external/bsd/libarchive/dist/tar/util.c

   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 #include "bsdtar_platform.h"
  27 __FBSDID("$FreeBSD: src/usr.bin/tar/util.c,v 1.23 2008/12/15 06:00:25 kientzle Exp $");
  28
  29 #ifdef HAVE_SYS_STAT_H
  30 #include <sys/stat.h>
  31 #endif
  32 #ifdef HAVE_SYS_TYPES_H
  33 #include <sys/types.h>  /* Linux doesn't define mode_t, etc. in sys/stat.h. */
  34 #endif
  35 #include <ctype.h>
  36 #ifdef HAVE_ERRNO_H
  37 #include <errno.h>
  38 #endif
  39 #ifdef HAVE_IO_H
  40 #include <io.h>
  41 #endif
  42 #ifdef HAVE_STDARG_H
  43 #include <stdarg.h>
  44 #endif
  45 #ifdef HAVE_STDINT_H
  46 #include <stdint.h>
  47 #endif
  48 #include <stdio.h>
  49 #ifdef HAVE_STDLIB_H
  50 #include <stdlib.h>
  51 #endif
  52 #ifdef HAVE_STRING_H
  53 #include <string.h>
  54 #endif
  55 #ifdef HAVE_WCTYPE_H
  56 #include <wctype.h>
  57 #else
  58 /* If we don't have wctype, we need to hack up some version of iswprint(). */
  59 #define iswprint isprint
  60 #endif
  61
  62 #include "bsdtar.h"
  63 #include "err.h"
  64
  65 static size_t   bsdtar_expand_char(char *, size_t, char);
  66 static const char *strip_components(const char *path, int elements);
  67
  68 #if defined(_WIN32) && !defined(__CYGWIN__)
  69 #define read _read
  70 #endif
  71
  72 /* TODO:  Hack up a version of mbtowc for platforms with no wide
  73  * character support at all.  I think the following might suffice,
  74  * but it needs careful testing.
  75  * #if !HAVE_MBTOWC
  76  * #define mbtowc(wcp, p, n) ((*wcp = *p), 1)
  77  * #endif
  78  */
  79
  80 /*
  81  * Print a string, taking care with any non-printable characters.
  82  *
  83  * Note that we use a stack-allocated buffer to receive the formatted
  84  * string if we can.  This is partly performance (avoiding a call to
  85  * malloc()), partly out of expedience (we have to call vsnprintf()
  86  * before malloc() anyway to find out how big a buffer we need; we may
  87  * as well point that first call at a small local buffer in case it
  88  * works), but mostly for safety (so we can use this to print messages
  89  * about out-of-memory conditions).
  90  */
  91
  92 void
  93 safe_fprintf(FILE *f, const char *fmt, ...)
  94 {
  95         char fmtbuff_stack[256]; /* Place to format the printf() string. */
  96         char outbuff[256]; /* Buffer for outgoing characters. */
  97         char *fmtbuff_heap; /* If fmtbuff_stack is too small, we use malloc */
  98         char *fmtbuff;  /* Pointer to fmtbuff_stack or fmtbuff_heap. */
  99         int fmtbuff_length;
 100         int length, n;
 101         va_list ap;
 102         const char *p;
 103         unsigned i;
 104         wchar_t wc;
 105         char try_wc;
 106
 107         /* Use a stack-allocated buffer if we can, for speed and safety. */
 108         fmtbuff_heap = NULL;
 109         fmtbuff_length = sizeof(fmtbuff_stack);
 110         fmtbuff = fmtbuff_stack;
 111
 112         /* Try formatting into the stack buffer. */
 113         va_start(ap, fmt);
 114         length = vsnprintf(fmtbuff, fmtbuff_length, fmt, ap);
 115         va_end(ap);
 116
 117         /* If the result was too large, allocate a buffer on the heap. */
 118         if (length >= fmtbuff_length) {
 119                 fmtbuff_length = length+1;
 120                 fmtbuff_heap = malloc(fmtbuff_length);
 121
 122                 /* Reformat the result into the heap buffer if we can. */
 123                 if (fmtbuff_heap != NULL) {
 124                         fmtbuff = fmtbuff_heap;
 125                         va_start(ap, fmt);
 126                         length = vsnprintf(fmtbuff, fmtbuff_length, fmt, ap);
 127                         va_end(ap);
 128                 } else {
 129                         /* Leave fmtbuff pointing to the truncated
 130                          * string in fmtbuff_stack. */
 131                         length = sizeof(fmtbuff_stack) - 1;
 132                 }
 133         }
 134
 135         /* Note: mbrtowc() has a cleaner API, but mbtowc() seems a bit
 136          * more portable, so we use that here instead. */
 137         n = mbtowc(NULL, NULL, 1); /* Reset the shift state. */
 138
 139         /* Write data, expanding unprintable characters. */
 140         p = fmtbuff;
 141         i = 0;
 142         try_wc = 1;
 143         while (*p != '\0') {
 144
 145                 /* Convert to wide char, test if the wide
 146                  * char is printable in the current locale. */
 147                 if (try_wc && (n = mbtowc(&wc, p, length)) != -1) {
 148                         length -= n;
 149                         if (iswprint(wc) && wc != L'\\') {
 150                                 /* Printable, copy the bytes through. */
 151                                 while (n-- > 0)
 152                                         outbuff[i++] = *p++;
 153                         } else {
 154                                 /* Not printable, format the bytes. */
 155                                 while (n-- > 0)
 156                                         i += (unsigned)bsdtar_expand_char(
 157                                             outbuff, i, *p++);
 158                         }
 159                 } else {
 160                         /* After any conversion failure, don't bother
 161                          * trying to convert the rest. */
 162                         i += (unsigned)bsdtar_expand_char(outbuff, i, *p++);
 163                         try_wc = 0;
 164                 }
 165
 166                 /* If our output buffer is full, dump it and keep going. */
 167                 if (i > (sizeof(outbuff) - 20)) {
 168                         outbuff[i] = '\0';
 169                         fprintf(f, "%s", outbuff);
 170                         i = 0;
 171                 }
 172         }
 173         outbuff[i] = '\0';
 174         fprintf(f, "%s", outbuff);
 175
 176         /* If we allocated a heap-based formatting buffer, free it now. */
 177         if (fmtbuff_heap != NULL)
 178                 free(fmtbuff_heap);
 179 }
 180
 181 /*
 182  * Render an arbitrary sequence of bytes into printable ASCII characters.
 183  */
 184 static size_t
 185 bsdtar_expand_char(char *buff, size_t offset, char c)
 186 {
 187         size_t i = offset;
 188
 189         if (isprint((unsigned char)c) && c != '\\')
 190                 buff[i++] = c;
 191         else {
 192                 buff[i++] = '\\';
 193                 switch (c) {
 194                 case '\a': buff[i++] = 'a'; break;
 195                 case '\b': buff[i++] = 'b'; break;
 196                 case '\f': buff[i++] = 'f'; break;
 197                 case '\n': buff[i++] = 'n'; break;
 198 #if '\r' != '\n'
 199                 /* On some platforms, \n and \r are the same. */
 200                 case '\r': buff[i++] = 'r'; break;
 201 #endif
 202                 case '\t': buff[i++] = 't'; break;
 203                 case '\v': buff[i++] = 'v'; break;
 204                 case '\\': buff[i++] = '\\'; break;
 205                 default:
 206                         sprintf(buff + i, "%03o", 0xFF & (int)c);
 207                         i += 3;
 208                 }
 209         }
 210
 211         return (i - offset);
 212 }
 213
 214 int
 215 yes(const char *fmt, ...)
 216 {
 217         char buff[32];
 218         char *p;
 219         ssize_t l;
 220
 221         va_list ap;
 222         va_start(ap, fmt);
 223         vfprintf(stderr, fmt, ap);
 224         va_end(ap);
 225         fprintf(stderr, " (y/N)? ");
 226         fflush(stderr);
 227
 228         l = read(2, buff, sizeof(buff) - 1);
 229         if (l <= 0)
 230                 return (0);
 231         buff[l] = 0;
 232
 233         for (p = buff; *p != '\0'; p++) {
 234                 if (isspace((unsigned char)*p))
 235                         continue;
 236                 switch(*p) {
 237                 case 'y': case 'Y':
 238                         return (1);
 239                 case 'n': case 'N':
 240                         return (0);
 241                 default:
 242                         return (0);
 243                 }
 244         }
 245
 246         return (0);
 247 }
 248
 249 /*-
 250  * The logic here for -C <dir> attempts to avoid
 251  * chdir() as long as possible.  For example:
 252  * "-C /foo -C /bar file"          needs chdir("/bar") but not chdir("/foo")
 253  * "-C /foo -C bar file"           needs chdir("/foo/bar")
 254  * "-C /foo -C bar /file1"         does not need chdir()
 255  * "-C /foo -C bar /file1 file2"   needs chdir("/foo/bar") before file2
 256  *
 257  * The only correct way to handle this is to record a "pending" chdir
 258  * request and combine multiple requests intelligently until we
 259  * need to process a non-absolute file.  set_chdir() adds the new dir
 260  * to the pending list; do_chdir() actually executes any pending chdir.
 261  *
 262  * This way, programs that build tar command lines don't have to worry
 263  * about -C with non-existent directories; such requests will only
 264  * fail if the directory must be accessed.
 265  *
 266  * TODO: Make this handle Windows paths correctly.
 267  */
 268 void
 269 set_chdir(struct bsdtar *bsdtar, const char *newdir)
 270 {
 271         if (newdir[0] == '/') {
 272                 /* The -C /foo -C /bar case; dump first one. */
 273                 free(bsdtar->pending_chdir);
 274                 bsdtar->pending_chdir = NULL;
 275         }
 276         if (bsdtar->pending_chdir == NULL)
 277                 /* Easy case: no previously-saved dir. */
 278                 bsdtar->pending_chdir = strdup(newdir);
 279         else {
 280                 /* The -C /foo -C bar case; concatenate */
 281                 char *old_pending = bsdtar->pending_chdir;
 282                 size_t old_len = strlen(old_pending);
 283                 bsdtar->pending_chdir = malloc(old_len + strlen(newdir) + 2);
 284                 if (old_pending[old_len - 1] == '/')
 285                         old_pending[old_len - 1] = '\0';
 286                 if (bsdtar->pending_chdir != NULL)
 287                         sprintf(bsdtar->pending_chdir, "%s/%s",
 288                             old_pending, newdir);
 289                 free(old_pending);
 290         }
 291         if (bsdtar->pending_chdir == NULL)
 292                 lafe_errc(1, errno, "No memory");
 293 }
 294
 295 void
 296 do_chdir(struct bsdtar *bsdtar)
 297 {
 298         if (bsdtar->pending_chdir == NULL)
 299                 return;
 300
 301         if (chdir(bsdtar->pending_chdir) != 0) {
 302                 lafe_errc(1, 0, "could not chdir to '%s'\n",
 303                     bsdtar->pending_chdir);
 304         }
 305         free(bsdtar->pending_chdir);
 306         bsdtar->pending_chdir = NULL;
 307 }
 308
 309 static const char *
 310 strip_components(const char *p, int elements)
 311 {
 312         /* Skip as many elements as necessary. */
 313         while (elements > 0) {
 314                 switch (*p++) {
 315                 case '/':
 316 #if defined(_WIN32) && !defined(__CYGWIN__)
 317                 case '\\': /* Support \ path sep on Windows ONLY. */
 318 #endif
 319                         elements--;
 320                         break;
 321                 case '\0':
 322                         /* Path is too short, skip it. */
 323                         return (NULL);
 324                 }
 325         }
 326
 327         /* Skip any / characters.  This handles short paths that have
 328          * additional / termination.  This also handles the case where
 329          * the logic above stops in the middle of a duplicate //
 330          * sequence (which would otherwise get converted to an
 331          * absolute path). */
 332         for (;;) {
 333                 switch (*p) {
 334                 case '/':
 335 #if defined(_WIN32) && !defined(__CYGWIN__)
 336                 case '\\': /* Support \ path sep on Windows ONLY. */
 337 #endif
 338                         ++p;
 339                         break;
 340                 case '\0':
 341                         return (NULL);
 342                 default:
 343                         return (p);
 344                 }
 345         }
 346 }
 347
 348 /*
 349  * Handle --strip-components and any future path-rewriting options.
 350  * Returns non-zero if the pathname should not be extracted.
 351  *
 352  * TODO: Support pax-style regex path rewrites.
 353  */
 354 int
 355 edit_pathname(struct bsdtar *bsdtar, struct archive_entry *entry)
 356 {
 357         const char *name = archive_entry_pathname(entry);
 358 #if HAVE_REGEX_H
 359         char *subst_name;
 360         int r;
 361 #endif
 362
 363 #if HAVE_REGEX_H
 364         r = apply_substitution(bsdtar, name, &subst_name, 0);
 365         if (r == -1) {
 366                 lafe_warnc(0, "Invalid substitution, skipping entry");
 367                 return 1;
 368         }
 369         if (r == 1) {
 370                 archive_entry_copy_pathname(entry, subst_name);
 371                 if (*subst_name == '\0') {
 372                         free(subst_name);
 373                         return -1;
 374                 } else
 375                         free(subst_name);
 376                 name = archive_entry_pathname(entry);
 377         }
 378
 379         if (archive_entry_hardlink(entry)) {
 380                 r = apply_substitution(bsdtar, archive_entry_hardlink(entry), &subst_name, 1);
 381                 if (r == -1) {
 382                         lafe_warnc(0, "Invalid substitution, skipping entry");
 383                         return 1;
 384                 }
 385                 if (r == 1) {
 386                         archive_entry_copy_hardlink(entry, subst_name);
 387                         free(subst_name);
 388                 }
 389         }
 390         if (archive_entry_symlink(entry) != NULL) {
 391                 r = apply_substitution(bsdtar, archive_entry_symlink(entry), &subst_name, 1);
 392                 if (r == -1) {
 393                         lafe_warnc(0, "Invalid substitution, skipping entry");
 394                         return 1;
 395                 }
 396                 if (r == 1) {
 397                         archive_entry_copy_symlink(entry, subst_name);
 398                         free(subst_name);
 399                 }
 400         }
 401 #endif
 402
 403         /* Strip leading dir names as per --strip-components option. */
 404         if (bsdtar->strip_components > 0) {
 405                 const char *linkname = archive_entry_hardlink(entry);
 406
 407                 name = strip_components(name, bsdtar->strip_components);
 408                 if (name == NULL)
 409                         return (1);
 410
 411                 if (linkname != NULL) {
 412                         linkname = strip_components(linkname,
 413                             bsdtar->strip_components);
 414                         if (linkname == NULL)
 415                                 return (1);
 416                         archive_entry_copy_hardlink(entry, linkname);
 417                 }
 418         }
 419
 420         /* By default, don't write or restore absolute pathnames. */
 421         if (!bsdtar->option_absolute_paths) {
 422                 const char *rp, *p = name;
 423                 int slashonly = 1;
 424
 425                 /* Remove leading "//./" or "//?/" or "//?/UNC/"
 426                  * (absolute path prefixes used by Windows API) */
 427                 if ((p[0] == '/' || p[0] == '\\') &&
 428                     (p[1] == '/' || p[1] == '\\') &&
 429                     (p[2] == '.' || p[2] == '?') &&
 430                     (p[3] == '/' || p[3] == '\\'))
 431                 {
 432                         if (p[2] == '?' &&
 433                             (p[4] == 'U' || p[4] == 'u') &&
 434                             (p[5] == 'N' || p[5] == 'n') &&
 435                             (p[6] == 'C' || p[6] == 'c') &&
 436                             (p[7] == '/' || p[7] == '\\'))
 437                                 p += 8;
 438                         else
 439                                 p += 4;
 440                         slashonly = 0;
 441                 }
 442                 do {
 443                         rp = p;
 444                         /* Remove leading drive letter from archives created
 445                          * on Windows. */
 446                         if (((p[0] >= 'a' && p[0] <= 'z') ||
 447                              (p[0] >= 'A' && p[0] <= 'Z')) &&
 448                                  p[1] == ':') {
 449                                 p += 2;
 450                                 slashonly = 0;
 451                         }
 452                         /* Remove leading "/../", "//", etc. */
 453                         while (p[0] == '/' || p[0] == '\\') {
 454                                 if (p[1] == '.' && p[2] == '.' &&
 455                                         (p[3] == '/' || p[3] == '\\')) {
 456                                         p += 3; /* Remove "/..", leave "/"
 457                                                          * for next pass. */
 458                                         slashonly = 0;
 459                                 } else
 460                                         p += 1; /* Remove "/". */
 461                         }
 462                 } while (rp != p);
 463
 464                 if (p != name && !bsdtar->warned_lead_slash) {
 465                         /* Generate a warning the first time this happens. */
 466                         if (slashonly)
 467                                 lafe_warnc(0,
 468                                     "Removing leading '%c' from member names",
 469                                     name[0]);
 470                         else
 471                                 lafe_warnc(0,
 472                                     "Removing leading drive letter from "
 473                                     "member names");
 474                         bsdtar->warned_lead_slash = 1;
 475                 }
 476
 477                 /* Special case: Stripping everything yields ".". */
 478                 if (*p == '\0')
 479                         name = ".";
 480                 else
 481                         name = p;
 482         } else {
 483                 /* Strip redundant leading '/' characters. */
 484                 while (name[0] == '/' && name[1] == '/')
 485                         name++;
 486         }
 487
 488         /* Safely replace name in archive_entry. */
 489         if (name != archive_entry_pathname(entry)) {
 490                 char *q = strdup(name);
 491                 archive_entry_copy_pathname(entry, q);
 492                 free(q);
 493         }
 494         return (0);
 495 }
 496
 497 /*
 498  * It would be nice to just use printf() for formatting large numbers,
 499  * but the compatibility problems are quite a headache.  Hence the
 500  * following simple utility function.
 501  */
 502 const char *
 503 tar_i64toa(int64_t n0)
 504 {
 505         static char buff[24];
 506         int64_t n = n0 < 0 ? -n0 : n0;
 507         char *p = buff + sizeof(buff);
 508
 509         *--p = '\0';
 510         do {
 511                 *--p = '0' + (int)(n % 10);
 512                 n /= 10;
 513         } while (n > 0);
 514         if (n0 < 0)
 515                 *--p = '-';
 516         return p;
 517 }
 518
 519 /*
 520  * Like strcmp(), but try to be a little more aware of the fact that
 521  * we're comparing two paths.  Right now, it just handles leading
 522  * "./" and trailing '/' specially, so that "a/b/" == "./a/b"
 523  *
 524  * TODO: Make this better, so that "./a//b/./c/" == "a/b/c"
 525  * TODO: After this works, push it down into libarchive.
 526  * TODO: Publish the path normalization routines in libarchive so
 527  * that bsdtar can normalize paths and use fast strcmp() instead
 528  * of this.
 529  *
 530  * Note: This is currently only used within write.c, so should
 531  * not handle \ path separators.
 532  */
 533
 534 int
 535 pathcmp(const char *a, const char *b)
 536 {
 537         /* Skip leading './' */
 538         if (a[0] == '.' && a[1] == '/' && a[2] != '\0')
 539                 a += 2;
 540         if (b[0] == '.' && b[1] == '/' && b[2] != '\0')
 541                 b += 2;
 542         /* Find the first difference, or return (0) if none. */
 543         while (*a == *b) {
 544                 if (*a == '\0')
 545                         return (0);
 546                 a++;
 547                 b++;
 548         }
 549         /*
 550          * If one ends in '/' and the other one doesn't,
 551          * they're the same.
 552          */
 553         if (a[0] == '/' && a[1] == '\0' && b[0] == '\0')
 554                 return (0);
 555         if (a[0] == '\0' && b[0] == '/' && b[1] == '\0')
 556                 return (0);
 557         /* They're really different, return the correct sign. */
 558         return (*(const unsigned char *)a - *(const unsigned char *)b);
 559 }