2107 | /* Reinit G2.xxx */
|
---|
| 2108 | memset(&G2, 0, sizeof(G2));
|
---|
| 2109 | G2.l_desc.dyn_tree = G2.dyn_ltree;
|
---|
| 2110 | G2.l_desc.static_tree = G2.static_ltree;
|
---|
| 2111 | G2.l_desc.extra_bits = extra_lbits;
|
---|
| 2112 | G2.l_desc.extra_base = LITERALS + 1;
|
---|
| 2113 | G2.l_desc.elems = L_CODES;
|
---|
| 2114 | G2.l_desc.max_length = MAX_BITS;
|
---|
| 2115 | //G2.l_desc.max_code = 0;
|
---|
| 2116 | G2.d_desc.dyn_tree = G2.dyn_dtree;
|
---|
| 2117 | G2.d_desc.static_tree = G2.static_dtree;
|
---|
| 2118 | G2.d_desc.extra_bits = extra_dbits;
|
---|
| 2119 | //G2.d_desc.extra_base = 0;
|
---|
| 2120 | G2.d_desc.elems = D_CODES;
|
---|
| 2121 | G2.d_desc.max_length = MAX_BITS;
|
---|
| 2122 | //G2.d_desc.max_code = 0;
|
---|
| 2123 | G2.bl_desc.dyn_tree = G2.bl_tree;
|
---|
| 2124 | //G2.bl_desc.static_tree = NULL;
|
---|
| 2125 | G2.bl_desc.extra_bits = extra_blbits,
|
---|
| 2126 | //G2.bl_desc.extra_base = 0;
|
---|
| 2127 | G2.bl_desc.elems = BL_CODES;
|
---|
| 2128 | G2.bl_desc.max_length = MAX_BL_BITS;
|
---|
| 2129 | //G2.bl_desc.max_code = 0;
|
---|
| 2130 |
|
---|
[3621] | 2131 | #if 0
|
---|
| 2132 | /* Saving of timestamp is disabled. Why?
|
---|
| 2133 | * - it is not Y2038-safe.
|
---|
| 2134 | * - some people want deterministic results
|
---|
| 2135 | * (normally they'd use -n, but our -n is a nop).
|
---|
| 2136 | * - it's bloat.
|
---|
| 2137 | * Per RFC 1952, gzfile.time=0 is "no timestamp".
|
---|
| 2138 | * If users will demand this to be reinstated,
|
---|
| 2139 | * implement -n "don't save timestamp".
|
---|
| 2140 | */
|
---|
| 2141 | struct stat s;
|
---|
[1765] | 2142 | s.st_ctime = 0;
|
---|
| 2143 | fstat(STDIN_FILENO, &s);
|
---|
| 2144 | zip(s.st_ctime);
|
---|
[3621] | 2145 | #else
|
---|
| 2146 | zip();
|
---|
| 2147 | #endif
|
---|
[1765] | 2148 | return 0;
|
---|
[821] | 2149 | }
|
---|
| 2150 |
|
---|
[2725] | 2151 | #if ENABLE_FEATURE_GZIP_LONG_OPTIONS
|
---|
| 2152 | static const char gzip_longopts[] ALIGN1 =
|
---|
| 2153 | "stdout\0" No_argument "c"
|
---|
| 2154 | "to-stdout\0" No_argument "c"
|
---|
| 2155 | "force\0" No_argument "f"
|
---|
| 2156 | "verbose\0" No_argument "v"
|
---|
| 2157 | #if ENABLE_GUNZIP
|
---|
| 2158 | "decompress\0" No_argument "d"
|
---|
| 2159 | "uncompress\0" No_argument "d"
|
---|
| 2160 | "test\0" No_argument "t"
|
---|
| 2161 | #endif
|
---|
| 2162 | "quiet\0" No_argument "q"
|
---|
| 2163 | "fast\0" No_argument "1"
|
---|
| 2164 | "best\0" No_argument "9"
|
---|
[3621] | 2165 | "no-name\0" No_argument "n"
|
---|
[2725] | 2166 | ;
|
---|
| 2167 | #endif
|
---|
| 2168 |
|
---|
| 2169 | /*
|
---|
| 2170 | * Linux kernel build uses gzip -d -n. We accept and ignore -n.
|
---|
| 2171 | * Man page says:
|
---|
| 2172 | * -n --no-name
|
---|
| 2173 | * gzip: do not save the original file name and time stamp.
|
---|
| 2174 | * (The original name is always saved if the name had to be truncated.)
|
---|
| 2175 | * gunzip: do not restore the original file name/time even if present
|
---|
| 2176 | * (remove only the gzip suffix from the compressed file name).
|
---|
| 2177 | * This option is the default when decompressing.
|
---|
| 2178 | * -N --name
|
---|
| 2179 | * gzip: always save the original file name and time stamp (this is the default)
|
---|
| 2180 | * gunzip: restore the original file name and time stamp if present.
|
---|
| 2181 | */
|
---|
| 2182 |
|
---|
| 2183 | int gzip_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
|
---|
| 2184 | #if ENABLE_GUNZIP
|
---|
[1765] | 2185 | int gzip_main(int argc, char **argv)
|
---|
[2725] | 2186 | #else
|
---|
| 2187 | int gzip_main(int argc UNUSED_PARAM, char **argv)
|
---|
| 2188 | #endif
|
---|
[821] | 2189 | {
|
---|
[1765] | 2190 | unsigned opt;
|
---|
[3621] | 2191 | #ifdef ENABLE_FEATURE_GZIP_LEVELS
|
---|
| 2192 | static const struct {
|
---|
| 2193 | uint8_t good;
|
---|
| 2194 | uint8_t chain_shift;
|
---|
| 2195 | uint8_t lazy2;
|
---|
| 2196 | uint8_t nice2;
|
---|
| 2197 | } gzip_level_config[6] = {
|
---|
| 2198 | {4, 4, 4/2, 16/2}, /* Level 4 */
|
---|
| 2199 | {8, 5, 16/2, 32/2}, /* Level 5 */
|
---|
| 2200 | {8, 7, 16/2, 128/2}, /* Level 6 */
|
---|
| 2201 | {8, 8, 32/2, 128/2}, /* Level 7 */
|
---|
| 2202 | {32, 10, 128/2, 258/2}, /* Level 8 */
|
---|
| 2203 | {32, 12, 258/2, 258/2}, /* Level 9 */
|
---|
| 2204 | };
|
---|
| 2205 | #endif
|
---|
[821] | 2206 |
|
---|
[3621] | 2207 | SET_PTR_TO_GLOBALS((char *)xzalloc(sizeof(struct globals)+sizeof(struct globals2))
|
---|
| 2208 | + sizeof(struct globals));
|
---|
| 2209 |
|
---|
[2725] | 2210 | #if ENABLE_FEATURE_GZIP_LONG_OPTIONS
|
---|
| 2211 | applet_long_options = gzip_longopts;
|
---|
| 2212 | #endif
|
---|
[1765] | 2213 | /* Must match bbunzip's constants OPT_STDOUT, OPT_FORCE! */
|
---|
[3621] | 2214 | opt = getopt32(argv, "cfv" IF_GUNZIP("dt") "qn123456789");
|
---|
[2725] | 2215 | #if ENABLE_GUNZIP /* gunzip_main may not be visible... */
|
---|
| 2216 | if (opt & 0x18) // -d and/or -t
|
---|
| 2217 | return gunzip_main(argc, argv);
|
---|
| 2218 | #endif
|
---|
[3621] | 2219 | #ifdef ENABLE_FEATURE_GZIP_LEVELS
|
---|
| 2220 | opt >>= ENABLE_GUNZIP ? 7 : 5; /* drop cfv[dt]qn bits */
|
---|
| 2221 | if (opt == 0)
|
---|
| 2222 | opt = 1 << 6; /* default: 6 */
|
---|
| 2223 | opt = ffs(opt >> 4); /* Maps -1..-4 to [0], -5 to [1] ... -9 to [5] */
|
---|
| 2224 | max_chain_length = 1 << gzip_level_config[opt].chain_shift;
|
---|
| 2225 | good_match = gzip_level_config[opt].good;
|
---|
| 2226 | max_lazy_match = gzip_level_config[opt].lazy2 * 2;
|
---|
| 2227 | nice_match = gzip_level_config[opt].nice2 * 2;
|
---|
| 2228 | #endif
|
---|
| 2229 | option_mask32 &= 0x7; /* retain only -cfv */
|
---|
[1765] | 2230 |
|
---|
| 2231 | /* Allocate all global buffers (for DYN_ALLOC option) */
|
---|
| 2232 | ALLOC(uch, G1.l_buf, INBUFSIZ);
|
---|
| 2233 | ALLOC(uch, G1.outbuf, OUTBUFSIZ);
|
---|
| 2234 | ALLOC(ush, G1.d_buf, DIST_BUFSIZE);
|
---|
| 2235 | ALLOC(uch, G1.window, 2L * WSIZE);
|
---|
| 2236 | ALLOC(ush, G1.prev, 1L << BITS);
|
---|
| 2237 |
|
---|
[2725] | 2238 | /* Initialize the CRC32 table */
|
---|
| 2239 | global_crc32_table = crc32_filltable(NULL, 0);
|
---|
[1765] | 2240 |
|
---|
[3621] | 2241 | argv += optind;
|
---|
[2725] | 2242 | return bbunpack(argv, pack_gzip, append_ext, "gz");
|
---|
[821] | 2243 | }
|
---|