contrib/libarchive-2/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 1.17 2007/04/18 04:36:11 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_STDARG_H
  40 #include <stdarg.h>
  41 #endif
  42 #include <stdio.h>
  43 #ifdef HAVE_STDLIB_H
  44 #include <stdlib.h>
  45 #endif
  46 #ifdef HAVE_STRING_H
  47 #include <string.h>
  48 #endif
  49
  50 #include "bsdtar.h"
  51
  52 static void     bsdtar_vwarnc(struct bsdtar *, int code,
  53                     const char *fmt, va_list ap);
  54
  55 /*
  56  * Print a string, taking care with any non-printable characters.
  57  */
  58
  59 void
  60 safe_fprintf(FILE *f, const char *fmt, ...)
  61 {
  62         char *buff;
  63         char *buff_heap;
  64         int buff_length;
  65         int length;
  66         va_list ap;
  67         char *p;
  68         unsigned i;
  69         char buff_stack[256];
  70         char copy_buff[256];
  71
  72         /* Use a stack-allocated buffer if we can, for speed and safety. */
  73         buff_heap = NULL;
  74         buff_length = sizeof(buff_stack);
  75         buff = buff_stack;
  76
  77         va_start(ap, fmt);
  78         length = vsnprintf(buff, buff_length, fmt, ap);
  79         va_end(ap);
  80         /* If the result is too large, allocate a buffer on the heap. */
  81         if (length >= buff_length) {
  82                 buff_length = length+1;
  83                 buff_heap = malloc(buff_length);
  84                 /* Failsafe: use the truncated string if malloc fails. */
  85                 if (buff_heap != NULL) {
  86                         buff = buff_heap;
  87                         va_start(ap, fmt);
  88                         length = vsnprintf(buff, buff_length, fmt, ap);
  89                         va_end(ap);
  90                 }
  91         }
  92
  93         /* Write data, expanding unprintable characters. */
  94         p = buff;
  95         i = 0;
  96         while (*p != '\0') {
  97                 unsigned char c = *p++;
  98
  99                 if (isprint(c) && c != '\\')
 100                         copy_buff[i++] = c;
 101                 else {
 102                         copy_buff[i++] = '\\';
 103                         switch (c) {
 104                         case '\a': copy_buff[i++] = 'a'; break;
 105                         case '\b': copy_buff[i++] = 'b'; break;
 106                         case '\f': copy_buff[i++] = 'f'; break;
 107                         case '\n': copy_buff[i++] = 'n'; break;
 108 #if '\r' != '\n'
 109                         /* On some platforms, \n and \r are the same. */
 110                         case '\r': copy_buff[i++] = 'r'; break;
 111 #endif
 112                         case '\t': copy_buff[i++] = 't'; break;
 113                         case '\v': copy_buff[i++] = 'v'; break;
 114                         case '\\': copy_buff[i++] = '\\'; break;
 115                         default:
 116                                 sprintf(copy_buff + i, "%03o", c);
 117                                 i += 3;
 118                         }
 119                 }
 120
 121                 /* If our temp buffer is full, dump it and keep going. */
 122                 if (i > (sizeof(copy_buff) - 8)) {
 123                         copy_buff[i++] = '\0';
 124                         fprintf(f, "%s", copy_buff);
 125                         i = 0;
 126                 }
 127         }
 128         copy_buff[i++] = '\0';
 129         fprintf(f, "%s", copy_buff);
 130
 131         /* If we allocated a heap-based buffer, free it now. */
 132         if (buff_heap != NULL)
 133                 free(buff_heap);
 134 }
 135
 136 static void
 137 bsdtar_vwarnc(struct bsdtar *bsdtar, int code, const char *fmt, va_list ap)
 138 {
 139         fprintf(stderr, "%s: ", bsdtar->progname);
 140         vfprintf(stderr, fmt, ap);
 141         if (code != 0)
 142                 fprintf(stderr, ": %s", strerror(code));
 143         fprintf(stderr, "\n");
 144 }
 145
 146 void
 147 bsdtar_warnc(struct bsdtar *bsdtar, int code, const char *fmt, ...)
 148 {
 149         va_list ap;
 150
 151         va_start(ap, fmt);
 152         bsdtar_vwarnc(bsdtar, code, fmt, ap);
 153         va_end(ap);
 154 }
 155
 156 void
 157 bsdtar_errc(struct bsdtar *bsdtar, int eval, int code, const char *fmt, ...)
 158 {
 159         va_list ap;
 160
 161         va_start(ap, fmt);
 162         bsdtar_vwarnc(bsdtar, code, fmt, ap);
 163         va_end(ap);
 164         exit(eval);
 165 }
 166
 167 int
 168 yes(const char *fmt, ...)
 169 {
 170         char buff[32];
 171         char *p;
 172         ssize_t l;
 173
 174         va_list ap;
 175         va_start(ap, fmt);
 176         vfprintf(stderr, fmt, ap);
 177         va_end(ap);
 178         fprintf(stderr, " (y/N)? ");
 179         fflush(stderr);
 180
 181         l = read(2, buff, sizeof(buff));
 182         if (l <= 0)
 183                 return (0);
 184         buff[l] = 0;
 185
 186         for (p = buff; *p != '\0'; p++) {
 187                 if (isspace(0xff & (int)*p))
 188                         continue;
 189                 switch(*p) {
 190                 case 'y': case 'Y':
 191                         return (1);
 192                 case 'n': case 'N':
 193                         return (0);
 194                 default:
 195                         return (0);
 196                 }
 197         }
 198
 199         return (0);
 200 }
 201
 202 /*
 203  * Read lines from file and do something with each one.  If option_null
 204  * is set, lines are terminated with zero bytes; otherwise, they're
 205  * terminated with newlines.
 206  *
 207  * This uses a self-sizing buffer to handle arbitrarily-long lines.
 208  * If the "process" function returns non-zero for any line, this
 209  * function will return non-zero after attempting to process all
 210  * remaining lines.
 211  */
 212 int
 213 process_lines(struct bsdtar *bsdtar, const char *pathname,
 214     int (*process)(struct bsdtar *, const char *))
 215 {
 216         FILE *f;
 217         char *buff, *buff_end, *line_start, *line_end, *p;
 218         size_t buff_length, bytes_read, bytes_wanted;
 219         int separator;
 220         int ret;
 221
 222         separator = bsdtar->option_null ? '\0' : '\n';
 223         ret = 0;
 224
 225         if (strcmp(pathname, "-") == 0)
 226                 f = stdin;
 227         else
 228                 f = fopen(pathname, "r");
 229         if (f == NULL)
 230                 bsdtar_errc(bsdtar, 1, errno, "Couldn't open %s", pathname);
 231         buff_length = 8192;
 232         buff = malloc(buff_length);
 233         if (buff == NULL)
 234                 bsdtar_errc(bsdtar, 1, ENOMEM, "Can't read %s", pathname);
 235         line_start = line_end = buff_end = buff;
 236         for (;;) {
 237                 /* Get some more data into the buffer. */
 238                 bytes_wanted = buff + buff_length - buff_end;
 239                 bytes_read = fread(buff_end, 1, bytes_wanted, f);
 240                 buff_end += bytes_read;
 241                 /* Process all complete lines in the buffer. */
 242                 while (line_end < buff_end) {
 243                         if (*line_end == separator) {
 244                                 *line_end = '\0';
 245                                 if ((*process)(bsdtar, line_start) != 0)
 246                                         ret = -1;
 247                                 line_start = line_end + 1;
 248                                 line_end = line_start;
 249                         } else
 250                                 line_end++;
 251                 }
 252                 if (feof(f))
 253                         break;
 254                 if (ferror(f))
 255                         bsdtar_errc(bsdtar, 1, errno,
 256                             "Can't read %s", pathname);
 257                 if (line_start > buff) {
 258                         /* Move a leftover fractional line to the beginning. */
 259                         memmove(buff, line_start, buff_end - line_start);
 260                         buff_end -= line_start - buff;
 261                         line_end -= line_start - buff;
 262                         line_start = buff;
 263                 } else {
 264                         /* Line is too big; enlarge the buffer. */
 265                         p = realloc(buff, buff_length *= 2);
 266                         if (p == NULL)
 267                                 bsdtar_errc(bsdtar, 1, ENOMEM,
 268                                     "Line too long in %s", pathname);
 269                         buff_end = p + (buff_end - buff);
 270                         line_end = p + (line_end - buff);
 271                         line_start = buff = p;
 272                 }
 273         }
 274         /* At end-of-file, handle the final line. */
 275         if (line_end > line_start) {
 276                 *line_end = '\0';
 277                 if ((*process)(bsdtar, line_start) != 0)
 278                         ret = -1;
 279         }
 280         free(buff);
 281         if (f != stdin)
 282                 fclose(f);
 283         return (ret);
 284 }
 285
 286 /*-
 287  * The logic here for -C <dir> attempts to avoid
 288  * chdir() as long as possible.  For example:
 289  * "-C /foo -C /bar file"          needs chdir("/bar") but not chdir("/foo")
 290  * "-C /foo -C bar file"           needs chdir("/foo/bar")
 291  * "-C /foo -C bar /file1"         does not need chdir()
 292  * "-C /foo -C bar /file1 file2"   needs chdir("/foo/bar") before file2
 293  *
 294  * The only correct way to handle this is to record a "pending" chdir
 295  * request and combine multiple requests intelligently until we
 296  * need to process a non-absolute file.  set_chdir() adds the new dir
 297  * to the pending list; do_chdir() actually executes any pending chdir.
 298  *
 299  * This way, programs that build tar command lines don't have to worry
 300  * about -C with non-existent directories; such requests will only
 301  * fail if the directory must be accessed.
 302  */
 303 void
 304 set_chdir(struct bsdtar *bsdtar, const char *newdir)
 305 {
 306         if (newdir[0] == '/') {
 307                 /* The -C /foo -C /bar case; dump first one. */
 308                 free(bsdtar->pending_chdir);
 309                 bsdtar->pending_chdir = NULL;
 310         }
 311         if (bsdtar->pending_chdir == NULL)
 312                 /* Easy case: no previously-saved dir. */
 313                 bsdtar->pending_chdir = strdup(newdir);
 314         else {
 315                 /* The -C /foo -C bar case; concatenate */
 316                 char *old_pending = bsdtar->pending_chdir;
 317                 size_t old_len = strlen(old_pending);
 318                 bsdtar->pending_chdir = malloc(old_len + strlen(newdir) + 2);
 319                 if (old_pending[old_len - 1] == '/')
 320                         old_pending[old_len - 1] = '\0';
 321                 if (bsdtar->pending_chdir != NULL)
 322                         sprintf(bsdtar->pending_chdir, "%s/%s",
 323                             old_pending, newdir);
 324                 free(old_pending);
 325         }
 326         if (bsdtar->pending_chdir == NULL)
 327                 bsdtar_errc(bsdtar, 1, errno, "No memory");
 328 }
 329
 330 void
 331 do_chdir(struct bsdtar *bsdtar)
 332 {
 333         if (bsdtar->pending_chdir == NULL)
 334                 return;
 335
 336         if (chdir(bsdtar->pending_chdir) != 0) {
 337                 bsdtar_errc(bsdtar, 1, 0, "could not chdir to '%s'\n",
 338                     bsdtar->pending_chdir);
 339         }
 340         free(bsdtar->pending_chdir);
 341         bsdtar->pending_chdir = NULL;
 342 }
 343
 344 /*
 345  * Handle --strip-components and any future path-rewriting options.
 346  * Returns non-zero if the pathname should not be extracted.
 347  *
 348  * TODO: Support pax-style regex path rewrites.
 349  */
 350 int
 351 edit_pathname(struct bsdtar *bsdtar, struct archive_entry *entry)
 352 {
 353         const char *name = archive_entry_pathname(entry);
 354
 355         /* Strip leading dir names as per --strip-components option. */
 356         if (bsdtar->strip_components > 0) {
 357                 int r = bsdtar->strip_components;
 358                 const char *p = name;
 359
 360                 while (r > 0) {
 361                         switch (*p++) {
 362                         case '/':
 363                                 r--;
 364                                 name = p;
 365                                 break;
 366                         case '\0':
 367                                 /* Path is too short, skip it. */
 368                                 return (1);
 369                         }
 370                 }
 371         }
 372
 373         /* Strip redundant leading '/' characters. */
 374         while (name[0] == '/' && name[1] == '/')
 375                 name++;
 376
 377         /* Strip leading '/' unless user has asked us not to. */
 378         if (name[0] == '/' && !bsdtar->option_absolute_paths) {
 379                 /* Generate a warning the first time this happens. */
 380                 if (!bsdtar->warned_lead_slash) {
 381                         bsdtar_warnc(bsdtar, 0,
 382                             "Removing leading '/' from member names");
 383                         bsdtar->warned_lead_slash = 1;
 384                 }
 385                 name++;
 386                 /* Special case: Stripping leading '/' from "/" yields ".". */
 387                 if (*name == '\0')
 388                         name = ".";
 389         }
 390
 391         /* Safely replace name in archive_entry. */
 392         if (name != archive_entry_pathname(entry)) {
 393                 char *q = strdup(name);
 394                 archive_entry_copy_pathname(entry, q);
 395                 free(q);
 396         }
 397         return (0);
 398 }
 399
 400 /*
 401  * Like strcmp(), but try to be a little more aware of the fact that
 402  * we're comparing two paths.  Right now, it just handles leading
 403  * "./" and trailing '/' specially, so that "a/b/" == "./a/b"
 404  *
 405  * TODO: Make this better, so that "./a//b/./c/" == "a/b/c"
 406  * TODO: After this works, push it down into libarchive.
 407  * TODO: Publish the path normalization routines in libarchive so
 408  * that bsdtar can normalize paths and use fast strcmp() instead
 409  * of this.
 410  */
 411
 412 int
 413 pathcmp(const char *a, const char *b)
 414 {
 415         /* Skip leading './' */
 416         if (a[0] == '.' && a[1] == '/' && a[2] != '\0')
 417                 a += 2;
 418         if (b[0] == '.' && b[1] == '/' && b[2] != '\0')
 419                 b += 2;
 420         /* Find the first difference, or return (0) if none. */
 421         while (*a == *b) {
 422                 if (*a == '\0')
 423                         return (0);
 424                 a++;
 425                 b++;
 426         }
 427         /*
 428          * If one ends in '/' and the other one doesn't,
 429          * they're the same.
 430          */
 431         if (a[0] == '/' && a[1] == '\0' && b[0] == '\0')
 432                 return (0);
 433         if (a[0] == '\0' && b[0] == '/' && b[1] == '\0')
 434                 return (0);
 435         /* They're really different, return the correct sign. */
 436         return (*(const unsigned char *)a - *(const unsigned char *)b);
 437 }