David Li | e54acca | 2011-03-18 16:21:06 -0700 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (c) 2006 Stefan Traby <stefan@hello-penguin.com> |
| 3 | * |
| 4 | * Redistribution and use in source and binary forms, with or without modifica- |
| 5 | * tion, are permitted provided that the following conditions are met: |
| 6 | * |
| 7 | * 1. Redistributions of source code must retain the above copyright notice, |
| 8 | * this list of conditions and the following disclaimer. |
| 9 | * |
| 10 | * 2. Redistributions in binary form must reproduce the above copyright |
| 11 | * notice, this list of conditions and the following disclaimer in the |
| 12 | * documentation and/or other materials provided with the distribution. |
| 13 | * |
| 14 | * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED |
| 15 | * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER- |
| 16 | * CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO |
| 17 | * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPE- |
| 18 | * CIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, |
| 19 | * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; |
| 20 | * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, |
| 21 | * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTH- |
| 22 | * ERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED |
| 23 | * OF THE POSSIBILITY OF SUCH DAMAGE. |
| 24 | * |
| 25 | * Alternatively, the contents of this file may be used under the terms of |
| 26 | * the GNU General Public License ("GPL") version 2 or any later version, |
| 27 | * in which case the provisions of the GPL are applicable instead of |
| 28 | * the above. If you wish to allow the use of your version of this file |
| 29 | * only under the terms of the GPL and not to allow others to use your |
| 30 | * version of this file under the BSD license, indicate your decision |
| 31 | * by deleting the provisions above and replace them with the notice |
| 32 | * and other provisions required by the GPL. If you do not delete the |
| 33 | * provisions above, a recipient may use your version of this file under |
| 34 | * either the BSD or the GPL. |
| 35 | */ |
| 36 | |
| 37 | #include "config.h" |
| 38 | #include <stdio.h> |
| 39 | #include <string.h> |
| 40 | #include <stdlib.h> |
| 41 | #include <unistd.h> |
| 42 | #include <sys/types.h> |
| 43 | #include <sys/stat.h> |
| 44 | #include <fcntl.h> |
| 45 | #include <errno.h> |
| 46 | #include <limits.h> |
| 47 | #include "lzf.h" |
| 48 | |
| 49 | #ifdef HAVE_GETOPT_H |
| 50 | # include <getopt.h> |
| 51 | #endif |
| 52 | |
| 53 | #define BLOCKSIZE (1024 * 64 - 1) |
| 54 | #define MAX_BLOCKSIZE BLOCKSIZE |
| 55 | |
| 56 | typedef unsigned char u8; |
| 57 | |
| 58 | static off_t nr_read, nr_written; |
| 59 | |
| 60 | static const char *imagename; |
| 61 | static enum { compress, uncompress, lzcat } mode = compress; |
| 62 | static int verbose = 0; |
| 63 | static int force = 0; |
| 64 | static long blocksize = BLOCKSIZE; |
| 65 | |
| 66 | #ifdef HAVE_GETOPT_LONG |
| 67 | |
| 68 | struct option longopts[] = { |
| 69 | {"compress", 0, 0, 'c'}, |
| 70 | {"decompress", 0, 0, 'd'}, |
| 71 | {"uncompress", 0, 0, 'd'}, |
| 72 | {"force", 0, 0, 'f'}, |
| 73 | {"help", 0, 0, 'h'}, |
| 74 | {"verbose", 0, 0, 'v'}, |
| 75 | {"blocksize", 1, 0, 'b'}, |
| 76 | {0, 0, 0, 0} |
| 77 | }; |
| 78 | |
| 79 | static const char *opt = |
| 80 | "-c --compress compress\n" |
| 81 | "-d --decompress decompress\n" |
| 82 | "-f --force force overwrite of output file\n" |
| 83 | "-h --help give this help\n" "-v --verbose verbose mode\n" "-b # --blocksize # set blocksize\n" "\n"; |
| 84 | |
| 85 | #else |
| 86 | |
| 87 | static const char *opt = |
| 88 | "-c compress\n" |
| 89 | "-d decompress\n" |
| 90 | "-f force overwrite of output file\n" |
| 91 | "-h give this help\n" |
| 92 | "-v verbose mode\n" |
| 93 | "-b # set blocksize\n" |
| 94 | "\n"; |
| 95 | |
| 96 | #endif |
| 97 | |
| 98 | static void |
| 99 | usage (int rc) |
| 100 | { |
| 101 | fprintf (stderr, "\n" |
| 102 | "lzf, a very lightweight compression/decompression utility written by Stefan Traby.\n" |
| 103 | "uses liblzf written by Marc Lehmann <schmorp@schmorp.de> You can find more info at\n" |
| 104 | "http://liblzf.plan9.de/\n" |
| 105 | "\n" |
| 106 | "usage: lzf [-dufhvb] [file ...]\n" |
| 107 | " unlzf [file ...]\n" |
| 108 | " lzcat [file ...]\n" |
| 109 | "\n%s", |
| 110 | opt); |
| 111 | |
| 112 | exit (rc); |
| 113 | } |
| 114 | |
| 115 | static inline ssize_t |
| 116 | rread (int fd, void *buf, size_t len) |
| 117 | { |
| 118 | ssize_t rc = 0, offset = 0; |
| 119 | char *p = buf; |
| 120 | |
| 121 | while (len && (rc = read (fd, &p[offset], len)) > 0) |
| 122 | { |
| 123 | offset += rc; |
| 124 | len -= rc; |
| 125 | } |
| 126 | |
| 127 | nr_read += offset; |
| 128 | |
| 129 | if (rc < 0) |
| 130 | return rc; |
| 131 | |
| 132 | return offset; |
| 133 | } |
| 134 | |
| 135 | /* returns 0 if all written else -1 */ |
| 136 | static inline ssize_t |
| 137 | wwrite (int fd, void *buf, size_t len) |
| 138 | { |
| 139 | ssize_t rc; |
| 140 | char *b = buf; |
| 141 | size_t l = len; |
| 142 | |
| 143 | while (l) |
| 144 | { |
| 145 | rc = write (fd, b, l); |
| 146 | if (rc < 0) |
| 147 | { |
| 148 | fprintf (stderr, "%s: write error: ", imagename); |
| 149 | perror (""); |
| 150 | return -1; |
| 151 | } |
| 152 | |
| 153 | l -= rc; |
| 154 | b += rc; |
| 155 | } |
| 156 | |
| 157 | nr_written += len; |
| 158 | return 0; |
| 159 | } |
| 160 | |
| 161 | /* |
| 162 | * Anatomy: an lzf file consists of any number of blocks in the following format: |
| 163 | * |
| 164 | * \x00 EOF (optional) |
| 165 | * "ZV\0" 2-byte-usize <uncompressed data> |
| 166 | * "ZV\1" 2-byte-csize 2-byte-usize <compressed data> |
| 167 | * "ZV\2" 4-byte-crc32-0xdebb20e3 (NYI) |
| 168 | */ |
| 169 | |
| 170 | |
| 171 | #define TYPE0_HDR_SIZE 5 |
| 172 | #define TYPE1_HDR_SIZE 7 |
| 173 | #define MAX_HDR_SIZE 7 |
| 174 | #define MIN_HDR_SIZE 5 |
| 175 | |
| 176 | static int |
| 177 | compress_fd (int from, int to) |
| 178 | { |
| 179 | ssize_t us, cs, len; |
| 180 | u8 buf1[MAX_BLOCKSIZE + MAX_HDR_SIZE + 16]; |
| 181 | u8 buf2[MAX_BLOCKSIZE + MAX_HDR_SIZE + 16]; |
| 182 | u8 *header; |
| 183 | |
| 184 | nr_read = nr_written = 0; |
| 185 | while ((us = rread (from, &buf1[MAX_HDR_SIZE], blocksize)) > 0) |
| 186 | { |
| 187 | cs = lzf_compress (&buf1[MAX_HDR_SIZE], us, &buf2[MAX_HDR_SIZE], us > 4 ? us - 4 : us); |
| 188 | if (cs) |
| 189 | { |
| 190 | header = &buf2[MAX_HDR_SIZE - TYPE1_HDR_SIZE]; |
| 191 | header[0] = 'Z'; |
| 192 | header[1] = 'V'; |
| 193 | header[2] = 1; |
| 194 | header[3] = cs >> 8; |
| 195 | header[4] = cs & 0xff; |
| 196 | header[5] = us >> 8; |
| 197 | header[6] = us & 0xff; |
| 198 | len = cs + TYPE1_HDR_SIZE; |
| 199 | } |
| 200 | else |
| 201 | { // write uncompressed |
| 202 | header = &buf1[MAX_HDR_SIZE - TYPE0_HDR_SIZE]; |
| 203 | header[0] = 'Z'; |
| 204 | header[1] = 'V'; |
| 205 | header[2] = 0; |
| 206 | header[3] = us >> 8; |
| 207 | header[4] = us & 0xff; |
| 208 | len = us + TYPE0_HDR_SIZE; |
| 209 | } |
| 210 | |
| 211 | if (wwrite (to, header, len) == -1) |
| 212 | return -1; |
| 213 | } |
| 214 | |
| 215 | return 0; |
| 216 | } |
| 217 | |
| 218 | static int |
| 219 | uncompress_fd (int from, int to) |
| 220 | { |
| 221 | u8 header[MAX_HDR_SIZE]; |
| 222 | u8 buf1[MAX_BLOCKSIZE + MAX_HDR_SIZE + 16]; |
| 223 | u8 buf2[MAX_BLOCKSIZE + MAX_HDR_SIZE + 16]; |
| 224 | u8 *p; |
| 225 | int l, rd; |
| 226 | ssize_t rc, cs, us, bytes, over = 0; |
| 227 | |
| 228 | nr_read = nr_written = 0; |
| 229 | while (1) |
| 230 | { |
| 231 | rc = rread (from, header + over, MAX_HDR_SIZE - over); |
| 232 | if (rc < 0) |
| 233 | { |
| 234 | fprintf (stderr, "%s: read error: ", imagename); |
| 235 | perror (""); |
| 236 | return -1; |
| 237 | } |
| 238 | |
| 239 | rc += over; |
| 240 | over = 0; |
| 241 | if (!rc || header[0] == 0) |
| 242 | return 0; |
| 243 | |
| 244 | if (rc < MIN_HDR_SIZE || header[0] != 'Z' || header[1] != 'V') |
| 245 | { |
| 246 | fprintf (stderr, "%s: invalid data stream - magic not found or short header\n", imagename); |
| 247 | return -1; |
| 248 | } |
| 249 | |
| 250 | switch (header[2]) |
| 251 | { |
| 252 | case 0: |
| 253 | cs = -1; |
| 254 | us = (header[3] << 8) | header[4]; |
| 255 | p = &header[TYPE0_HDR_SIZE]; |
| 256 | break; |
| 257 | case 1: |
| 258 | if (rc < TYPE1_HDR_SIZE) |
| 259 | { |
| 260 | goto short_read; |
| 261 | } |
| 262 | cs = (header[3] << 8) | header[4]; |
| 263 | us = (header[5] << 8) | header[6]; |
| 264 | p = &header[TYPE1_HDR_SIZE]; |
| 265 | break; |
| 266 | default: |
| 267 | fprintf (stderr, "%s: unknown blocktype\n", imagename); |
| 268 | return -1; |
| 269 | } |
| 270 | |
| 271 | bytes = cs == -1 ? us : cs; |
| 272 | l = &header[rc] - p; |
| 273 | |
| 274 | if (l > 0) |
| 275 | memcpy (buf1, p, l); |
| 276 | |
| 277 | if (l > bytes) |
| 278 | { |
| 279 | over = l - bytes; |
| 280 | memmove (header, &p[bytes], over); |
| 281 | } |
| 282 | |
| 283 | p = &buf1[l]; |
| 284 | rd = bytes - l; |
| 285 | if (rd > 0) |
| 286 | if ((rc = rread (from, p, rd)) != rd) |
| 287 | goto short_read; |
| 288 | |
| 289 | if (cs == -1) |
| 290 | { |
| 291 | if (wwrite (to, buf1, us)) |
| 292 | return -1; |
| 293 | } |
| 294 | else |
| 295 | { |
| 296 | if (lzf_decompress (buf1, cs, buf2, us) != us) |
| 297 | { |
| 298 | fprintf (stderr, "%s: decompress: invalid stream - data corrupted\n", imagename); |
| 299 | return -1; |
| 300 | } |
| 301 | |
| 302 | if (wwrite (to, buf2, us)) |
| 303 | return -1; |
| 304 | } |
| 305 | } |
| 306 | |
| 307 | return 0; |
| 308 | |
| 309 | short_read: |
| 310 | fprintf (stderr, "%s: short data\n", imagename); |
| 311 | return -1; |
| 312 | } |
| 313 | |
| 314 | static int |
| 315 | open_out (const char *name) |
| 316 | { |
| 317 | int fd; |
| 318 | int m = O_EXCL; |
| 319 | |
| 320 | if (force) |
| 321 | m = 0; |
| 322 | |
| 323 | fd = open (name, O_CREAT | O_WRONLY | O_TRUNC | m, 600); |
| 324 | #if defined(__MINGW32__) |
| 325 | _setmode(fd, _O_BINARY); |
| 326 | #endif |
| 327 | return fd; |
| 328 | } |
| 329 | |
| 330 | static int |
| 331 | compose_name (const char *fname, char *oname) |
| 332 | { |
| 333 | char *p; |
| 334 | |
| 335 | if (mode == compress) |
| 336 | { |
| 337 | if (strlen (fname) > PATH_MAX - 4) |
| 338 | { |
| 339 | fprintf (stderr, "%s: %s.lzf: name too long", imagename, fname); |
| 340 | return -1; |
| 341 | } |
| 342 | |
| 343 | strcpy (oname, fname); |
| 344 | strcat (oname, ".lzf"); |
| 345 | } |
| 346 | else |
| 347 | { |
| 348 | if (strlen (fname) > PATH_MAX) |
| 349 | { |
| 350 | fprintf (stderr, "%s: %s: name too long\n", imagename, fname); |
| 351 | return -1; |
| 352 | } |
| 353 | |
| 354 | strcpy (oname, fname); |
| 355 | p = &oname[strlen (oname)] - 4; |
| 356 | if (p < oname || strcmp (p, ".lzf")) |
| 357 | { |
| 358 | fprintf (stderr, "%s: %s: unknown suffix\n", imagename, fname); |
| 359 | return -1; |
| 360 | } |
| 361 | |
| 362 | *p = 0; |
| 363 | } |
| 364 | |
| 365 | return 0; |
| 366 | } |
| 367 | |
| 368 | static int |
| 369 | run_file (const char *fname) |
| 370 | { |
| 371 | int fd, fd2; |
| 372 | int rc; |
| 373 | struct stat mystat; |
| 374 | char oname[PATH_MAX + 1]; |
| 375 | |
| 376 | if (mode != lzcat) |
| 377 | if (compose_name (fname, oname)) |
| 378 | return -1; |
| 379 | |
| 380 | #if !defined(__MINGW32__) |
| 381 | rc = lstat (fname, &mystat); |
| 382 | #else |
| 383 | rc = stat (fname, &mystat); |
| 384 | #endif |
| 385 | fd = open (fname, O_RDONLY); |
| 386 | #if defined(__MINGW32__) |
| 387 | _setmode(fd, _O_BINARY); |
| 388 | #endif |
| 389 | if (rc || fd == -1) |
| 390 | { |
| 391 | fprintf (stderr, "%s: %s: ", imagename, fname); |
| 392 | perror (""); |
| 393 | return -1; |
| 394 | } |
| 395 | |
| 396 | if (!S_ISREG (mystat.st_mode)) |
| 397 | { |
| 398 | fprintf (stderr, "%s: %s: not a regular file.\n", imagename, fname); |
| 399 | close (fd); |
| 400 | return -1; |
| 401 | } |
| 402 | |
| 403 | if (mode == lzcat) |
| 404 | { |
| 405 | rc = uncompress_fd (fd, 1); |
| 406 | close (fd); |
| 407 | return rc; |
| 408 | } |
| 409 | |
| 410 | fd2 = open_out (oname); |
| 411 | if (fd2 == -1) |
| 412 | { |
| 413 | fprintf (stderr, "%s: %s: ", imagename, oname); |
| 414 | perror (""); |
| 415 | close (fd); |
| 416 | return -1; |
| 417 | } |
| 418 | |
| 419 | if (mode == compress) |
| 420 | { |
| 421 | rc = compress_fd (fd, fd2); |
| 422 | if (!rc && verbose) |
| 423 | fprintf (stderr, "%s: %5.1f%% -- replaced with %s\n", |
| 424 | fname, nr_read == 0 ? 0 : 100.0 - nr_written / ((double) nr_read / 100.0), oname); |
| 425 | } |
| 426 | else |
| 427 | { |
| 428 | rc = uncompress_fd (fd, fd2); |
| 429 | if (!rc && verbose) |
| 430 | fprintf (stderr, "%s: %5.1f%% -- replaced with %s\n", |
| 431 | fname, nr_written == 0 ? 0 : 100.0 - nr_read / ((double) nr_written / 100.0), oname); |
| 432 | } |
| 433 | |
| 434 | #if !defined(__MINGW32__) |
| 435 | fchmod (fd2, mystat.st_mode); |
| 436 | #else |
| 437 | chmod (oname, mystat.st_mode); |
| 438 | #endif |
| 439 | close (fd); |
| 440 | close (fd2); |
| 441 | |
| 442 | if (!rc) |
| 443 | unlink (fname); |
| 444 | |
| 445 | return rc; |
| 446 | } |
| 447 | |
| 448 | int |
| 449 | main (int argc, char *argv[]) |
| 450 | { |
| 451 | char *p = argv[0]; |
| 452 | int optc; |
| 453 | int rc = 0; |
| 454 | |
| 455 | errno = 0; |
| 456 | p = getenv ("LZF_BLOCKSIZE"); |
| 457 | if (p) |
| 458 | { |
| 459 | blocksize = strtoul (p, 0, 0); |
| 460 | if (errno || !blocksize || blocksize > MAX_BLOCKSIZE) |
| 461 | blocksize = BLOCKSIZE; |
| 462 | } |
| 463 | |
| 464 | p = strrchr (argv[0], '/'); |
| 465 | imagename = p ? ++p : argv[0]; |
| 466 | |
| 467 | if (!strncmp (imagename, "un", 2) || !strncmp (imagename, "de", 2)) |
| 468 | mode = uncompress; |
| 469 | |
| 470 | if (strstr (imagename, "cat")) |
| 471 | mode = lzcat; |
| 472 | |
| 473 | #ifdef HAVE_GETOPT_LONG |
| 474 | while ((optc = getopt_long (argc, argv, "cdfhvb:", longopts, 0)) != -1) |
| 475 | #else |
| 476 | while ((optc = getopt (argc, argv, "cdfhvb:")) != -1) |
| 477 | #endif |
| 478 | { |
| 479 | switch (optc) |
| 480 | { |
| 481 | case 'c': |
| 482 | mode = compress; |
| 483 | break; |
| 484 | case 'd': |
| 485 | mode = uncompress; |
| 486 | break; |
| 487 | case 'f': |
| 488 | force = 1; |
| 489 | break; |
| 490 | case 'h': |
| 491 | usage (0); |
| 492 | break; |
| 493 | case 'v': |
| 494 | verbose = 1; |
| 495 | break; |
| 496 | case 'b': |
| 497 | errno = 0; |
| 498 | blocksize = strtoul (optarg, 0, 0); |
| 499 | if (errno || !blocksize || blocksize > MAX_BLOCKSIZE) |
| 500 | blocksize = BLOCKSIZE; |
| 501 | break; |
| 502 | default: |
| 503 | usage (1); |
| 504 | break; |
| 505 | } |
| 506 | } |
| 507 | |
| 508 | if (optind == argc) |
| 509 | { // stdin stdout |
| 510 | if (!force) |
| 511 | { |
| 512 | if ((mode == uncompress || mode == lzcat) && isatty (0)) |
| 513 | { |
| 514 | fprintf (stderr, "%s: compressed data not read from a terminal. Use -f to force decompression.\n", imagename); |
| 515 | exit (1); |
| 516 | } |
| 517 | if (mode == compress && isatty (1)) |
| 518 | { |
| 519 | fprintf (stderr, "%s: compressed data not written to a terminal. Use -f to force compression.\n", imagename); |
| 520 | exit (1); |
| 521 | } |
| 522 | } |
| 523 | |
| 524 | if (mode == compress) |
| 525 | rc = compress_fd (0, 1); |
| 526 | else |
| 527 | rc = uncompress_fd (0, 1); |
| 528 | |
| 529 | exit (rc ? 1 : 0); |
| 530 | } |
| 531 | |
| 532 | while (optind < argc) |
| 533 | rc |= run_file (argv[optind++]); |
| 534 | |
| 535 | exit (rc ? 1 : 0); |
| 536 | } |
| 537 | |