1 /*- 2 * Copyright (c) 2003-2007 Tim Kientzle 3 * Copyright (c) 2008 Joerg Sonnenberger 4 * Copyright (c) 2011-2012 Michihiro NAKAJIMA 5 * All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR 17 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 18 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 19 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT, 20 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 21 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 22 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 23 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 24 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 25 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 26 */ 27 28 #include "archive_platform.h" 29 __FBSDID("$FreeBSD$"); 30 31 #ifdef HAVE_SYS_STAT_H 32 #include <sys/stat.h> 33 #endif 34 #ifdef HAVE_ERRNO_H 35 #include <errno.h> 36 #endif 37 #ifdef HAVE_FCNTL_H 38 #include <fcntl.h> 39 #endif 40 #include <stddef.h> 41 /* #include <stdint.h> */ /* See archive_platform.h */ 42 #ifdef HAVE_STDLIB_H 43 #include <stdlib.h> 44 #endif 45 #ifdef HAVE_STRING_H 46 #include <string.h> 47 #endif 48 49 #include "archive.h" 50 #include "archive_entry.h" 51 #include "archive_private.h" 52 #include "archive_read_private.h" 53 #include "archive_string.h" 54 #include "archive_pack_dev.h" 55 56 #ifndef O_BINARY 57 #define O_BINARY 0 58 #endif 59 #ifndef O_CLOEXEC 60 #define O_CLOEXEC 0 61 #endif 62 63 #define MTREE_HAS_DEVICE 0x0001 64 #define MTREE_HAS_FFLAGS 0x0002 65 #define MTREE_HAS_GID 0x0004 66 #define MTREE_HAS_GNAME 0x0008 67 #define MTREE_HAS_MTIME 0x0010 68 #define MTREE_HAS_NLINK 0x0020 69 #define MTREE_HAS_PERM 0x0040 70 #define MTREE_HAS_SIZE 0x0080 71 #define MTREE_HAS_TYPE 0x0100 72 #define MTREE_HAS_UID 0x0200 73 #define MTREE_HAS_UNAME 0x0400 74 75 #define MTREE_HAS_OPTIONAL 0x0800 76 #define MTREE_HAS_NOCHANGE 0x1000 /* FreeBSD specific */ 77 78 #define MTREE_HASHTABLE_SIZE 1024 79 80 struct mtree_option { 81 struct mtree_option *next; 82 char *value; 83 }; 84 85 struct mtree_entry { 86 struct mtree_entry *next; 87 struct mtree_option *options; 88 char *name; 89 char full; 90 char used; 91 unsigned int name_hash; 92 struct mtree_entry *hashtable_next; 93 }; 94 95 struct mtree { 96 struct archive_string line; 97 size_t buffsize; 98 char *buff; 99 int64_t offset; 100 int fd; 101 int archive_format; 102 const char *archive_format_name; 103 struct mtree_entry *entries; 104 struct mtree_entry *this_entry; 105 struct mtree_entry *entry_hashtable[MTREE_HASHTABLE_SIZE]; 106 struct archive_string current_dir; 107 struct archive_string contents_name; 108 109 struct archive_entry_linkresolver *resolver; 110 111 int64_t cur_size; 112 char checkfs; 113 }; 114 115 static int bid_keycmp(const char *, const char *, ssize_t); 116 static int cleanup(struct archive_read *); 117 static int detect_form(struct archive_read *, int *); 118 static unsigned int hash(const char *); 119 static int mtree_bid(struct archive_read *, int); 120 static int parse_file(struct archive_read *, struct archive_entry *, 121 struct mtree *, struct mtree_entry *, int *); 122 static void parse_escapes(char *, struct mtree_entry *); 123 static int parse_line(struct archive_read *, struct archive_entry *, 124 struct mtree *, struct mtree_entry *, int *); 125 static int parse_keyword(struct archive_read *, struct mtree *, 126 struct archive_entry *, struct mtree_option *, int *); 127 static int read_data(struct archive_read *a, 128 const void **buff, size_t *size, int64_t *offset); 129 static ssize_t readline(struct archive_read *, struct mtree *, char **, ssize_t); 130 static int skip(struct archive_read *a); 131 static int read_header(struct archive_read *, 132 struct archive_entry *); 133 static int64_t mtree_atol10(char **); 134 static int64_t mtree_atol8(char **); 135 static int64_t mtree_atol(char **); 136 137 /* 138 * There's no standard for TIME_T_MAX/TIME_T_MIN. So we compute them 139 * here. TODO: Move this to configure time, but be careful 140 * about cross-compile environments. 141 */ 142 static int64_t 143 get_time_t_max(void) 144 { 145 #if defined(TIME_T_MAX) 146 return TIME_T_MAX; 147 #else 148 /* ISO C allows time_t to be a floating-point type, 149 but POSIX requires an integer type. The following 150 should work on any system that follows the POSIX 151 conventions. */ 152 if (((time_t)0) < ((time_t)-1)) { 153 /* Time_t is unsigned */ 154 return (~(time_t)0); 155 } else { 156 /* Time_t is signed. */ 157 /* Assume it's the same as int64_t or int32_t */ 158 if (sizeof(time_t) == sizeof(int64_t)) { 159 return (time_t)INT64_MAX; 160 } else { 161 return (time_t)INT32_MAX; 162 } 163 } 164 #endif 165 } 166 167 static int64_t 168 get_time_t_min(void) 169 { 170 #if defined(TIME_T_MIN) 171 return TIME_T_MIN; 172 #else 173 if (((time_t)0) < ((time_t)-1)) { 174 /* Time_t is unsigned */ 175 return (time_t)0; 176 } else { 177 /* Time_t is signed. */ 178 if (sizeof(time_t) == sizeof(int64_t)) { 179 return (time_t)INT64_MIN; 180 } else { 181 return (time_t)INT32_MIN; 182 } 183 } 184 #endif 185 } 186 187 static int 188 archive_read_format_mtree_options(struct archive_read *a, 189 const char *key, const char *val) 190 { 191 struct mtree *mtree; 192 193 mtree = (struct mtree *)(a->format->data); 194 if (strcmp(key, "checkfs") == 0) { 195 /* Allows to read information missing from the mtree from the file system */ 196 if (val == NULL || val[0] == 0) { 197 mtree->checkfs = 0; 198 } else { 199 mtree->checkfs = 1; 200 } 201 return (ARCHIVE_OK); 202 } 203 204 /* Note: The "warn" return is just to inform the options 205 * supervisor that we didn't handle it. It will generate 206 * a suitable error if no one used this option. */ 207 return (ARCHIVE_WARN); 208 } 209 210 static void 211 free_options(struct mtree_option *head) 212 { 213 struct mtree_option *next; 214 215 for (; head != NULL; head = next) { 216 next = head->next; 217 free(head->value); 218 free(head); 219 } 220 } 221 222 int 223 archive_read_support_format_mtree(struct archive *_a) 224 { 225 struct archive_read *a = (struct archive_read *)_a; 226 struct mtree *mtree; 227 int r; 228 229 archive_check_magic(_a, ARCHIVE_READ_MAGIC, 230 ARCHIVE_STATE_NEW, "archive_read_support_format_mtree"); 231 232 mtree = (struct mtree *)calloc(1, sizeof(*mtree)); 233 if (mtree == NULL) { 234 archive_set_error(&a->archive, ENOMEM, 235 "Can't allocate mtree data"); 236 return (ARCHIVE_FATAL); 237 } 238 mtree->fd = -1; 239 240 r = __archive_read_register_format(a, mtree, "mtree", 241 mtree_bid, archive_read_format_mtree_options, read_header, read_data, skip, NULL, cleanup, NULL, NULL); 242 243 if (r != ARCHIVE_OK) 244 free(mtree); 245 return (ARCHIVE_OK); 246 } 247 248 static int 249 cleanup(struct archive_read *a) 250 { 251 struct mtree *mtree; 252 struct mtree_entry *p, *q; 253 254 mtree = (struct mtree *)(a->format->data); 255 256 p = mtree->entries; 257 while (p != NULL) { 258 q = p->next; 259 free(p->name); 260 free_options(p->options); 261 free(p); 262 p = q; 263 } 264 archive_string_free(&mtree->line); 265 archive_string_free(&mtree->current_dir); 266 archive_string_free(&mtree->contents_name); 267 archive_entry_linkresolver_free(mtree->resolver); 268 269 free(mtree->buff); 270 free(mtree); 271 (a->format->data) = NULL; 272 return (ARCHIVE_OK); 273 } 274 275 static ssize_t 276 get_line_size(const char *b, ssize_t avail, ssize_t *nlsize) 277 { 278 ssize_t len; 279 280 len = 0; 281 while (len < avail) { 282 switch (*b) { 283 case '\0':/* Non-ascii character or control character. */ 284 if (nlsize != NULL) 285 *nlsize = 0; 286 return (-1); 287 case '\r': 288 if (avail-len > 1 && b[1] == '\n') { 289 if (nlsize != NULL) 290 *nlsize = 2; 291 return (len+2); 292 } 293 /* FALL THROUGH */ 294 case '\n': 295 if (nlsize != NULL) 296 *nlsize = 1; 297 return (len+1); 298 default: 299 b++; 300 len++; 301 break; 302 } 303 } 304 if (nlsize != NULL) 305 *nlsize = 0; 306 return (avail); 307 } 308 309 /* 310 * <---------------- ravail ---------------------> 311 * <-- diff ------> <--- avail -----------------> 312 * <---- len -----------> 313 * | Previous lines | line being parsed nl extra | 314 * ^ 315 * b 316 * 317 */ 318 static ssize_t 319 next_line(struct archive_read *a, 320 const char **b, ssize_t *avail, ssize_t *ravail, ssize_t *nl) 321 { 322 ssize_t len; 323 int quit; 324 325 quit = 0; 326 if (*avail == 0) { 327 *nl = 0; 328 len = 0; 329 } else 330 len = get_line_size(*b, *avail, nl); 331 /* 332 * Read bytes more while it does not reach the end of line. 333 */ 334 while (*nl == 0 && len == *avail && !quit) { 335 ssize_t diff = *ravail - *avail; 336 size_t nbytes_req = (*ravail+1023) & ~1023U; 337 ssize_t tested; 338 339 /* Increase reading bytes if it is not enough to at least 340 * new two lines. */ 341 if (nbytes_req < (size_t)*ravail + 160) 342 nbytes_req <<= 1; 343 344 *b = __archive_read_ahead(a, nbytes_req, avail); 345 if (*b == NULL) { 346 if (*ravail >= *avail) 347 return (0); 348 /* Reading bytes reaches the end of file. */ 349 *b = __archive_read_ahead(a, *avail, avail); 350 quit = 1; 351 } 352 *ravail = *avail; 353 *b += diff; 354 *avail -= diff; 355 tested = len;/* Skip some bytes we already determinated. */ 356 len = get_line_size(*b + len, *avail - len, nl); 357 if (len >= 0) 358 len += tested; 359 } 360 return (len); 361 } 362 363 /* 364 * Compare characters with a mtree keyword. 365 * Returns the length of a mtree keyword if matched. 366 * Returns 0 if not matched. 367 */ 368 static int 369 bid_keycmp(const char *p, const char *key, ssize_t len) 370 { 371 int match_len = 0; 372 373 while (len > 0 && *p && *key) { 374 if (*p == *key) { 375 --len; 376 ++p; 377 ++key; 378 ++match_len; 379 continue; 380 } 381 return (0);/* Not match */ 382 } 383 if (*key != '\0') 384 return (0);/* Not match */ 385 386 /* A following character should be specified characters */ 387 if (p[0] == '=' || p[0] == ' ' || p[0] == '\t' || 388 p[0] == '\n' || p[0] == '\r' || 389 (p[0] == '\\' && (p[1] == '\n' || p[1] == '\r'))) 390 return (match_len); 391 return (0);/* Not match */ 392 } 393 394 /* 395 * Test whether the characters 'p' has is mtree keyword. 396 * Returns the length of a detected keyword. 397 * Returns 0 if any keywords were not found. 398 */ 399 static int 400 bid_keyword(const char *p, ssize_t len) 401 { 402 static const char * const keys_c[] = { 403 "content", "contents", "cksum", NULL 404 }; 405 static const char * const keys_df[] = { 406 "device", "flags", NULL 407 }; 408 static const char * const keys_g[] = { 409 "gid", "gname", NULL 410 }; 411 static const char * const keys_il[] = { 412 "ignore", "inode", "link", NULL 413 }; 414 static const char * const keys_m[] = { 415 "md5", "md5digest", "mode", NULL 416 }; 417 static const char * const keys_no[] = { 418 "nlink", "nochange", "optional", NULL 419 }; 420 static const char * const keys_r[] = { 421 "resdevice", "rmd160", "rmd160digest", NULL 422 }; 423 static const char * const keys_s[] = { 424 "sha1", "sha1digest", 425 "sha256", "sha256digest", 426 "sha384", "sha384digest", 427 "sha512", "sha512digest", 428 "size", NULL 429 }; 430 static const char * const keys_t[] = { 431 "tags", "time", "type", NULL 432 }; 433 static const char * const keys_u[] = { 434 "uid", "uname", NULL 435 }; 436 const char * const *keys; 437 int i; 438 439 switch (*p) { 440 case 'c': keys = keys_c; break; 441 case 'd': case 'f': keys = keys_df; break; 442 case 'g': keys = keys_g; break; 443 case 'i': case 'l': keys = keys_il; break; 444 case 'm': keys = keys_m; break; 445 case 'n': case 'o': keys = keys_no; break; 446 case 'r': keys = keys_r; break; 447 case 's': keys = keys_s; break; 448 case 't': keys = keys_t; break; 449 case 'u': keys = keys_u; break; 450 default: return (0);/* Unknown key */ 451 } 452 453 for (i = 0; keys[i] != NULL; i++) { 454 int l = bid_keycmp(p, keys[i], len); 455 if (l > 0) 456 return (l); 457 } 458 return (0);/* Unknown key */ 459 } 460 461 /* 462 * Test whether there is a set of mtree keywords. 463 * Returns the number of keyword. 464 * Returns -1 if we got incorrect sequence. 465 * This function expects a set of "<space characters>keyword=value". 466 * When "unset" is specified, expects a set of "<space characters>keyword". 467 */ 468 static int 469 bid_keyword_list(const char *p, ssize_t len, int unset, int last_is_path) 470 { 471 int l; 472 int keycnt = 0; 473 474 while (len > 0 && *p) { 475 int blank = 0; 476 477 /* Test whether there are blank characters in the line. */ 478 while (len >0 && (*p == ' ' || *p == '\t')) { 479 ++p; 480 --len; 481 blank = 1; 482 } 483 if (*p == '\n' || *p == '\r') 484 break; 485 if (p[0] == '\\' && (p[1] == '\n' || p[1] == '\r')) 486 break; 487 if (!blank && !last_is_path) /* No blank character. */ 488 return (-1); 489 if (last_is_path && len == 0) 490 return (keycnt); 491 492 if (unset) { 493 l = bid_keycmp(p, "all", len); 494 if (l > 0) 495 return (1); 496 } 497 /* Test whether there is a correct key in the line. */ 498 l = bid_keyword(p, len); 499 if (l == 0) 500 return (-1);/* Unknown keyword was found. */ 501 p += l; 502 len -= l; 503 keycnt++; 504 505 /* Skip value */ 506 if (*p == '=') { 507 int value = 0; 508 ++p; 509 --len; 510 while (len > 0 && *p != ' ' && *p != '\t') { 511 ++p; 512 --len; 513 value = 1; 514 } 515 /* A keyword should have a its value unless 516 * "/unset" operation. */ 517 if (!unset && value == 0) 518 return (-1); 519 } 520 } 521 return (keycnt); 522 } 523 524 static int 525 bid_entry(const char *p, ssize_t len, ssize_t nl, int *last_is_path) 526 { 527 int f = 0; 528 static const unsigned char safe_char[256] = { 529 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 00 - 0F */ 530 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 10 - 1F */ 531 /* !"$%&'()*+,-./ EXCLUSION:( )(#) */ 532 0, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 20 - 2F */ 533 /* 0123456789:;<>? EXCLUSION:(=) */ 534 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, /* 30 - 3F */ 535 /* @ABCDEFGHIJKLMNO */ 536 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 40 - 4F */ 537 /* PQRSTUVWXYZ[\]^_ */ 538 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 50 - 5F */ 539 /* `abcdefghijklmno */ 540 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 60 - 6F */ 541 /* pqrstuvwxyz{|}~ */ 542 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, /* 70 - 7F */ 543 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 80 - 8F */ 544 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 90 - 9F */ 545 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* A0 - AF */ 546 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* B0 - BF */ 547 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* C0 - CF */ 548 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* D0 - DF */ 549 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* E0 - EF */ 550 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* F0 - FF */ 551 }; 552 ssize_t ll; 553 const char *pp = p; 554 const char * const pp_end = pp + len; 555 556 *last_is_path = 0; 557 /* 558 * Skip the path-name which is quoted. 559 */ 560 for (;pp < pp_end; ++pp) { 561 if (!safe_char[*(const unsigned char *)pp]) { 562 if (*pp != ' ' && *pp != '\t' && *pp != '\r' 563 && *pp != '\n') 564 f = 0; 565 break; 566 } 567 f = 1; 568 } 569 ll = pp_end - pp; 570 571 /* If a path-name was not found at the first, try to check 572 * a mtree format(a.k.a form D) ``NetBSD's mtree -D'' creates, 573 * which places the path-name at the last. */ 574 if (f == 0) { 575 const char *pb = p + len - nl; 576 int name_len = 0; 577 int slash; 578 579 /* The form D accepts only a single line for an entry. */ 580 if (pb-2 >= p && 581 pb[-1] == '\\' && (pb[-2] == ' ' || pb[-2] == '\t')) 582 return (-1); 583 if (pb-1 >= p && pb[-1] == '\\') 584 return (-1); 585 586 slash = 0; 587 while (p <= --pb && *pb != ' ' && *pb != '\t') { 588 if (!safe_char[*(const unsigned char *)pb]) 589 return (-1); 590 name_len++; 591 /* The pathname should have a slash in this 592 * format. */ 593 if (*pb == '/') 594 slash = 1; 595 } 596 if (name_len == 0 || slash == 0) 597 return (-1); 598 /* If '/' is placed at the first in this field, this is not 599 * a valid filename. */ 600 if (pb[1] == '/') 601 return (-1); 602 ll = len - nl - name_len; 603 pp = p; 604 *last_is_path = 1; 605 } 606 607 return (bid_keyword_list(pp, ll, 0, *last_is_path)); 608 } 609 610 #define MAX_BID_ENTRY 3 611 612 static int 613 mtree_bid(struct archive_read *a, int best_bid) 614 { 615 const char *signature = "#mtree"; 616 const char *p; 617 618 (void)best_bid; /* UNUSED */ 619 620 /* Now let's look at the actual header and see if it matches. */ 621 p = __archive_read_ahead(a, strlen(signature), NULL); 622 if (p == NULL) 623 return (-1); 624 625 if (memcmp(p, signature, strlen(signature)) == 0) 626 return (8 * (int)strlen(signature)); 627 628 /* 629 * There is not a mtree signature. Let's try to detect mtree format. 630 */ 631 return (detect_form(a, NULL)); 632 } 633 634 static int 635 detect_form(struct archive_read *a, int *is_form_d) 636 { 637 const char *p; 638 ssize_t avail, ravail; 639 ssize_t detected_bytes = 0, len, nl; 640 int entry_cnt = 0, multiline = 0; 641 int form_D = 0;/* The archive is generated by `NetBSD mtree -D' 642 * (In this source we call it `form D') . */ 643 644 if (is_form_d != NULL) 645 *is_form_d = 0; 646 p = __archive_read_ahead(a, 1, &avail); 647 if (p == NULL) 648 return (-1); 649 ravail = avail; 650 for (;;) { 651 len = next_line(a, &p, &avail, &ravail, &nl); 652 /* The terminal character of the line should be 653 * a new line character, '\r\n' or '\n'. */ 654 if (len <= 0 || nl == 0) 655 break; 656 if (!multiline) { 657 /* Leading whitespace is never significant, 658 * ignore it. */ 659 while (len > 0 && (*p == ' ' || *p == '\t')) { 660 ++p; 661 --avail; 662 --len; 663 } 664 /* Skip comment or empty line. */ 665 if (p[0] == '#' || p[0] == '\n' || p[0] == '\r') { 666 p += len; 667 avail -= len; 668 continue; 669 } 670 } else { 671 /* A continuance line; the terminal 672 * character of previous line was '\' character. */ 673 if (bid_keyword_list(p, len, 0, 0) <= 0) 674 break; 675 if (multiline == 1) 676 detected_bytes += len; 677 if (p[len-nl-1] != '\\') { 678 if (multiline == 1 && 679 ++entry_cnt >= MAX_BID_ENTRY) 680 break; 681 multiline = 0; 682 } 683 p += len; 684 avail -= len; 685 continue; 686 } 687 if (p[0] != '/') { 688 int last_is_path, keywords; 689 690 keywords = bid_entry(p, len, nl, &last_is_path); 691 if (keywords >= 0) { 692 detected_bytes += len; 693 if (form_D == 0) { 694 if (last_is_path) 695 form_D = 1; 696 else if (keywords > 0) 697 /* This line is not `form D'. */ 698 form_D = -1; 699 } else if (form_D == 1) { 700 if (!last_is_path && keywords > 0) 701 /* This this is not `form D' 702 * and We cannot accept mixed 703 * format. */ 704 break; 705 } 706 if (!last_is_path && p[len-nl-1] == '\\') 707 /* This line continues. */ 708 multiline = 1; 709 else { 710 /* We've got plenty of correct lines 711 * to assume that this file is a mtree 712 * format. */ 713 if (++entry_cnt >= MAX_BID_ENTRY) 714 break; 715 } 716 } else 717 break; 718 } else if (len > 4 && strncmp(p, "/set", 4) == 0) { 719 if (bid_keyword_list(p+4, len-4, 0, 0) <= 0) 720 break; 721 /* This line continues. */ 722 if (p[len-nl-1] == '\\') 723 multiline = 2; 724 } else if (len > 6 && strncmp(p, "/unset", 6) == 0) { 725 if (bid_keyword_list(p+6, len-6, 1, 0) <= 0) 726 break; 727 /* This line continues. */ 728 if (p[len-nl-1] == '\\') 729 multiline = 2; 730 } else 731 break; 732 733 /* Test next line. */ 734 p += len; 735 avail -= len; 736 } 737 if (entry_cnt >= MAX_BID_ENTRY || (entry_cnt > 0 && len == 0)) { 738 if (is_form_d != NULL) { 739 if (form_D == 1) 740 *is_form_d = 1; 741 } 742 return (32); 743 } 744 745 return (0); 746 } 747 748 /* 749 * The extended mtree format permits multiple lines specifying 750 * attributes for each file. For those entries, only the last line 751 * is actually used. Practically speaking, that means we have 752 * to read the entire mtree file into memory up front. 753 * 754 * The parsing is done in two steps. First, it is decided if a line 755 * changes the global defaults and if it is, processed accordingly. 756 * Otherwise, the options of the line are merged with the current 757 * global options. 758 */ 759 static int 760 add_option(struct archive_read *a, struct mtree_option **global, 761 const char *value, size_t len) 762 { 763 struct mtree_option *opt; 764 765 if ((opt = malloc(sizeof(*opt))) == NULL) { 766 archive_set_error(&a->archive, errno, "Can't allocate memory"); 767 return (ARCHIVE_FATAL); 768 } 769 if ((opt->value = malloc(len + 1)) == NULL) { 770 free(opt); 771 archive_set_error(&a->archive, errno, "Can't allocate memory"); 772 return (ARCHIVE_FATAL); 773 } 774 memcpy(opt->value, value, len); 775 opt->value[len] = '\0'; 776 opt->next = *global; 777 *global = opt; 778 return (ARCHIVE_OK); 779 } 780 781 static void 782 remove_option(struct mtree_option **global, const char *value, size_t len) 783 { 784 struct mtree_option *iter, *last; 785 786 last = NULL; 787 for (iter = *global; iter != NULL; last = iter, iter = iter->next) { 788 if (strncmp(iter->value, value, len) == 0 && 789 (iter->value[len] == '\0' || 790 iter->value[len] == '=')) 791 break; 792 } 793 if (iter == NULL) 794 return; 795 if (last == NULL) 796 *global = iter->next; 797 else 798 last->next = iter->next; 799 800 free(iter->value); 801 free(iter); 802 } 803 804 static int 805 process_global_set(struct archive_read *a, 806 struct mtree_option **global, const char *line) 807 { 808 const char *next, *eq; 809 size_t len; 810 int r; 811 812 line += 4; 813 for (;;) { 814 next = line + strspn(line, " \t\r\n"); 815 if (*next == '\0') 816 return (ARCHIVE_OK); 817 line = next; 818 next = line + strcspn(line, " \t\r\n"); 819 eq = strchr(line, '='); 820 if (eq > next) 821 len = next - line; 822 else 823 len = eq - line; 824 825 remove_option(global, line, len); 826 r = add_option(a, global, line, next - line); 827 if (r != ARCHIVE_OK) 828 return (r); 829 line = next; 830 } 831 } 832 833 static int 834 process_global_unset(struct archive_read *a, 835 struct mtree_option **global, const char *line) 836 { 837 const char *next; 838 size_t len; 839 840 line += 6; 841 if (strchr(line, '=') != NULL) { 842 archive_set_error(&a->archive, ARCHIVE_ERRNO_MISC, 843 "/unset shall not contain `='"); 844 return ARCHIVE_FATAL; 845 } 846 847 for (;;) { 848 next = line + strspn(line, " \t\r\n"); 849 if (*next == '\0') 850 return (ARCHIVE_OK); 851 line = next; 852 len = strcspn(line, " \t\r\n"); 853 854 if (len == 3 && strncmp(line, "all", 3) == 0) { 855 free_options(*global); 856 *global = NULL; 857 } else { 858 remove_option(global, line, len); 859 } 860 861 line += len; 862 } 863 } 864 865 static int 866 process_add_entry(struct archive_read *a, struct mtree *mtree, 867 struct mtree_option **global, const char *line, ssize_t line_len, 868 struct mtree_entry **last_entry, int is_form_d) 869 { 870 struct mtree_entry *entry, *ht_iter; 871 struct mtree_option *iter; 872 const char *next, *eq, *name, *end; 873 size_t name_len, len; 874 int r, i; 875 unsigned int ht_idx; 876 877 if ((entry = malloc(sizeof(*entry))) == NULL) { 878 archive_set_error(&a->archive, errno, "Can't allocate memory"); 879 return (ARCHIVE_FATAL); 880 } 881 entry->next = NULL; 882 entry->options = NULL; 883 entry->name = NULL; 884 entry->used = 0; 885 entry->full = 0; 886 entry->name_hash = 0; 887 entry->hashtable_next = NULL; 888 889 /* Add this entry to list. */ 890 if (*last_entry == NULL) 891 mtree->entries = entry; 892 else 893 (*last_entry)->next = entry; 894 *last_entry = entry; 895 896 if (is_form_d) { 897 /* Filename is last item on line. */ 898 /* Adjust line_len to trim trailing whitespace */ 899 while (line_len > 0) { 900 char last_character = line[line_len - 1]; 901 if (last_character == '\r' 902 || last_character == '\n' 903 || last_character == '\t' 904 || last_character == ' ') { 905 line_len--; 906 } else { 907 break; 908 } 909 } 910 /* Name starts after the last whitespace separator */ 911 name = line; 912 for (i = 0; i < line_len; i++) { 913 if (line[i] == '\r' 914 || line[i] == '\n' 915 || line[i] == '\t' 916 || line[i] == ' ') { 917 name = line + i + 1; 918 } 919 } 920 name_len = line + line_len - name; 921 end = name; 922 } else { 923 /* Filename is first item on line */ 924 name_len = strcspn(line, " \t\r\n"); 925 name = line; 926 line += name_len; 927 end = line + line_len; 928 } 929 /* name/name_len is the name within the line. */ 930 /* line..end brackets the entire line except the name */ 931 932 if ((entry->name = malloc(name_len + 1)) == NULL) { 933 archive_set_error(&a->archive, errno, "Can't allocate memory"); 934 return (ARCHIVE_FATAL); 935 } 936 937 memcpy(entry->name, name, name_len); 938 entry->name[name_len] = '\0'; 939 parse_escapes(entry->name, entry); 940 entry->name_hash = hash(entry->name); 941 942 ht_idx = entry->name_hash % MTREE_HASHTABLE_SIZE; 943 if ((ht_iter = mtree->entry_hashtable[ht_idx]) != NULL) { 944 while (ht_iter->hashtable_next) 945 ht_iter = ht_iter->hashtable_next; 946 ht_iter->hashtable_next = entry; 947 } else { 948 mtree->entry_hashtable[ht_idx] = entry; 949 } 950 951 for (iter = *global; iter != NULL; iter = iter->next) { 952 r = add_option(a, &entry->options, iter->value, 953 strlen(iter->value)); 954 if (r != ARCHIVE_OK) 955 return (r); 956 } 957 958 for (;;) { 959 next = line + strspn(line, " \t\r\n"); 960 if (*next == '\0') 961 return (ARCHIVE_OK); 962 if (next >= end) 963 return (ARCHIVE_OK); 964 line = next; 965 next = line + strcspn(line, " \t\r\n"); 966 eq = strchr(line, '='); 967 if (eq == NULL || eq > next) 968 len = next - line; 969 else 970 len = eq - line; 971 972 remove_option(&entry->options, line, len); 973 r = add_option(a, &entry->options, line, next - line); 974 if (r != ARCHIVE_OK) 975 return (r); 976 line = next; 977 } 978 } 979 980 static int 981 read_mtree(struct archive_read *a, struct mtree *mtree) 982 { 983 ssize_t len; 984 uintmax_t counter; 985 char *p; 986 struct mtree_option *global; 987 struct mtree_entry *last_entry; 988 int r, is_form_d; 989 990 mtree->archive_format = ARCHIVE_FORMAT_MTREE; 991 mtree->archive_format_name = "mtree"; 992 993 global = NULL; 994 last_entry = NULL; 995 996 (void)detect_form(a, &is_form_d); 997 998 for (counter = 1; ; ++counter) { 999 len = readline(a, mtree, &p, 65536); 1000 if (len == 0) { 1001 mtree->this_entry = mtree->entries; 1002 free_options(global); 1003 return (ARCHIVE_OK); 1004 } 1005 if (len < 0) { 1006 free_options(global); 1007 return ((int)len); 1008 } 1009 /* Leading whitespace is never significant, ignore it. */ 1010 while (*p == ' ' || *p == '\t') { 1011 ++p; 1012 --len; 1013 } 1014 /* Skip content lines and blank lines. */ 1015 if (*p == '#') 1016 continue; 1017 if (*p == '\r' || *p == '\n' || *p == '\0') 1018 continue; 1019 if (*p != '/') { 1020 r = process_add_entry(a, mtree, &global, p, len, 1021 &last_entry, is_form_d); 1022 } else if (len > 4 && strncmp(p, "/set", 4) == 0) { 1023 if (p[4] != ' ' && p[4] != '\t') 1024 break; 1025 r = process_global_set(a, &global, p); 1026 } else if (len > 6 && strncmp(p, "/unset", 6) == 0) { 1027 if (p[6] != ' ' && p[6] != '\t') 1028 break; 1029 r = process_global_unset(a, &global, p); 1030 } else 1031 break; 1032 1033 if (r != ARCHIVE_OK) { 1034 free_options(global); 1035 return r; 1036 } 1037 } 1038 1039 archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT, 1040 "Can't parse line %ju", counter); 1041 free_options(global); 1042 return (ARCHIVE_FATAL); 1043 } 1044 1045 /* 1046 * Read in the entire mtree file into memory on the first request. 1047 * Then use the next unused file to satisfy each header request. 1048 */ 1049 static int 1050 read_header(struct archive_read *a, struct archive_entry *entry) 1051 { 1052 struct mtree *mtree; 1053 char *p; 1054 int r, use_next; 1055 1056 mtree = (struct mtree *)(a->format->data); 1057 1058 if (mtree->fd >= 0) { 1059 close(mtree->fd); 1060 mtree->fd = -1; 1061 } 1062 1063 if (mtree->entries == NULL) { 1064 mtree->resolver = archive_entry_linkresolver_new(); 1065 if (mtree->resolver == NULL) 1066 return ARCHIVE_FATAL; 1067 archive_entry_linkresolver_set_strategy(mtree->resolver, 1068 ARCHIVE_FORMAT_MTREE); 1069 r = read_mtree(a, mtree); 1070 if (r != ARCHIVE_OK) 1071 return (r); 1072 } 1073 1074 a->archive.archive_format = mtree->archive_format; 1075 a->archive.archive_format_name = mtree->archive_format_name; 1076 1077 for (;;) { 1078 if (mtree->this_entry == NULL) 1079 return (ARCHIVE_EOF); 1080 if (strcmp(mtree->this_entry->name, "..") == 0) { 1081 mtree->this_entry->used = 1; 1082 if (archive_strlen(&mtree->current_dir) > 0) { 1083 /* Roll back current path. */ 1084 p = mtree->current_dir.s 1085 + mtree->current_dir.length - 1; 1086 while (p >= mtree->current_dir.s && *p != '/') 1087 --p; 1088 if (p >= mtree->current_dir.s) 1089 --p; 1090 mtree->current_dir.length 1091 = p - mtree->current_dir.s + 1; 1092 } 1093 } 1094 if (!mtree->this_entry->used) { 1095 use_next = 0; 1096 r = parse_file(a, entry, mtree, mtree->this_entry, 1097 &use_next); 1098 if (use_next == 0) 1099 return (r); 1100 } 1101 mtree->this_entry = mtree->this_entry->next; 1102 } 1103 } 1104 1105 /* 1106 * A single file can have multiple lines contribute specifications. 1107 * Parse as many lines as necessary, then pull additional information 1108 * from a backing file on disk as necessary. 1109 */ 1110 static int 1111 parse_file(struct archive_read *a, struct archive_entry *entry, 1112 struct mtree *mtree, struct mtree_entry *mentry, int *use_next) 1113 { 1114 const char *path; 1115 struct stat st_storage, *st; 1116 struct mtree_entry *mp; 1117 struct archive_entry *sparse_entry; 1118 int r = ARCHIVE_OK, r1, parsed_kws; 1119 1120 mentry->used = 1; 1121 1122 /* Initialize reasonable defaults. */ 1123 archive_entry_set_filetype(entry, AE_IFREG); 1124 archive_entry_set_size(entry, 0); 1125 archive_string_empty(&mtree->contents_name); 1126 1127 /* Parse options from this line. */ 1128 parsed_kws = 0; 1129 r = parse_line(a, entry, mtree, mentry, &parsed_kws); 1130 1131 if (mentry->full) { 1132 archive_entry_copy_pathname(entry, mentry->name); 1133 /* 1134 * "Full" entries are allowed to have multiple lines 1135 * and those lines aren't required to be adjacent. We 1136 * don't support multiple lines for "relative" entries 1137 * nor do we make any attempt to merge data from 1138 * separate "relative" and "full" entries. (Merging 1139 * "relative" and "full" entries would require dealing 1140 * with pathname canonicalization, which is a very 1141 * tricky subject.) 1142 */ 1143 for (mp = mentry->hashtable_next; mp != NULL; mp = mp->hashtable_next) { 1144 if (mp->full && !mp->used 1145 && mentry->name_hash == mp->name_hash 1146 && strcmp(mentry->name, mp->name) == 0) { 1147 /* Later lines override earlier ones. */ 1148 mp->used = 1; 1149 r1 = parse_line(a, entry, mtree, mp, 1150 &parsed_kws); 1151 if (r1 < r) 1152 r = r1; 1153 } 1154 } 1155 } else { 1156 /* 1157 * Relative entries require us to construct 1158 * the full path and possibly update the 1159 * current directory. 1160 */ 1161 size_t n = archive_strlen(&mtree->current_dir); 1162 if (n > 0) 1163 archive_strcat(&mtree->current_dir, "/"); 1164 archive_strcat(&mtree->current_dir, mentry->name); 1165 archive_entry_copy_pathname(entry, mtree->current_dir.s); 1166 if (archive_entry_filetype(entry) != AE_IFDIR) 1167 mtree->current_dir.length = n; 1168 } 1169 1170 if (mtree->checkfs) { 1171 /* 1172 * Try to open and stat the file to get the real size 1173 * and other file info. It would be nice to avoid 1174 * this here so that getting a listing of an mtree 1175 * wouldn't require opening every referenced contents 1176 * file. But then we wouldn't know the actual 1177 * contents size, so I don't see a really viable way 1178 * around this. (Also, we may want to someday pull 1179 * other unspecified info from the contents file on 1180 * disk.) 1181 */ 1182 mtree->fd = -1; 1183 if (archive_strlen(&mtree->contents_name) > 0) 1184 path = mtree->contents_name.s; 1185 else 1186 path = archive_entry_pathname(entry); 1187 1188 if (archive_entry_filetype(entry) == AE_IFREG || 1189 archive_entry_filetype(entry) == AE_IFDIR) { 1190 mtree->fd = open(path, O_RDONLY | O_BINARY | O_CLOEXEC); 1191 __archive_ensure_cloexec_flag(mtree->fd); 1192 if (mtree->fd == -1 && 1193 (errno != ENOENT || 1194 archive_strlen(&mtree->contents_name) > 0)) { 1195 archive_set_error(&a->archive, errno, 1196 "Can't open %s", path); 1197 r = ARCHIVE_WARN; 1198 } 1199 } 1200 1201 st = &st_storage; 1202 if (mtree->fd >= 0) { 1203 if (fstat(mtree->fd, st) == -1) { 1204 archive_set_error(&a->archive, errno, 1205 "Could not fstat %s", path); 1206 r = ARCHIVE_WARN; 1207 /* If we can't stat it, don't keep it open. */ 1208 close(mtree->fd); 1209 mtree->fd = -1; 1210 st = NULL; 1211 } 1212 } else if (lstat(path, st) == -1) { 1213 st = NULL; 1214 } 1215 1216 /* 1217 * Check for a mismatch between the type in the specification 1218 * and the type of the contents object on disk. 1219 */ 1220 if (st != NULL) { 1221 if (((st->st_mode & S_IFMT) == S_IFREG && 1222 archive_entry_filetype(entry) == AE_IFREG) 1223 #ifdef S_IFLNK 1224 ||((st->st_mode & S_IFMT) == S_IFLNK && 1225 archive_entry_filetype(entry) == AE_IFLNK) 1226 #endif 1227 #ifdef S_IFSOCK 1228 ||((st->st_mode & S_IFSOCK) == S_IFSOCK && 1229 archive_entry_filetype(entry) == AE_IFSOCK) 1230 #endif 1231 #ifdef S_IFCHR 1232 ||((st->st_mode & S_IFMT) == S_IFCHR && 1233 archive_entry_filetype(entry) == AE_IFCHR) 1234 #endif 1235 #ifdef S_IFBLK 1236 ||((st->st_mode & S_IFMT) == S_IFBLK && 1237 archive_entry_filetype(entry) == AE_IFBLK) 1238 #endif 1239 ||((st->st_mode & S_IFMT) == S_IFDIR && 1240 archive_entry_filetype(entry) == AE_IFDIR) 1241 #ifdef S_IFIFO 1242 ||((st->st_mode & S_IFMT) == S_IFIFO && 1243 archive_entry_filetype(entry) == AE_IFIFO) 1244 #endif 1245 ) { 1246 /* Types match. */ 1247 } else { 1248 /* Types don't match; bail out gracefully. */ 1249 if (mtree->fd >= 0) 1250 close(mtree->fd); 1251 mtree->fd = -1; 1252 if (parsed_kws & MTREE_HAS_OPTIONAL) { 1253 /* It's not an error for an optional 1254 * entry to not match disk. */ 1255 *use_next = 1; 1256 } else if (r == ARCHIVE_OK) { 1257 archive_set_error(&a->archive, 1258 ARCHIVE_ERRNO_MISC, 1259 "mtree specification has different" 1260 " type for %s", 1261 archive_entry_pathname(entry)); 1262 r = ARCHIVE_WARN; 1263 } 1264 return (r); 1265 } 1266 } 1267 1268 /* 1269 * If there is a contents file on disk, pick some of the 1270 * metadata from that file. For most of these, we only 1271 * set it from the contents if it wasn't already parsed 1272 * from the specification. 1273 */ 1274 if (st != NULL) { 1275 if (((parsed_kws & MTREE_HAS_DEVICE) == 0 || 1276 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) && 1277 (archive_entry_filetype(entry) == AE_IFCHR || 1278 archive_entry_filetype(entry) == AE_IFBLK)) 1279 archive_entry_set_rdev(entry, st->st_rdev); 1280 if ((parsed_kws & (MTREE_HAS_GID | MTREE_HAS_GNAME)) 1281 == 0 || 1282 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) 1283 archive_entry_set_gid(entry, st->st_gid); 1284 if ((parsed_kws & (MTREE_HAS_UID | MTREE_HAS_UNAME)) 1285 == 0 || 1286 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) 1287 archive_entry_set_uid(entry, st->st_uid); 1288 if ((parsed_kws & MTREE_HAS_MTIME) == 0 || 1289 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) { 1290 #if HAVE_STRUCT_STAT_ST_MTIMESPEC_TV_NSEC 1291 archive_entry_set_mtime(entry, st->st_mtime, 1292 st->st_mtimespec.tv_nsec); 1293 #elif HAVE_STRUCT_STAT_ST_MTIM_TV_NSEC 1294 archive_entry_set_mtime(entry, st->st_mtime, 1295 st->st_mtim.tv_nsec); 1296 #elif HAVE_STRUCT_STAT_ST_MTIME_N 1297 archive_entry_set_mtime(entry, st->st_mtime, 1298 st->st_mtime_n); 1299 #elif HAVE_STRUCT_STAT_ST_UMTIME 1300 archive_entry_set_mtime(entry, st->st_mtime, 1301 st->st_umtime*1000); 1302 #elif HAVE_STRUCT_STAT_ST_MTIME_USEC 1303 archive_entry_set_mtime(entry, st->st_mtime, 1304 st->st_mtime_usec*1000); 1305 #else 1306 archive_entry_set_mtime(entry, st->st_mtime, 0); 1307 #endif 1308 } 1309 if ((parsed_kws & MTREE_HAS_NLINK) == 0 || 1310 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) 1311 archive_entry_set_nlink(entry, st->st_nlink); 1312 if ((parsed_kws & MTREE_HAS_PERM) == 0 || 1313 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) 1314 archive_entry_set_perm(entry, st->st_mode); 1315 if ((parsed_kws & MTREE_HAS_SIZE) == 0 || 1316 (parsed_kws & MTREE_HAS_NOCHANGE) != 0) 1317 archive_entry_set_size(entry, st->st_size); 1318 archive_entry_set_ino(entry, st->st_ino); 1319 archive_entry_set_dev(entry, st->st_dev); 1320 1321 archive_entry_linkify(mtree->resolver, &entry, 1322 &sparse_entry); 1323 } else if (parsed_kws & MTREE_HAS_OPTIONAL) { 1324 /* 1325 * Couldn't open the entry, stat it or the on-disk type 1326 * didn't match. If this entry is optional, just 1327 * ignore it and read the next header entry. 1328 */ 1329 *use_next = 1; 1330 return ARCHIVE_OK; 1331 } 1332 } 1333 1334 mtree->cur_size = archive_entry_size(entry); 1335 mtree->offset = 0; 1336 1337 return r; 1338 } 1339 1340 /* 1341 * Each line contains a sequence of keywords. 1342 */ 1343 static int 1344 parse_line(struct archive_read *a, struct archive_entry *entry, 1345 struct mtree *mtree, struct mtree_entry *mp, int *parsed_kws) 1346 { 1347 struct mtree_option *iter; 1348 int r = ARCHIVE_OK, r1; 1349 1350 for (iter = mp->options; iter != NULL; iter = iter->next) { 1351 r1 = parse_keyword(a, mtree, entry, iter, parsed_kws); 1352 if (r1 < r) 1353 r = r1; 1354 } 1355 if (r == ARCHIVE_OK && (*parsed_kws & MTREE_HAS_TYPE) == 0) { 1356 archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT, 1357 "Missing type keyword in mtree specification"); 1358 return (ARCHIVE_WARN); 1359 } 1360 return (r); 1361 } 1362 1363 /* 1364 * Device entries have one of the following forms: 1365 * - raw dev_t 1366 * - format,major,minor[,subdevice] 1367 * When parsing succeeded, `pdev' will contain the appropriate dev_t value. 1368 */ 1369 1370 /* strsep() is not in C90, but strcspn() is. */ 1371 /* Taken from http://unixpapa.com/incnote/string.html */ 1372 static char * 1373 la_strsep(char **sp, const char *sep) 1374 { 1375 char *p, *s; 1376 if (sp == NULL || *sp == NULL || **sp == '\0') 1377 return(NULL); 1378 s = *sp; 1379 p = s + strcspn(s, sep); 1380 if (*p != '\0') 1381 *p++ = '\0'; 1382 *sp = p; 1383 return(s); 1384 } 1385 1386 static int 1387 parse_device(dev_t *pdev, struct archive *a, char *val) 1388 { 1389 #define MAX_PACK_ARGS 3 1390 unsigned long numbers[MAX_PACK_ARGS]; 1391 char *p, *dev; 1392 int argc; 1393 pack_t *pack; 1394 dev_t result; 1395 const char *error = NULL; 1396 1397 memset(pdev, 0, sizeof(*pdev)); 1398 if ((dev = strchr(val, ',')) != NULL) { 1399 /* 1400 * Device's major/minor are given in a specified format. 1401 * Decode and pack it accordingly. 1402 */ 1403 *dev++ = '\0'; 1404 if ((pack = pack_find(val)) == NULL) { 1405 archive_set_error(a, ARCHIVE_ERRNO_FILE_FORMAT, 1406 "Unknown format `%s'", val); 1407 return ARCHIVE_WARN; 1408 } 1409 argc = 0; 1410 while ((p = la_strsep(&dev, ",")) != NULL) { 1411 if (*p == '\0') { 1412 archive_set_error(a, ARCHIVE_ERRNO_FILE_FORMAT, 1413 "Missing number"); 1414 return ARCHIVE_WARN; 1415 } 1416 if (argc >= MAX_PACK_ARGS) { 1417 archive_set_error(a, ARCHIVE_ERRNO_FILE_FORMAT, 1418 "Too many arguments"); 1419 return ARCHIVE_WARN; 1420 } 1421 numbers[argc++] = (unsigned long)mtree_atol(&p); 1422 } 1423 if (argc < 2) { 1424 archive_set_error(a, ARCHIVE_ERRNO_FILE_FORMAT, 1425 "Not enough arguments"); 1426 return ARCHIVE_WARN; 1427 } 1428 result = (*pack)(argc, numbers, &error); 1429 if (error != NULL) { 1430 archive_set_error(a, ARCHIVE_ERRNO_FILE_FORMAT, 1431 "%s", error); 1432 return ARCHIVE_WARN; 1433 } 1434 } else { 1435 /* file system raw value. */ 1436 result = (dev_t)mtree_atol(&val); 1437 } 1438 *pdev = result; 1439 return ARCHIVE_OK; 1440 #undef MAX_PACK_ARGS 1441 } 1442 1443 /* 1444 * Parse a single keyword and its value. 1445 */ 1446 static int 1447 parse_keyword(struct archive_read *a, struct mtree *mtree, 1448 struct archive_entry *entry, struct mtree_option *opt, int *parsed_kws) 1449 { 1450 char *val, *key; 1451 1452 key = opt->value; 1453 1454 if (*key == '\0') 1455 return (ARCHIVE_OK); 1456 1457 if (strcmp(key, "nochange") == 0) { 1458 *parsed_kws |= MTREE_HAS_NOCHANGE; 1459 return (ARCHIVE_OK); 1460 } 1461 if (strcmp(key, "optional") == 0) { 1462 *parsed_kws |= MTREE_HAS_OPTIONAL; 1463 return (ARCHIVE_OK); 1464 } 1465 if (strcmp(key, "ignore") == 0) { 1466 /* 1467 * The mtree processing is not recursive, so 1468 * recursion will only happen for explicitly listed 1469 * entries. 1470 */ 1471 return (ARCHIVE_OK); 1472 } 1473 1474 val = strchr(key, '='); 1475 if (val == NULL) { 1476 archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT, 1477 "Malformed attribute \"%s\" (%d)", key, key[0]); 1478 return (ARCHIVE_WARN); 1479 } 1480 1481 *val = '\0'; 1482 ++val; 1483 1484 switch (key[0]) { 1485 case 'c': 1486 if (strcmp(key, "content") == 0 1487 || strcmp(key, "contents") == 0) { 1488 parse_escapes(val, NULL); 1489 archive_strcpy(&mtree->contents_name, val); 1490 break; 1491 } 1492 if (strcmp(key, "cksum") == 0) 1493 break; 1494 case 'd': 1495 if (strcmp(key, "device") == 0) { 1496 /* stat(2) st_rdev field, e.g. the major/minor IDs 1497 * of a char/block special file */ 1498 int r; 1499 dev_t dev; 1500 1501 *parsed_kws |= MTREE_HAS_DEVICE; 1502 r = parse_device(&dev, &a->archive, val); 1503 if (r == ARCHIVE_OK) 1504 archive_entry_set_rdev(entry, dev); 1505 return r; 1506 } 1507 case 'f': 1508 if (strcmp(key, "flags") == 0) { 1509 *parsed_kws |= MTREE_HAS_FFLAGS; 1510 archive_entry_copy_fflags_text(entry, val); 1511 break; 1512 } 1513 case 'g': 1514 if (strcmp(key, "gid") == 0) { 1515 *parsed_kws |= MTREE_HAS_GID; 1516 archive_entry_set_gid(entry, mtree_atol10(&val)); 1517 break; 1518 } 1519 if (strcmp(key, "gname") == 0) { 1520 *parsed_kws |= MTREE_HAS_GNAME; 1521 archive_entry_copy_gname(entry, val); 1522 break; 1523 } 1524 case 'i': 1525 if (strcmp(key, "inode") == 0) { 1526 archive_entry_set_ino(entry, mtree_atol10(&val)); 1527 break; 1528 } 1529 case 'l': 1530 if (strcmp(key, "link") == 0) { 1531 archive_entry_copy_symlink(entry, val); 1532 break; 1533 } 1534 case 'm': 1535 if (strcmp(key, "md5") == 0 || strcmp(key, "md5digest") == 0) 1536 break; 1537 if (strcmp(key, "mode") == 0) { 1538 if (val[0] >= '0' && val[0] <= '9') { 1539 *parsed_kws |= MTREE_HAS_PERM; 1540 archive_entry_set_perm(entry, 1541 (mode_t)mtree_atol8(&val)); 1542 } else { 1543 archive_set_error(&a->archive, 1544 ARCHIVE_ERRNO_FILE_FORMAT, 1545 "Symbolic mode \"%s\" unsupported", val); 1546 return ARCHIVE_WARN; 1547 } 1548 break; 1549 } 1550 case 'n': 1551 if (strcmp(key, "nlink") == 0) { 1552 *parsed_kws |= MTREE_HAS_NLINK; 1553 archive_entry_set_nlink(entry, 1554 (unsigned int)mtree_atol10(&val)); 1555 break; 1556 } 1557 case 'r': 1558 if (strcmp(key, "resdevice") == 0) { 1559 /* stat(2) st_dev field, e.g. the device ID where the 1560 * inode resides */ 1561 int r; 1562 dev_t dev; 1563 1564 r = parse_device(&dev, &a->archive, val); 1565 if (r == ARCHIVE_OK) 1566 archive_entry_set_dev(entry, dev); 1567 return r; 1568 } 1569 if (strcmp(key, "rmd160") == 0 || 1570 strcmp(key, "rmd160digest") == 0) 1571 break; 1572 case 's': 1573 if (strcmp(key, "sha1") == 0 || strcmp(key, "sha1digest") == 0) 1574 break; 1575 if (strcmp(key, "sha256") == 0 || 1576 strcmp(key, "sha256digest") == 0) 1577 break; 1578 if (strcmp(key, "sha384") == 0 || 1579 strcmp(key, "sha384digest") == 0) 1580 break; 1581 if (strcmp(key, "sha512") == 0 || 1582 strcmp(key, "sha512digest") == 0) 1583 break; 1584 if (strcmp(key, "size") == 0) { 1585 archive_entry_set_size(entry, mtree_atol10(&val)); 1586 break; 1587 } 1588 case 't': 1589 if (strcmp(key, "tags") == 0) { 1590 /* 1591 * Comma delimited list of tags. 1592 * Ignore the tags for now, but the interface 1593 * should be extended to allow inclusion/exclusion. 1594 */ 1595 break; 1596 } 1597 if (strcmp(key, "time") == 0) { 1598 int64_t m; 1599 int64_t my_time_t_max = get_time_t_max(); 1600 int64_t my_time_t_min = get_time_t_min(); 1601 long ns = 0; 1602 1603 *parsed_kws |= MTREE_HAS_MTIME; 1604 m = mtree_atol10(&val); 1605 /* Replicate an old mtree bug: 1606 * 123456789.1 represents 123456789 1607 * seconds and 1 nanosecond. */ 1608 if (*val == '.') { 1609 ++val; 1610 ns = (long)mtree_atol10(&val); 1611 if (ns < 0) 1612 ns = 0; 1613 else if (ns > 999999999) 1614 ns = 999999999; 1615 } 1616 if (m > my_time_t_max) 1617 m = my_time_t_max; 1618 else if (m < my_time_t_min) 1619 m = my_time_t_min; 1620 archive_entry_set_mtime(entry, (time_t)m, ns); 1621 break; 1622 } 1623 if (strcmp(key, "type") == 0) { 1624 switch (val[0]) { 1625 case 'b': 1626 if (strcmp(val, "block") == 0) { 1627 archive_entry_set_filetype(entry, AE_IFBLK); 1628 break; 1629 } 1630 case 'c': 1631 if (strcmp(val, "char") == 0) { 1632 archive_entry_set_filetype(entry, 1633 AE_IFCHR); 1634 break; 1635 } 1636 case 'd': 1637 if (strcmp(val, "dir") == 0) { 1638 archive_entry_set_filetype(entry, 1639 AE_IFDIR); 1640 break; 1641 } 1642 case 'f': 1643 if (strcmp(val, "fifo") == 0) { 1644 archive_entry_set_filetype(entry, 1645 AE_IFIFO); 1646 break; 1647 } 1648 if (strcmp(val, "file") == 0) { 1649 archive_entry_set_filetype(entry, 1650 AE_IFREG); 1651 break; 1652 } 1653 case 'l': 1654 if (strcmp(val, "link") == 0) { 1655 archive_entry_set_filetype(entry, 1656 AE_IFLNK); 1657 break; 1658 } 1659 default: 1660 archive_set_error(&a->archive, 1661 ARCHIVE_ERRNO_FILE_FORMAT, 1662 "Unrecognized file type \"%s\"; " 1663 "assuming \"file\"", val); 1664 archive_entry_set_filetype(entry, AE_IFREG); 1665 return (ARCHIVE_WARN); 1666 } 1667 *parsed_kws |= MTREE_HAS_TYPE; 1668 break; 1669 } 1670 case 'u': 1671 if (strcmp(key, "uid") == 0) { 1672 *parsed_kws |= MTREE_HAS_UID; 1673 archive_entry_set_uid(entry, mtree_atol10(&val)); 1674 break; 1675 } 1676 if (strcmp(key, "uname") == 0) { 1677 *parsed_kws |= MTREE_HAS_UNAME; 1678 archive_entry_copy_uname(entry, val); 1679 break; 1680 } 1681 default: 1682 archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT, 1683 "Unrecognized key %s=%s", key, val); 1684 return (ARCHIVE_WARN); 1685 } 1686 return (ARCHIVE_OK); 1687 } 1688 1689 static int 1690 read_data(struct archive_read *a, const void **buff, size_t *size, 1691 int64_t *offset) 1692 { 1693 size_t bytes_to_read; 1694 ssize_t bytes_read; 1695 struct mtree *mtree; 1696 1697 mtree = (struct mtree *)(a->format->data); 1698 if (mtree->fd < 0) { 1699 *buff = NULL; 1700 *offset = 0; 1701 *size = 0; 1702 return (ARCHIVE_EOF); 1703 } 1704 if (mtree->buff == NULL) { 1705 mtree->buffsize = 64 * 1024; 1706 mtree->buff = malloc(mtree->buffsize); 1707 if (mtree->buff == NULL) { 1708 archive_set_error(&a->archive, ENOMEM, 1709 "Can't allocate memory"); 1710 return (ARCHIVE_FATAL); 1711 } 1712 } 1713 1714 *buff = mtree->buff; 1715 *offset = mtree->offset; 1716 if ((int64_t)mtree->buffsize > mtree->cur_size - mtree->offset) 1717 bytes_to_read = (size_t)(mtree->cur_size - mtree->offset); 1718 else 1719 bytes_to_read = mtree->buffsize; 1720 bytes_read = read(mtree->fd, mtree->buff, bytes_to_read); 1721 if (bytes_read < 0) { 1722 archive_set_error(&a->archive, errno, "Can't read"); 1723 return (ARCHIVE_WARN); 1724 } 1725 if (bytes_read == 0) { 1726 *size = 0; 1727 return (ARCHIVE_EOF); 1728 } 1729 mtree->offset += bytes_read; 1730 *size = bytes_read; 1731 return (ARCHIVE_OK); 1732 } 1733 1734 /* Skip does nothing except possibly close the contents file. */ 1735 static int 1736 skip(struct archive_read *a) 1737 { 1738 struct mtree *mtree; 1739 1740 mtree = (struct mtree *)(a->format->data); 1741 if (mtree->fd >= 0) { 1742 close(mtree->fd); 1743 mtree->fd = -1; 1744 } 1745 return (ARCHIVE_OK); 1746 } 1747 1748 /* 1749 * Since parsing backslash sequences always makes strings shorter, 1750 * we can always do this conversion in-place. 1751 */ 1752 static void 1753 parse_escapes(char *src, struct mtree_entry *mentry) 1754 { 1755 char *dest = src; 1756 char c; 1757 1758 if (mentry != NULL && strcmp(src, ".") == 0) 1759 mentry->full = 1; 1760 1761 while (*src != '\0') { 1762 c = *src++; 1763 if (c == '/' && mentry != NULL) 1764 mentry->full = 1; 1765 if (c == '\\') { 1766 switch (src[0]) { 1767 case '0': 1768 if (src[1] < '0' || src[1] > '7') { 1769 c = 0; 1770 ++src; 1771 break; 1772 } 1773 /* FALLTHROUGH */ 1774 case '1': 1775 case '2': 1776 case '3': 1777 if (src[1] >= '0' && src[1] <= '7' && 1778 src[2] >= '0' && src[2] <= '7') { 1779 c = (src[0] - '0') << 6; 1780 c |= (src[1] - '0') << 3; 1781 c |= (src[2] - '0'); 1782 src += 3; 1783 } 1784 break; 1785 case 'a': 1786 c = '\a'; 1787 ++src; 1788 break; 1789 case 'b': 1790 c = '\b'; 1791 ++src; 1792 break; 1793 case 'f': 1794 c = '\f'; 1795 ++src; 1796 break; 1797 case 'n': 1798 c = '\n'; 1799 ++src; 1800 break; 1801 case 'r': 1802 c = '\r'; 1803 ++src; 1804 break; 1805 case 's': 1806 c = ' '; 1807 ++src; 1808 break; 1809 case 't': 1810 c = '\t'; 1811 ++src; 1812 break; 1813 case 'v': 1814 c = '\v'; 1815 ++src; 1816 break; 1817 case '\\': 1818 c = '\\'; 1819 ++src; 1820 break; 1821 } 1822 } 1823 *dest++ = c; 1824 } 1825 *dest = '\0'; 1826 } 1827 1828 /* 1829 * Note that this implementation does not (and should not!) obey 1830 * locale settings; you cannot simply substitute strtol here, since 1831 * it does obey locale. 1832 */ 1833 static int64_t 1834 mtree_atol8(char **p) 1835 { 1836 int64_t l, limit, last_digit_limit; 1837 int digit, base; 1838 1839 base = 8; 1840 limit = INT64_MAX / base; 1841 last_digit_limit = INT64_MAX % base; 1842 1843 l = 0; 1844 digit = **p - '0'; 1845 while (digit >= 0 && digit < base) { 1846 if (l>limit || (l == limit && digit > last_digit_limit)) { 1847 l = INT64_MAX; /* Truncate on overflow. */ 1848 break; 1849 } 1850 l = (l * base) + digit; 1851 digit = *++(*p) - '0'; 1852 } 1853 return (l); 1854 } 1855 1856 /* 1857 * Note that this implementation does not (and should not!) obey 1858 * locale settings; you cannot simply substitute strtol here, since 1859 * it does obey locale. 1860 * 1861 * Convert the number pointed to by 'p' into a 64-bit signed integer. 1862 * On return, 'p' points to the first non-digit following the number. 1863 * On overflow, the function returns INT64_MIN or INT64_MAX. 1864 */ 1865 static int64_t 1866 mtree_atol10(char **p) 1867 { 1868 const int base = 10; 1869 const int64_t limit = INT64_MAX / base; 1870 const int64_t last_digit_limit = INT64_MAX % base; 1871 int64_t l; 1872 int sign; 1873 1874 if (**p == '-') { 1875 sign = -1; 1876 ++(*p); 1877 } else { 1878 sign = 1; 1879 } 1880 1881 l = 0; 1882 while (**p >= '0' && **p < '0' + base) { 1883 int digit = **p - '0'; 1884 if (l > limit || (l == limit && digit > last_digit_limit)) { 1885 while (**p >= '0' && **p < '0' + base) { 1886 ++(*p); 1887 } 1888 return (sign < 0) ? INT64_MIN : INT64_MAX; 1889 } 1890 l = (l * base) + digit; 1891 ++(*p); 1892 } 1893 return (sign < 0) ? -l : l; 1894 } 1895 1896 /* Parse a hex digit. */ 1897 static int 1898 parsehex(char c) 1899 { 1900 if (c >= '0' && c <= '9') 1901 return c - '0'; 1902 else if (c >= 'a' && c <= 'f') 1903 return c - 'a'; 1904 else if (c >= 'A' && c <= 'F') 1905 return c - 'A'; 1906 else 1907 return -1; 1908 } 1909 1910 /* 1911 * Note that this implementation does not (and should not!) obey 1912 * locale settings; you cannot simply substitute strtol here, since 1913 * it does obey locale. 1914 */ 1915 static int64_t 1916 mtree_atol16(char **p) 1917 { 1918 int64_t l, limit, last_digit_limit; 1919 int base, digit, sign; 1920 1921 base = 16; 1922 1923 if (**p == '-') { 1924 sign = -1; 1925 limit = ((uint64_t)(INT64_MAX) + 1) / base; 1926 last_digit_limit = ((uint64_t)(INT64_MAX) + 1) % base; 1927 ++(*p); 1928 } else { 1929 sign = 1; 1930 limit = INT64_MAX / base; 1931 last_digit_limit = INT64_MAX % base; 1932 } 1933 1934 l = 0; 1935 digit = parsehex(**p); 1936 while (digit >= 0 && digit < base) { 1937 if (l > limit || (l == limit && digit > last_digit_limit)) 1938 return (sign < 0) ? INT64_MIN : INT64_MAX; 1939 l = (l * base) + digit; 1940 digit = parsehex(*++(*p)); 1941 } 1942 return (sign < 0) ? -l : l; 1943 } 1944 1945 static int64_t 1946 mtree_atol(char **p) 1947 { 1948 if (**p != '0') 1949 return mtree_atol10(p); 1950 if ((*p)[1] == 'x' || (*p)[1] == 'X') { 1951 *p += 2; 1952 return mtree_atol16(p); 1953 } 1954 return mtree_atol8(p); 1955 } 1956 1957 /* 1958 * Returns length of line (including trailing newline) 1959 * or negative on error. 'start' argument is updated to 1960 * point to first character of line. 1961 */ 1962 static ssize_t 1963 readline(struct archive_read *a, struct mtree *mtree, char **start, 1964 ssize_t limit) 1965 { 1966 ssize_t bytes_read; 1967 ssize_t total_size = 0; 1968 ssize_t find_off = 0; 1969 const void *t; 1970 void *nl; 1971 char *u; 1972 1973 /* Accumulate line in a line buffer. */ 1974 for (;;) { 1975 /* Read some more. */ 1976 t = __archive_read_ahead(a, 1, &bytes_read); 1977 if (t == NULL) 1978 return (0); 1979 if (bytes_read < 0) 1980 return (ARCHIVE_FATAL); 1981 nl = memchr(t, '\n', bytes_read); 1982 /* If we found '\n', trim the read to end exactly there. */ 1983 if (nl != NULL) { 1984 bytes_read = ((const char *)nl) - ((const char *)t) + 1; 1985 } 1986 if (total_size + bytes_read + 1 > limit) { 1987 archive_set_error(&a->archive, 1988 ARCHIVE_ERRNO_FILE_FORMAT, 1989 "Line too long"); 1990 return (ARCHIVE_FATAL); 1991 } 1992 if (archive_string_ensure(&mtree->line, 1993 total_size + bytes_read + 1) == NULL) { 1994 archive_set_error(&a->archive, ENOMEM, 1995 "Can't allocate working buffer"); 1996 return (ARCHIVE_FATAL); 1997 } 1998 /* Append new bytes to string. */ 1999 memcpy(mtree->line.s + total_size, t, bytes_read); 2000 __archive_read_consume(a, bytes_read); 2001 total_size += bytes_read; 2002 mtree->line.s[total_size] = '\0'; 2003 2004 for (u = mtree->line.s + find_off; *u; ++u) { 2005 if (u[0] == '\n') { 2006 /* Ends with unescaped newline. */ 2007 *start = mtree->line.s; 2008 return total_size; 2009 } else if (u[0] == '#') { 2010 /* Ends with comment sequence #...\n */ 2011 if (nl == NULL) { 2012 /* But we've not found the \n yet */ 2013 break; 2014 } 2015 } else if (u[0] == '\\') { 2016 if (u[1] == '\n') { 2017 /* Trim escaped newline. */ 2018 total_size -= 2; 2019 mtree->line.s[total_size] = '\0'; 2020 break; 2021 } else if (u[1] != '\0') { 2022 /* Skip the two-char escape sequence */ 2023 ++u; 2024 } 2025 } 2026 } 2027 find_off = u - mtree->line.s; 2028 } 2029 } 2030 2031 static unsigned int 2032 hash(const char *p) 2033 { 2034 /* A 32-bit version of Peter Weinberger's (PJW) hash algorithm, 2035 as used by ELF for hashing function names. */ 2036 unsigned g, h = 0; 2037 while (*p != '\0') { 2038 h = (h << 4) + *p++; 2039 if ((g = h & 0xF0000000) != 0) { 2040 h ^= g >> 24; 2041 h &= 0x0FFFFFFF; 2042 } 2043 } 2044 return h; 2045 } 2046