1 /*- 2 * Copyright (c) 2003-2007 Tim Kientzle 3 * All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 8 * 1. Redistributions of source code must retain the above copyright 9 * notice, this list of conditions and the following disclaimer. 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(S) ``AS IS'' AND ANY EXPRESS OR 15 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 16 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 17 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT, 18 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 19 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 20 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 21 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 22 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 23 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 24 */ 25 26 /* 27 * This code borrows heavily from "compress" source code, which is 28 * protected by the following copyright. (Clause 3 dropped by request 29 * of the Regents.) 30 */ 31 32 /*- 33 * Copyright (c) 1985, 1986, 1992, 1993 34 * The Regents of the University of California. All rights reserved. 35 * 36 * This code is derived from software contributed to Berkeley by 37 * Diomidis Spinellis and James A. Woods, derived from original 38 * work by Spencer Thomas and Joseph Orost. 39 * 40 * Redistribution and use in source and binary forms, with or without 41 * modification, are permitted provided that the following conditions 42 * are met: 43 * 1. Redistributions of source code must retain the above copyright 44 * notice, this list of conditions and the following disclaimer. 45 * 2. Redistributions in binary form must reproduce the above copyright 46 * notice, this list of conditions and the following disclaimer in the 47 * documentation and/or other materials provided with the distribution. 48 * 4. Neither the name of the University nor the names of its contributors 49 * may be used to endorse or promote products derived from this software 50 * without specific prior written permission. 51 * 52 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 53 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 54 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 55 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 56 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 57 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 58 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 59 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 60 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 61 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 62 * SUCH DAMAGE. 63 */ 64 65 66 #include "archive_platform.h" 67 __FBSDID("$FreeBSD$"); 68 69 #ifdef HAVE_ERRNO_H 70 #include <errno.h> 71 #endif 72 #ifdef HAVE_STDLIB_H 73 #include <stdlib.h> 74 #endif 75 #ifdef HAVE_STRING_H 76 #include <string.h> 77 #endif 78 #ifdef HAVE_UNISTD_H 79 #include <unistd.h> 80 #endif 81 82 #include "archive.h" 83 #include "archive_private.h" 84 #include "archive_read_private.h" 85 86 /* 87 * Because LZW decompression is pretty simple, I've just implemented 88 * the whole decompressor here (cribbing from "compress" source code, 89 * of course), rather than relying on an external library. I have 90 * made an effort to clarify and simplify the algorithm, so the 91 * names and structure here don't exactly match those used by compress. 92 */ 93 94 struct private_data { 95 /* Input variables. */ 96 const unsigned char *next_in; 97 size_t avail_in; 98 size_t consume_unnotified; 99 int bit_buffer; 100 int bits_avail; 101 size_t bytes_in_section; 102 103 /* Output variables. */ 104 size_t out_block_size; 105 void *out_block; 106 107 /* Decompression status variables. */ 108 int use_reset_code; 109 int end_of_stream; /* EOF status. */ 110 int maxcode; /* Largest code. */ 111 int maxcode_bits; /* Length of largest code. */ 112 int section_end_code; /* When to increase bits. */ 113 int bits; /* Current code length. */ 114 int oldcode; /* Previous code. */ 115 int finbyte; /* Last byte of prev code. */ 116 117 /* Dictionary. */ 118 int free_ent; /* Next dictionary entry. */ 119 unsigned char suffix[65536]; 120 uint16_t prefix[65536]; 121 122 /* 123 * Scratch area for expanding dictionary entries. Note: 124 * "worst" case here comes from compressing /dev/zero: the 125 * last code in the dictionary will code a sequence of 126 * 65536-256 zero bytes. Thus, we need stack space to expand 127 * a 65280-byte dictionary entry. (Of course, 32640:1 128 * compression could also be considered the "best" case. ;-) 129 */ 130 unsigned char *stackp; 131 unsigned char stack[65300]; 132 }; 133 134 static int compress_bidder_bid(struct archive_read_filter_bidder *, struct archive_read_filter *); 135 static int compress_bidder_init(struct archive_read_filter *); 136 static int compress_bidder_free(struct archive_read_filter_bidder *); 137 138 static ssize_t compress_filter_read(struct archive_read_filter *, const void **); 139 static int compress_filter_close(struct archive_read_filter *); 140 141 static int getbits(struct archive_read_filter *, int n); 142 static int next_code(struct archive_read_filter *); 143 144 #if ARCHIVE_VERSION_NUMBER < 4000000 145 /* Deprecated; remove in libarchive 4.0 */ 146 int 147 archive_read_support_compression_compress(struct archive *a) 148 { 149 return archive_read_support_filter_compress(a); 150 } 151 #endif 152 153 int 154 archive_read_support_filter_compress(struct archive *_a) 155 { 156 struct archive_read *a = (struct archive_read *)_a; 157 struct archive_read_filter_bidder *bidder; 158 159 archive_check_magic(_a, ARCHIVE_READ_MAGIC, 160 ARCHIVE_STATE_NEW, "archive_read_support_filter_compress"); 161 162 if (__archive_read_get_bidder(a, &bidder) != ARCHIVE_OK) 163 return (ARCHIVE_FATAL); 164 165 bidder->data = NULL; 166 bidder->bid = compress_bidder_bid; 167 bidder->init = compress_bidder_init; 168 bidder->options = NULL; 169 bidder->free = compress_bidder_free; 170 return (ARCHIVE_OK); 171 } 172 173 /* 174 * Test whether we can handle this data. 175 * This logic returns zero if any part of the signature fails. 176 */ 177 static int 178 compress_bidder_bid(struct archive_read_filter_bidder *self, 179 struct archive_read_filter *filter) 180 { 181 const unsigned char *buffer; 182 ssize_t avail; 183 int bits_checked; 184 185 (void)self; /* UNUSED */ 186 187 buffer = __archive_read_filter_ahead(filter, 2, &avail); 188 189 if (buffer == NULL) 190 return (0); 191 192 bits_checked = 0; 193 if (buffer[0] != 0x1F || buffer[1] != 0x9D) 194 return (0); 195 bits_checked += 16; 196 197 /* 198 * TODO: Verify more. 199 */ 200 201 return (bits_checked); 202 } 203 204 /* 205 * Setup the callbacks. 206 */ 207 static int 208 compress_bidder_init(struct archive_read_filter *self) 209 { 210 struct private_data *state; 211 static const size_t out_block_size = 64 * 1024; 212 void *out_block; 213 int code; 214 215 self->code = ARCHIVE_COMPRESSION_COMPRESS; 216 self->name = "compress (.Z)"; 217 218 state = (struct private_data *)calloc(sizeof(*state), 1); 219 out_block = malloc(out_block_size); 220 if (state == NULL || out_block == NULL) { 221 free(out_block); 222 free(state); 223 archive_set_error(&self->archive->archive, ENOMEM, 224 "Can't allocate data for %s decompression", 225 self->name); 226 return (ARCHIVE_FATAL); 227 } 228 229 self->data = state; 230 state->out_block_size = out_block_size; 231 state->out_block = out_block; 232 self->read = compress_filter_read; 233 self->skip = NULL; /* not supported */ 234 self->close = compress_filter_close; 235 236 /* XXX MOVE THE FOLLOWING OUT OF INIT() XXX */ 237 238 (void)getbits(self, 8); /* Skip first signature byte. */ 239 (void)getbits(self, 8); /* Skip second signature byte. */ 240 241 code = getbits(self, 8); 242 state->maxcode_bits = code & 0x1f; 243 state->maxcode = (1 << state->maxcode_bits); 244 state->use_reset_code = code & 0x80; 245 246 /* Initialize decompressor. */ 247 state->free_ent = 256; 248 state->stackp = state->stack; 249 if (state->use_reset_code) 250 state->free_ent++; 251 state->bits = 9; 252 state->section_end_code = (1<<state->bits) - 1; 253 state->oldcode = -1; 254 for (code = 255; code >= 0; code--) { 255 state->prefix[code] = 0; 256 state->suffix[code] = code; 257 } 258 next_code(self); 259 260 return (ARCHIVE_OK); 261 } 262 263 /* 264 * Return a block of data from the decompression buffer. Decompress more 265 * as necessary. 266 */ 267 static ssize_t 268 compress_filter_read(struct archive_read_filter *self, const void **pblock) 269 { 270 struct private_data *state; 271 unsigned char *p, *start, *end; 272 int ret; 273 274 state = (struct private_data *)self->data; 275 if (state->end_of_stream) { 276 *pblock = NULL; 277 return (0); 278 } 279 p = start = (unsigned char *)state->out_block; 280 end = start + state->out_block_size; 281 282 while (p < end && !state->end_of_stream) { 283 if (state->stackp > state->stack) { 284 *p++ = *--state->stackp; 285 } else { 286 ret = next_code(self); 287 if (ret == -1) 288 state->end_of_stream = ret; 289 else if (ret != ARCHIVE_OK) 290 return (ret); 291 } 292 } 293 294 *pblock = start; 295 return (p - start); 296 } 297 298 /* 299 * Clean up the reader. 300 */ 301 static int 302 compress_bidder_free(struct archive_read_filter_bidder *self) 303 { 304 self->data = NULL; 305 return (ARCHIVE_OK); 306 } 307 308 /* 309 * Close and release the filter. 310 */ 311 static int 312 compress_filter_close(struct archive_read_filter *self) 313 { 314 struct private_data *state = (struct private_data *)self->data; 315 316 free(state->out_block); 317 free(state); 318 return (ARCHIVE_OK); 319 } 320 321 /* 322 * Process the next code and fill the stack with the expansion 323 * of the code. Returns ARCHIVE_FATAL if there is a fatal I/O or 324 * format error, ARCHIVE_EOF if we hit end of data, ARCHIVE_OK otherwise. 325 */ 326 static int 327 next_code(struct archive_read_filter *self) 328 { 329 struct private_data *state = (struct private_data *)self->data; 330 int code, newcode; 331 332 static int debug_buff[1024]; 333 static unsigned debug_index; 334 335 code = newcode = getbits(self, state->bits); 336 if (code < 0) 337 return (code); 338 339 debug_buff[debug_index++] = code; 340 if (debug_index >= sizeof(debug_buff)/sizeof(debug_buff[0])) 341 debug_index = 0; 342 343 /* If it's a reset code, reset the dictionary. */ 344 if ((code == 256) && state->use_reset_code) { 345 /* 346 * The original 'compress' implementation blocked its 347 * I/O in a manner that resulted in junk bytes being 348 * inserted after every reset. The next section skips 349 * this junk. (Yes, the number of *bytes* to skip is 350 * a function of the current *bit* length.) 351 */ 352 int skip_bytes = state->bits - 353 (state->bytes_in_section % state->bits); 354 skip_bytes %= state->bits; 355 state->bits_avail = 0; /* Discard rest of this byte. */ 356 while (skip_bytes-- > 0) { 357 code = getbits(self, 8); 358 if (code < 0) 359 return (code); 360 } 361 /* Now, actually do the reset. */ 362 state->bytes_in_section = 0; 363 state->bits = 9; 364 state->section_end_code = (1 << state->bits) - 1; 365 state->free_ent = 257; 366 state->oldcode = -1; 367 return (next_code(self)); 368 } 369 370 if (code > state->free_ent) { 371 /* An invalid code is a fatal error. */ 372 archive_set_error(&(self->archive->archive), -1, 373 "Invalid compressed data"); 374 return (ARCHIVE_FATAL); 375 } 376 377 /* Special case for KwKwK string. */ 378 if (code >= state->free_ent) { 379 *state->stackp++ = state->finbyte; 380 code = state->oldcode; 381 } 382 383 /* Generate output characters in reverse order. */ 384 while (code >= 256) { 385 *state->stackp++ = state->suffix[code]; 386 code = state->prefix[code]; 387 } 388 *state->stackp++ = state->finbyte = code; 389 390 /* Generate the new entry. */ 391 code = state->free_ent; 392 if (code < state->maxcode && state->oldcode >= 0) { 393 state->prefix[code] = state->oldcode; 394 state->suffix[code] = state->finbyte; 395 ++state->free_ent; 396 } 397 if (state->free_ent > state->section_end_code) { 398 state->bits++; 399 state->bytes_in_section = 0; 400 if (state->bits == state->maxcode_bits) 401 state->section_end_code = state->maxcode; 402 else 403 state->section_end_code = (1 << state->bits) - 1; 404 } 405 406 /* Remember previous code. */ 407 state->oldcode = newcode; 408 return (ARCHIVE_OK); 409 } 410 411 /* 412 * Return next 'n' bits from stream. 413 * 414 * -1 indicates end of available data. 415 */ 416 static int 417 getbits(struct archive_read_filter *self, int n) 418 { 419 struct private_data *state = (struct private_data *)self->data; 420 int code; 421 ssize_t ret; 422 static const int mask[] = { 423 0x00, 0x01, 0x03, 0x07, 0x0f, 0x1f, 0x3f, 0x7f, 0xff, 424 0x1ff, 0x3ff, 0x7ff, 0xfff, 0x1fff, 0x3fff, 0x7fff, 0xffff 425 }; 426 427 while (state->bits_avail < n) { 428 if (state->avail_in <= 0) { 429 if (state->consume_unnotified) { 430 __archive_read_filter_consume(self->upstream, 431 state->consume_unnotified); 432 state->consume_unnotified = 0; 433 } 434 state->next_in 435 = __archive_read_filter_ahead(self->upstream, 436 1, &ret); 437 if (ret == 0) 438 return (-1); 439 if (ret < 0 || state->next_in == NULL) 440 return (ARCHIVE_FATAL); 441 state->consume_unnotified = state->avail_in = ret; 442 } 443 state->bit_buffer |= *state->next_in++ << state->bits_avail; 444 state->avail_in--; 445 state->bits_avail += 8; 446 state->bytes_in_section++; 447 } 448 449 code = state->bit_buffer; 450 state->bit_buffer >>= n; 451 state->bits_avail -= n; 452 453 return (code & mask[n]); 454 } 455