find/ftsfind.c

   1 /* find -- search for files in a directory hierarchy (fts version)
   2    Copyright (C) 1990, 91, 92, 93, 94, 2000,
   3                  2003, 2004, 2005 Free Software Foundation, Inc.
   4
   5    This program is free software; you can redistribute it and/or modify
   6    it under the terms of the GNU General Public License as published by
   7    the Free Software Foundation; either version 2, or (at your option)
   8    any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program; if not, write to the Free Software
  17    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301,
  18    USA.*/
  19
  20 /* This file was written by James Youngman, based on find.c.
  21
  22    GNU find was written by Eric Decker <cire@cisco.com>,
  23    with enhancements by David MacKenzie <djm@gnu.org>,
  24    Jay Plett <jay@silence.princeton.nj.us>,
  25    and Tim Wood <axolotl!tim@toad.com>.
  26    The idea for -print0 and xargs -0 came from
  27    Dan Bernstein <brnstnd@kramden.acf.nyu.edu>.
  28 */
  29
  30
  31 #include "defs.h"
  32
  33
  34 #define USE_SAFE_CHDIR 1
  35 #undef  STAT_MOUNTPOINTS
  36
  37
  38 #include <errno.h>
  39 #include <assert.h>
  40
  41 #ifdef HAVE_FCNTL_H
  42 #include <fcntl.h>
  43 #else
  44 #include <sys/file.h>
  45 #endif
  46
  47
  48 #include "../gnulib/lib/xalloc.h"
  49 #include "closeout.h"
  50 #include <modetype.h>
  51 #include "quotearg.h"
  52 #include "quote.h"
  53 #include "fts_.h"
  54
  55 #ifdef HAVE_LOCALE_H
  56 #include <locale.h>
  57 #endif
  58
  59 #if ENABLE_NLS
  60 # include <libintl.h>
  61 # define _(Text) gettext (Text)
  62 #else
  63 # define _(Text) Text
  64 #define textdomain(Domain)
  65 #define bindtextdomain(Package, Directory)
  66 #endif
  67 #ifdef gettext_noop
  68 # define N_(String) gettext_noop (String)
  69 #else
  70 /* See locate.c for explanation as to why not use (String) */
  71 # define N_(String) String
  72 #endif
  73
  74
  75 #ifdef STAT_MOUNTPOINTS
  76 static void init_mounted_dev_list(void);
  77 #endif
  78 \f
  79 /* We have encountered an error which shoudl affect the exit status.
  80  * This is normally used to change the exit status from 0 to 1.
  81  * However, if the exit status is already 2 for example, we don't want to
  82  * reduce it to 1.
  83  */
  84 static void
  85 error_severity(int level)
  86 {
  87   if (state.exit_status < level)
  88     state.exit_status = level;
  89 }
  90 \f
  91
  92 #define STRINGIFY(X) #X
  93 #define HANDLECASE(N) case N: return #N;
  94
  95 static char *
  96 get_fts_info_name(int info)
  97 {
  98   static char buf[10];
  99   switch (info)
 100     {
 101       HANDLECASE(FTS_D);
 102       HANDLECASE(FTS_DC);
 103       HANDLECASE(FTS_DEFAULT);
 104       HANDLECASE(FTS_DNR);
 105       HANDLECASE(FTS_DOT);
 106       HANDLECASE(FTS_DP);
 107       HANDLECASE(FTS_ERR);
 108       HANDLECASE(FTS_F);
 109       HANDLECASE(FTS_INIT);
 110       HANDLECASE(FTS_NS);
 111       HANDLECASE(FTS_NSOK);
 112       HANDLECASE(FTS_SL);
 113       HANDLECASE(FTS_SLNONE);
 114       HANDLECASE(FTS_W);
 115     default:
 116       sprintf(buf, "[%d]", info);
 117       return buf;
 118     }
 119 }
 120 \f
 121 static void
 122 visit(FTS *p, FTSENT *ent, struct stat *pstat)
 123 {
 124   struct predicate *eval_tree;
 125
 126   state.curdepth = ent->fts_level;
 127   state.have_stat = (ent->fts_info != FTS_NS) && (ent->fts_info != FTS_NSOK);
 128   state.rel_pathname = ent->fts_accpath;
 129
 130   /* Apply the predicates to this path. */
 131   eval_tree = get_eval_tree();
 132   (*(eval_tree)->pred_func)(ent->fts_path, pstat, eval_tree);
 133
 134   /* Deal with any side effects of applying the predicates. */
 135   if (state.stop_at_current_level)
 136     {
 137       fts_set(p, ent, FTS_SKIP);
 138     }
 139 }
 140 \f
 141 static const char*
 142 partial_quotearg_n(int n, char *s, size_t len, enum quoting_style style)
 143 {
 144   if (0 == len)
 145     {
 146       return quotearg_n_style(n, style, "");
 147     }
 148   else
 149     {
 150       char saved;
 151       const char *result;
 152
 153       saved = s[len];
 154       s[len] = 0;
 155       result = quotearg_n_style(n, style, s);
 156       s[len] = saved;
 157       return result;
 158     }
 159 }
 160
 161 \f
 162 /* We've detected a filesystem loop.   This is caused by one of
 163  * two things:
 164  *
 165  * 1. Option -L is in effect and we've hit a symbolic link that
 166  *    points to an ancestor.  This is harmless.  We won't traverse the
 167  *    symbolic link.
 168  *
 169  * 2. We have hit a real cycle in the directory hierarchy.  In this
 170  *    case, we issue a diagnostic message (POSIX requires this) and we
 171  *    skip that directory entry.
 172  */
 173 static void
 174 issue_loop_warning(FTSENT * ent)
 175 {
 176   if (S_ISLNK(ent->fts_statp->st_mode))
 177     {
 178       error(0, 0,
 179             _("Symbolic link %s is part of a loop in the directory hierarchy; we have already visited the directory to which it points."),
 180             quotearg_n_style(0, locale_quoting_style, ent->fts_path));
 181     }
 182   else
 183     {
 184       /* We have found an infinite loop.  POSIX requires us to
 185        * issue a diagnostic.  Usually we won't get to here
 186        * because when the leaf optimisation is on, it will cause
 187        * the subdirectory to be skipped.  If /a/b/c/d is a hard
 188        * link to /a/b, then the link count of /a/b/c is 2,
 189        * because the ".." entry of /b/b/c/d points to /a, not
 190        * to /a/b/c.
 191        */
 192       error(0, 0,
 193             _("Filesystem loop detected; "
 194               "%s is part of the same filesystem loop as %s."),
 195             quotearg_n_style(0, locale_quoting_style, ent->fts_path),
 196             partial_quotearg_n(1,
 197                                ent->fts_cycle->fts_path,
 198                                ent->fts_cycle->fts_pathlen,
 199                                locale_quoting_style));
 200     }
 201 }
 202 \f
 203 /*
 204  * Return true if NAME corresponds to a file which forms part of a
 205  * symbolic link loop.  The command
 206  *      rm -f a b; ln -s a b; ln -s b a
 207  * produces such a loop.
 208  */
 209 static boolean
 210 symlink_loop(const char *name)
 211 {
 212   struct stat stbuf;
 213   int rv;
 214   if (following_links())
 215     rv = stat(name, &stbuf);
 216   else
 217     rv = lstat(name, &stbuf);
 218   return (0 != rv) && (ELOOP == errno);
 219 }
 220
 221 \f
 222 static void
 223 consider_visiting(FTS *p, FTSENT *ent)
 224 {
 225   struct stat statbuf;
 226   mode_t mode;
 227   int ignore, isdir;
 228
 229   if (options.debug_options & DebugSearch)
 230     fprintf(stderr,
 231             "consider_visiting: end->fts_info=%s, ent->fts_path=%s\n",
 232             get_fts_info_name(ent->fts_info),
 233             quotearg_n_style(0, locale_quoting_style, ent->fts_path));
 234
 235   /* Cope with various error conditions. */
 236   if (ent->fts_info == FTS_ERR
 237       || ent->fts_info == FTS_NS
 238       || ent->fts_info == FTS_DNR)
 239     {
 240       error(0, ent->fts_errno, ent->fts_path);
 241       error_severity(1);
 242       return;
 243     }
 244   else if (ent->fts_info == FTS_DC)
 245     {
 246       issue_loop_warning(ent);
 247       error_severity(1);
 248       return;
 249     }
 250   else if (ent->fts_info == FTS_SLNONE)
 251     {
 252       /* fts_read() claims that ent->fts_accpath is a broken symbolic
 253        * link.  That would be fine, but if this is part of a symbolic
 254        * link loop, we diagnose the problem and also ensure that the
 255        * eventual return value is nonzero.   Note that while the path
 256        * we stat is local (fts_accpath), we print the fill path name
 257        * of the file (fts_path) in the error message.
 258        */
 259       if (symlink_loop(ent->fts_accpath))
 260         {
 261           error(0, ELOOP, ent->fts_path);
 262           error_severity(1);
 263           return;
 264         }
 265     }
 266
 267   /* Not an error, cope with the usual cases. */
 268   if (ent->fts_info == FTS_NSOK)
 269     {
 270       assert(!state.have_stat);
 271       assert(!state.have_type);
 272       state.type = mode = 0;
 273     }
 274   else
 275     {
 276       state.have_stat = true;
 277       state.have_type = true;
 278       statbuf = *(ent->fts_statp);
 279       state.type = mode = statbuf.st_mode;
 280     }
 281
 282   if (0 == ent->fts_level && (0u == state.starting_path_length))
 283     state.starting_path_length = ent->fts_pathlen;
 284
 285   if (mode)
 286     {
 287       if (!digest_mode(mode, ent->fts_path, ent->fts_name, &statbuf, 0))
 288         return;
 289     }
 290
 291   /* examine this item. */
 292   ignore = 0;
 293   isdir = S_ISDIR(statbuf.st_mode)
 294     || (FTS_D  == ent->fts_info)
 295     || (FTS_DP == ent->fts_info)
 296     || (FTS_DC == ent->fts_info);
 297
 298   if (isdir && (ent->fts_info == FTS_NSOK))
 299     {
 300       /* This is a directory, but fts did not stat it, so
 301        * presumably would not be planning to search its
 302        * children.  Force a stat of the file so that the
 303        * children can be checked.
 304        */
 305       fts_set(p, ent, FTS_AGAIN);
 306       return;
 307     }
 308
 309   if (options.maxdepth >= 0 && (ent->fts_level > options.maxdepth))
 310     {
 311       ignore = 1;
 312       fts_set(p, ent, FTS_SKIP);
 313     }
 314   else if ( (ent->fts_info == FTS_D) && !options.do_dir_first )
 315     {
 316       /* this is the preorder visit, but user said -depth */
 317       ignore = 1;
 318     }
 319   else if ( (ent->fts_info == FTS_DP) && options.do_dir_first )
 320     {
 321       /* this is the postorder visit, but user didn't say -depth */
 322       ignore = 1;
 323     }
 324   else if (ent->fts_level < options.mindepth)
 325     {
 326       ignore = 1;
 327     }
 328
 329   if (!ignore)
 330     {
 331       visit(p, ent, &statbuf);
 332     }
 333
 334
 335   if (ent->fts_info == FTS_DP)
 336     {
 337       /* we're leaving a directory. */
 338       state.stop_at_current_level = false;
 339       complete_pending_execdirs(get_eval_tree());
 340     }
 341 }
 342
 343 \f
 344 static void
 345 find(char *arg)
 346 {
 347   char * arglist[2];
 348   int ftsoptions;
 349   FTS *p;
 350   FTSENT *ent;
 351
 352
 353   arglist[0] = arg;
 354   arglist[1] = NULL;
 355
 356   ftsoptions = FTS_NOSTAT;
 357   switch (options.symlink_handling)
 358     {
 359     case SYMLINK_ALWAYS_DEREF:
 360       ftsoptions |= FTS_COMFOLLOW|FTS_LOGICAL;
 361       break;
 362
 363     case SYMLINK_DEREF_ARGSONLY:
 364       ftsoptions |= FTS_COMFOLLOW;
 365       break;
 366
 367     case SYMLINK_NEVER_DEREF:
 368       ftsoptions |= FTS_PHYSICAL;
 369       break;
 370     }
 371
 372   if (options.stay_on_filesystem)
 373     ftsoptions |= FTS_XDEV;
 374
 375   p = fts_open(arglist, ftsoptions, NULL);
 376   if (NULL == p)
 377     {
 378       error (0, errno,
 379              _("cannot search %s"),
 380              quotearg_n_style(0, locale_quoting_style, arg));
 381     }
 382   else
 383     {
 384       while ( (ent=fts_read(p)) != NULL )
 385         {
 386           state.have_stat = false;
 387           state.have_type = false;
 388           state.type = 0;
 389
 390           consider_visiting(p, ent);
 391         }
 392       fts_close(p);
 393       p = NULL;
 394     }
 395 }
 396
 397
 398 static void
 399 process_all_startpoints(int argc, char *argv[])
 400 {
 401   int i;
 402
 403   /* figure out how many start points there are */
 404   for (i = 0; i < argc && !looks_like_expression(argv[i], true); i++)
 405     {
 406       find(argv[i]);
 407     }
 408
 409   if (i == 0)
 410     {
 411       /*
 412        * We use a temporary variable here because some actions modify
 413        * the path temporarily.  Hence if we use a string constant,
 414        * we get a coredump.  The best example of this is if we say
 415        * "find -printf %H" (note, not "find . -printf %H").
 416        */
 417       char defaultpath[2] = ".";
 418       find(defaultpath);
 419     }
 420 }
 421
 422
 423
 424 \f
 425 int
 426 main (int argc, char **argv)
 427 {
 428   int end_of_leading_options = 0; /* First arg after any -H/-L etc. */
 429   struct predicate *eval_tree;
 430
 431   program_name = argv[0];
 432   state.exit_status = 0;
 433
 434
 435   /* Set the option defaults before we do the the locale
 436    * initialisation as check_nofollow() needs to be executed in the
 437    * POSIX locale.
 438    */
 439   set_option_defaults(&options);
 440
 441 #ifdef HAVE_SETLOCALE
 442   setlocale (LC_ALL, "");
 443 #endif
 444
 445   bindtextdomain (PACKAGE, LOCALEDIR);
 446   textdomain (PACKAGE);
 447   atexit (close_stdout);
 448
 449   /* Check for -P, -H or -L options.  Also -D and -O, which are
 450    * both GNU extensions.
 451    */
 452   end_of_leading_options = process_leading_options(argc, argv);
 453
 454   if (options.debug_options & DebugStat)
 455     options.xstat = debug_stat;
 456
 457 #ifdef DEBUG
 458   fprintf (stderr, "cur_day_start = %s", ctime (&options.cur_day_start));
 459 #endif /* DEBUG */
 460
 461
 462   /* We are now processing the part of the "find" command line
 463    * after the -H/-L options (if any).
 464    */
 465   eval_tree = build_expression_tree(argc, argv, end_of_leading_options);
 466
 467   /* safely_chdir() needs to check that it has ended up in the right place.
 468    * To avoid bailing out when something gets automounted, it checks if
 469    * the target directory appears to have had a directory mounted on it as
 470    * we chdir()ed.  The problem with this is that in order to notice that
 471    * a filesystem was mounted, we would need to lstat() all the mount points.
 472    * That strategy loses if our machine is a client of a dead NFS server.
 473    *
 474    * Hence if safely_chdir() and wd_sanity_check() can manage without needing
 475    * to know the mounted device list, we do that.
 476    */
 477   if (!options.open_nofollow_available)
 478     {
 479 #ifdef STAT_MOUNTPOINTS
 480       init_mounted_dev_list();
 481 #endif
 482     }
 483
 484
 485   starting_desc = open (".", O_RDONLY);
 486   if (0 <= starting_desc && fchdir (starting_desc) != 0)
 487     {
 488       close (starting_desc);
 489       starting_desc = -1;
 490     }
 491   if (starting_desc < 0)
 492     {
 493       starting_dir = xgetcwd ();
 494       if (! starting_dir)
 495         error (1, errno, _("cannot get current directory"));
 496     }
 497
 498
 499   process_all_startpoints(argc-end_of_leading_options, argv+end_of_leading_options);
 500
 501   /* If "-exec ... {} +" has been used, there may be some
 502    * partially-full command lines which have been built,
 503    * but which are not yet complete.   Execute those now.
 504    */
 505   cleanup();
 506   return state.exit_status;
 507 }
 508
 509 boolean is_fts_enabled()
 510 {
 511   /* this version of find (i.e. this main()) uses fts. */
 512   return true;
 513 }