src/pathchk.c

   1 /* pathchk -- check whether file names are valid or portable
   2    Copyright (C) 1991-2023 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation, either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  16
  17 #include <config.h>
  18 #include <stdio.h>
  19 #include <getopt.h>
  20 #include <sys/types.h>
  21 #include <wchar.h>
  22
  23 #include "system.h"
  24 #include "error.h"
  25 #include "quote.h"
  26
  27 /* The official name of this program (e.g., no 'g' prefix).  */
  28 #define PROGRAM_NAME "pathchk"
  29
  30 #define AUTHORS \
  31   proper_name ("Paul Eggert"), \
  32   proper_name ("David MacKenzie"), \
  33   proper_name ("Jim Meyering")
  34
  35 #ifndef _POSIX_PATH_MAX
  36 # define _POSIX_PATH_MAX 256
  37 #endif
  38 #ifndef _POSIX_NAME_MAX
  39 # define _POSIX_NAME_MAX 14
  40 #endif
  41
  42 #ifdef _XOPEN_NAME_MAX
  43 # define NAME_MAX_MINIMUM _XOPEN_NAME_MAX
  44 #else
  45 # define NAME_MAX_MINIMUM _POSIX_NAME_MAX
  46 #endif
  47 #ifdef _XOPEN_PATH_MAX
  48 # define PATH_MAX_MINIMUM _XOPEN_PATH_MAX
  49 #else
  50 # define PATH_MAX_MINIMUM _POSIX_PATH_MAX
  51 #endif
  52
  53 #if ! (HAVE_PATHCONF && defined _PC_NAME_MAX && defined _PC_PATH_MAX)
  54 # ifndef _PC_NAME_MAX
  55 #  define _PC_NAME_MAX 0
  56 #  define _PC_PATH_MAX 1
  57 # endif
  58 # ifndef pathconf
  59 #  define pathconf(file, flag) \
  60      (flag == _PC_NAME_MAX ? NAME_MAX_MINIMUM : PATH_MAX_MINIMUM)
  61 # endif
  62 #endif
  63
  64 static bool validate_file_name (char *, bool, bool);
  65
  66 /* For long options that have no equivalent short option, use a
  67    non-character as a pseudo short option, starting with CHAR_MAX + 1.  */
  68 enum
  69 {
  70   PORTABILITY_OPTION = CHAR_MAX + 1
  71 };
  72
  73 static struct option const longopts[] =
  74 {
  75   {"portability", no_argument, NULL, PORTABILITY_OPTION},
  76   {GETOPT_HELP_OPTION_DECL},
  77   {GETOPT_VERSION_OPTION_DECL},
  78   {NULL, 0, NULL, 0}
  79 };
  80
  81 void
  82 usage (int status)
  83 {
  84   if (status != EXIT_SUCCESS)
  85     emit_try_help ();
  86   else
  87     {
  88       printf (_("Usage: %s [OPTION]... NAME...\n"), program_name);
  89       fputs (_("\
  90 Diagnose invalid or unportable file names.\n\
  91 \n\
  92   -p                  check for most POSIX systems\n\
  93   -P                  check for empty names and leading \"-\"\n\
  94       --portability   check for all POSIX systems (equivalent to -p -P)\n\
  95 "), stdout);
  96       fputs (HELP_OPTION_DESCRIPTION, stdout);
  97       fputs (VERSION_OPTION_DESCRIPTION, stdout);
  98       emit_ancillary_info (PROGRAM_NAME);
  99     }
 100   exit (status);
 101 }
 102
 103 int
 104 main (int argc, char **argv)
 105 {
 106   bool ok = true;
 107   bool check_basic_portability = false;
 108   bool check_extra_portability = false;
 109   int optc;
 110
 111   initialize_main (&argc, &argv);
 112   set_program_name (argv[0]);
 113   setlocale (LC_ALL, "");
 114   bindtextdomain (PACKAGE, LOCALEDIR);
 115   textdomain (PACKAGE);
 116
 117   atexit (close_stdout);
 118
 119   while ((optc = getopt_long (argc, argv, "+pP", longopts, NULL)) != -1)
 120     {
 121       switch (optc)
 122         {
 123         case PORTABILITY_OPTION:
 124           check_basic_portability = true;
 125           check_extra_portability = true;
 126           break;
 127
 128         case 'p':
 129           check_basic_portability = true;
 130           break;
 131
 132         case 'P':
 133           check_extra_portability = true;
 134           break;
 135
 136         case_GETOPT_HELP_CHAR;
 137
 138         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 139
 140         default:
 141           usage (EXIT_FAILURE);
 142         }
 143     }
 144
 145   if (optind == argc)
 146     {
 147       error (0, 0, _("missing operand"));
 148       usage (EXIT_FAILURE);
 149     }
 150
 151   for (; optind < argc; ++optind)
 152     ok &= validate_file_name (argv[optind],
 153                               check_basic_portability, check_extra_portability);
 154
 155   return ok ? EXIT_SUCCESS : EXIT_FAILURE;
 156 }
 157
 158 /* If FILE contains a component with a leading "-", report an error
 159    and return false; otherwise, return true.  */
 160
 161 static bool
 162 no_leading_hyphen (char const *file)
 163 {
 164   char const *p;
 165
 166   for (p = file;  (p = strchr (p, '-'));  p++)
 167     if (p == file || p[-1] == '/')
 168       {
 169         error (0, 0, _("leading '-' in a component of file name %s"),
 170                quoteaf (file));
 171         return false;
 172       }
 173
 174   return true;
 175 }
 176
 177 /* If FILE (of length FILELEN) contains only portable characters,
 178    return true, else report an error and return false.  */
 179
 180 static bool
 181 portable_chars_only (char const *file, size_t filelen)
 182 {
 183   size_t validlen = strspn (file,
 184                             ("/"
 185                              "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
 186                              "abcdefghijklmnopqrstuvwxyz"
 187                              "0123456789._-"));
 188   char const *invalid = file + validlen;
 189
 190   if (*invalid)
 191     {
 192       mbstate_t mbstate = { 0, };
 193       size_t charlen = mbrlen (invalid, filelen - validlen, &mbstate);
 194       error (0, 0,
 195              _("nonportable character %s in file name %s"),
 196              quotearg_n_style_mem (1, locale_quoting_style, invalid,
 197                                    (charlen <= MB_LEN_MAX ? charlen : 1)),
 198              quoteaf_n (0, file));
 199       return false;
 200     }
 201
 202   return true;
 203 }
 204
 205 /* Return the address of the start of the next file name component in F.  */
 206
 207 ATTRIBUTE_PURE
 208 static char *
 209 component_start (char *f)
 210 {
 211   while (*f == '/')
 212     f++;
 213   return f;
 214 }
 215
 216 /* Return the size of the file name component F.  F must be nonempty.  */
 217
 218 ATTRIBUTE_PURE
 219 static size_t
 220 component_len (char const *f)
 221 {
 222   size_t len;
 223   for (len = 1; f[len] != '/' && f[len]; len++)
 224     continue;
 225   return len;
 226 }
 227
 228 /* Make sure that
 229    strlen (FILE) <= PATH_MAX
 230    && strlen (each-existing-directory-in-FILE) <= NAME_MAX
 231
 232    If CHECK_BASIC_PORTABILITY is true, compare against _POSIX_PATH_MAX and
 233    _POSIX_NAME_MAX instead, and make sure that FILE contains no
 234    characters not in the POSIX portable filename character set, which
 235    consists of A-Z, a-z, 0-9, ., _, - (plus / for separators).
 236
 237    If CHECK_BASIC_PORTABILITY is false, make sure that all leading directories
 238    along FILE that exist are searchable.
 239
 240    If CHECK_EXTRA_PORTABILITY is true, check that file name components do not
 241    begin with "-".
 242
 243    If either CHECK_BASIC_PORTABILITY or CHECK_EXTRA_PORTABILITY is true,
 244    check that the file name is not empty.
 245
 246    Return true if all of these tests are successful, false if any fail.  */
 247
 248 static bool
 249 validate_file_name (char *file, bool check_basic_portability,
 250                     bool check_extra_portability)
 251 {
 252   size_t filelen = strlen (file);
 253
 254   /* Start of file name component being checked.  */
 255   char *start;
 256
 257   /* True if component lengths need to be checked.  */
 258   bool check_component_lengths;
 259
 260   /* True if the file is known to exist.  */
 261   bool file_exists = false;
 262
 263   if (check_extra_portability && ! no_leading_hyphen (file))
 264     return false;
 265
 266   if ((check_basic_portability || check_extra_portability)
 267       && filelen == 0)
 268     {
 269       /* Fail, since empty names are not portable.  As of
 270          2005-01-06 POSIX does not address whether "pathchk -p ''"
 271          should (or is allowed to) fail, so this is not a
 272          conformance violation.  */
 273       error (0, 0, _("empty file name"));
 274       return false;
 275     }
 276
 277   if (check_basic_portability)
 278     {
 279       if (! portable_chars_only (file, filelen))
 280         return false;
 281     }
 282   else
 283     {
 284       /* Check whether a file name component is in a directory that
 285          is not searchable, or has some other serious problem.
 286          POSIX does not allow "" as a file name, but some non-POSIX
 287          hosts do (as an alias for "."), so allow "" if lstat does.  */
 288
 289       struct stat st;
 290       if (lstat (file, &st) == 0)
 291         file_exists = true;
 292       else if (errno != ENOENT || filelen == 0)
 293         {
 294           error (0, errno, "%s", quotef (file));
 295           return false;
 296         }
 297     }
 298
 299   if (check_basic_portability
 300       || (! file_exists && PATH_MAX_MINIMUM <= filelen))
 301     {
 302       size_t maxsize;
 303
 304       if (check_basic_portability)
 305         maxsize = _POSIX_PATH_MAX;
 306       else
 307         {
 308           long int size;
 309           char const *dir = (*file == '/' ? "/" : ".");
 310           errno = 0;
 311           size = pathconf (dir, _PC_PATH_MAX);
 312           if (size < 0 && errno != 0)
 313             {
 314               error (0, errno,
 315                      _("%s: unable to determine maximum file name length"),
 316                      dir);
 317               return false;
 318             }
 319           maxsize = MIN (size, SSIZE_MAX);
 320         }
 321
 322       if (maxsize <= filelen)
 323         {
 324           unsigned long int len = filelen;
 325           unsigned long int maxlen = maxsize - 1;
 326           error (0, 0, _("limit %lu exceeded by length %lu of file name %s"),
 327                  maxlen, len, quoteaf (file));
 328           return false;
 329         }
 330     }
 331
 332   /* Check whether pathconf (..., _PC_NAME_MAX) can be avoided, i.e.,
 333      whether all file name components are so short that they are valid
 334      in any file system on this platform.  If CHECK_BASIC_PORTABILITY, though,
 335      it's more convenient to check component lengths below.  */
 336
 337   check_component_lengths = check_basic_portability;
 338   if (! check_component_lengths && ! file_exists)
 339     {
 340       for (start = file; *(start = component_start (start)); )
 341         {
 342           size_t length = component_len (start);
 343
 344           if (NAME_MAX_MINIMUM < length)
 345             {
 346               check_component_lengths = true;
 347               break;
 348             }
 349
 350           start += length;
 351         }
 352     }
 353
 354   if (check_component_lengths)
 355     {
 356       /* The limit on file name components for the current component.
 357          This defaults to NAME_MAX_MINIMUM, for the sake of non-POSIX
 358          systems (NFS, say?) where pathconf fails on "." or "/" with
 359          errno == ENOENT.  */
 360       size_t name_max = NAME_MAX_MINIMUM;
 361
 362       /* If nonzero, the known limit on file name components.  */
 363       size_t known_name_max = (check_basic_portability ? _POSIX_NAME_MAX : 0);
 364
 365       for (start = file; *(start = component_start (start)); )
 366         {
 367           size_t length;
 368
 369           if (known_name_max)
 370             name_max = known_name_max;
 371           else
 372             {
 373               long int len;
 374               char const *dir = (start == file ? "." : file);
 375               char c = *start;
 376               errno = 0;
 377               *start = '\0';
 378               len = pathconf (dir, _PC_NAME_MAX);
 379               *start = c;
 380               if (0 <= len)
 381                 name_max = MIN (len, SSIZE_MAX);
 382               else
 383                 switch (errno)
 384                   {
 385                   case 0:
 386                     /* There is no limit.  */
 387                     name_max = SIZE_MAX;
 388                     break;
 389
 390                   case ENOENT:
 391                     /* DIR does not exist; use its parent's maximum.  */
 392                     known_name_max = name_max;
 393                     break;
 394
 395                   default:
 396                     *start = '\0';
 397                     error (0, errno, "%s", quotef (dir));
 398                     *start = c;
 399                     return false;
 400                   }
 401             }
 402
 403           length = component_len (start);
 404
 405           if (name_max < length)
 406             {
 407               unsigned long int len = length;
 408               unsigned long int maxlen = name_max;
 409               char c = start[len];
 410               start[len] = '\0';
 411               error (0, 0,
 412                      _("limit %lu exceeded by length %lu "
 413                        "of file name component %s"),
 414                      maxlen, len, quote (start));
 415               start[len] = c;
 416               return false;
 417             }
 418
 419           start += length;
 420         }
 421     }
 422
 423   return true;
 424 }