src/md5sum.c

   1 /* Compute MD5 or SHA1 checksum of files or strings
   2    Copyright (C) 1995-2000 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 2, or (at your option)
   7    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, write to the Free Software Foundation,
  16    Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  17
  18 /* Written by Ulrich Drepper <drepper@gnu.ai.mit.edu>.  */
  19
  20 #include <config.h>
  21
  22 #include <getopt.h>
  23 #include <stdio.h>
  24 #include <sys/types.h>
  25
  26 #include "system.h"
  27
  28 #include "md5.h"
  29 #include "sha.h"
  30 #include "checksum.h"
  31 #include "getline.h"
  32 #include "closeout.h"
  33 #include "error.h"
  34
  35 /* The official name of this program (e.g., no `g' prefix).  */
  36 #define PROGRAM_NAME (algorithm == ALG_MD5 ? "md5sum" : "shasum")
  37
  38 #define AUTHORS "Ulrich Drepper and Scott Miller"
  39
  40 /* Most systems do not distinguish between external and internal
  41    text representations.  */
  42 /* FIXME: This begs for an autoconf test.  */
  43 #if O_BINARY
  44 # define OPENOPTS(BINARY) ((BINARY) != 0 ? TEXT1TO1 : TEXTCNVT)
  45 # define TEXT1TO1 "rb"
  46 # define TEXTCNVT "r"
  47 #else
  48 # if defined VMS
  49 #  define OPENOPTS(BINARY) ((BINARY) != 0 ? TEXT1TO1 : TEXTCNVT)
  50 #  define TEXT1TO1 "rb", "ctx=stm"
  51 #  define TEXTCNVT "r", "ctx=stm"
  52 # else
  53 #  if UNIX || __UNIX__ || unix || __unix__ || _POSIX_VERSION
  54 #   define OPENOPTS(BINARY) "r"
  55 #  else
  56     /* The following line is intended to evoke an error.
  57        Using #error is not portable enough.  */
  58     "Cannot determine system type."
  59 #  endif
  60 # endif
  61 #endif
  62
  63
  64 #define DIGEST_TYPE_STRING(Alg) ((Alg) == ALG_MD5 ? "MD5" : "SHA1")
  65 #define DIGEST_STREAM(Alg) ((Alg) == ALG_MD5 ? md5_stream : sha_stream)
  66
  67 #define DIGEST_BITS(Alg) ((Alg) == ALG_MD5 ? 128 : 160)
  68 #define DIGEST_HEX_BYTES(Alg) (DIGEST_BITS (Alg) / 4)
  69 #define DIGEST_BIN_BYTES(Alg) (DIGEST_BITS (Alg) / 8)
  70
  71 #define MAX_DIGEST_BIN_BYTES MAX (DIGEST_BIN_BYTES (ALG_MD5), \
  72                                   DIGEST_BIN_BYTES (ALG_SHA1))
  73
  74 /* The minimum length of a valid digest line.  This length does
  75    not include any newline character at the end of a line.  */
  76 #define MIN_DIGEST_LINE_LENGTH(Alg) \
  77   (DIGEST_HEX_BYTES (Alg) /* length of hexadecimal message digest */ \
  78    + 2 /* blank and binary indicator */ \
  79    + 1 /* minimum filename length */ )
  80
  81 /* Nonzero if any of the files read were the standard input. */
  82 static int have_read_stdin;
  83
  84 /* The minimum length of a valid checksum line for the selected algorithm.  */
  85 static int min_digest_line_length;
  86
  87 /* Set to the length of a digest hex string for the selected algorithm.  */
  88 static size_t digest_hex_bytes;
  89
  90 /* With --check, don't generate any output.
  91    The exit code indicates success or failure.  */
  92 static int status_only = 0;
  93
  94 /* With --check, print a message to standard error warning about each
  95    improperly formatted checksum line.  */
  96 static int warn = 0;
  97
  98 /* Declared and set via one of the wrapper .c files.  */
  99 /* int algorithm = ALG_UNSPECIFIED; */
 100
 101 /* The name this program was run with.  */
 102 char *program_name;
 103
 104 static const struct option long_options[] =
 105 {
 106   { "binary", no_argument, 0, 'b' },
 107   { "check", no_argument, 0, 'c' },
 108   { "status", no_argument, 0, 2 },
 109   { "string", required_argument, 0, 1 },
 110   { "text", no_argument, 0, 't' },
 111   { "warn", no_argument, 0, 'w' },
 112   { GETOPT_HELP_OPTION_DECL },
 113   { GETOPT_VERSION_OPTION_DECL },
 114   { NULL, 0, NULL, 0 }
 115 };
 116
 117 void
 118 usage (int status)
 119 {
 120   if (status != 0)
 121     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 122              program_name);
 123   else
 124     {
 125       printf (_("\
 126 Usage: %s [OPTION] [FILE]...\n\
 127   or:  %s [OPTION] --check [FILE]\n\
 128 Print or check %s (%d-bit) checksums.\n\
 129 With no FILE, or when FILE is -, read standard input.\n\
 130 \n\
 131   -b, --binary            read files in binary mode (default on DOS/Windows)\n\
 132   -c, --check             check %s sums against given list\n\
 133   -t, --text              read files in text mode (default)\n\
 134 \n\
 135 The following two options are useful only when verifying checksums:\n\
 136       --status            don't output anything, status code shows success\n\
 137   -w, --warn              warn about improperly formated checksum lines\n\
 138 \n\
 139       --help              display this help and exit\n\
 140       --version           output version information and exit\n\
 141 \n\
 142 The sums are computed as described in %s.  When checking, the input\n\
 143 should be a former output of this program.  The default mode is to print\n\
 144 a line with checksum, a character indicating type (`*' for binary, ` ' for\n\
 145 text), and name for each FILE.\n"),
 146               program_name, program_name,
 147               DIGEST_TYPE_STRING (algorithm),
 148               DIGEST_BITS (algorithm),
 149               DIGEST_TYPE_STRING (algorithm),
 150               (algorithm == ALG_MD5 ? "RFC 1321" : "FIPS-180-1")
 151               );
 152       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
 153     }
 154
 155   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 156 }
 157
 158 static int
 159 split_3 (char *s, size_t s_len, unsigned char **u, int *binary, char **w)
 160 {
 161   size_t i;
 162   int escaped_filename = 0;
 163
 164 #define ISWHITE(c) ((c) == ' ' || (c) == '\t')
 165
 166   i = 0;
 167   while (ISWHITE (s[i]))
 168     ++i;
 169
 170   /* The line must have at least `min_digest_line_length - 1' (or one more, if
 171      the first is a backslash) more characters to contain correct message digest
 172      information.  Ignore this line if it is too short.  */
 173   if (!(s_len - i >= min_digest_line_length
 174         || (s[i] == '\\' && s_len - i >= 1 + min_digest_line_length)))
 175     return 1;
 176
 177   if (s[i] == '\\')
 178     {
 179       ++i;
 180       escaped_filename = 1;
 181     }
 182   *u = (unsigned char *) &s[i];
 183
 184   /* The first field has to be the n-character hexadecimal
 185      representation of the message digest.  If it is not followed
 186      immediately by a white space it's an error.  */
 187   i += digest_hex_bytes;
 188   if (!ISWHITE (s[i]))
 189     return 1;
 190
 191   s[i++] = '\0';
 192
 193   if (s[i] != ' ' && s[i] != '*')
 194     return 1;
 195   *binary = (s[i++] == '*');
 196
 197   /* All characters between the type indicator and end of line are
 198      significant -- that includes leading and trailing white space.  */
 199   *w = &s[i];
 200
 201   if (escaped_filename)
 202     {
 203       /* Translate each `\n' string in the file name to a NEWLINE,
 204          and each `\\' string to a backslash.  */
 205
 206       char *dst = &s[i];
 207
 208       while (i < s_len)
 209         {
 210           switch (s[i])
 211             {
 212             case '\\':
 213               if (i == s_len - 1)
 214                 {
 215                   /* A valid line does not end with a backslash.  */
 216                   return 1;
 217                 }
 218               ++i;
 219               switch (s[i++])
 220                 {
 221                 case 'n':
 222                   *dst++ = '\n';
 223                   break;
 224                 case '\\':
 225                   *dst++ = '\\';
 226                   break;
 227                 default:
 228                   /* Only `\' or `n' may follow a backslash.  */
 229                   return 1;
 230                 }
 231               break;
 232
 233             case '\0':
 234               /* The file name may not contain a NUL.  */
 235               return 1;
 236               break;
 237
 238             default:
 239               *dst++ = s[i++];
 240               break;
 241             }
 242         }
 243       *dst = '\0';
 244     }
 245   return 0;
 246 }
 247
 248 static int
 249 hex_digits (unsigned char const *s)
 250 {
 251   while (*s)
 252     {
 253       if (!ISXDIGIT (*s))
 254         return 0;
 255       ++s;
 256     }
 257   return 1;
 258 }
 259
 260 /* An interface to the function, DIGEST_STREAM, (either md5_stream or sha_stream).
 261    Operate on FILENAME (it may be "-") and put the result in *BIN_RESULT.
 262    Return non-zero upon failure, zero to indicate success.  */
 263
 264 static int
 265 digest_file (const char *filename, int binary, unsigned char *bin_result,
 266            int (*digest_stream)(FILE *, void *))
 267 {
 268   FILE *fp;
 269   int err;
 270
 271   if (STREQ (filename, "-"))
 272     {
 273       have_read_stdin = 1;
 274       fp = stdin;
 275 #if O_BINARY
 276       /* If we need binary reads from a pipe or redirected stdin, we need
 277          to switch it to BINARY mode here, since stdin is already open.  */
 278       if (binary)
 279         SET_BINARY (fileno (stdin));
 280 #endif
 281     }
 282   else
 283     {
 284       /* OPENOPTS is a macro.  It varies with the system.
 285          Some systems distinguish between internal and
 286          external text representations.  */
 287
 288       fp = fopen (filename, OPENOPTS (binary));
 289       if (fp == NULL)
 290         {
 291           error (0, errno, "%s", filename);
 292           return 1;
 293         }
 294     }
 295
 296   err = (*digest_stream) (fp, bin_result);
 297   if (err)
 298     {
 299       error (0, errno, "%s", filename);
 300       if (fp != stdin)
 301         fclose (fp);
 302       return 1;
 303     }
 304
 305   if (fp != stdin && fclose (fp) == EOF)
 306     {
 307       error (0, errno, "%s", filename);
 308       return 1;
 309     }
 310
 311   return 0;
 312 }
 313
 314 static int
 315 digest_check (const char *checkfile_name, int (*digest_stream)(FILE *, void *))
 316 {
 317   FILE *checkfile_stream;
 318   int n_properly_formated_lines = 0;
 319   int n_mismatched_checksums = 0;
 320   int n_open_or_read_failures = 0;
 321   unsigned char bin_buffer[MAX_DIGEST_BIN_BYTES];
 322   size_t line_number;
 323   char *line;
 324   size_t line_chars_allocated;
 325
 326   if (STREQ (checkfile_name, "-"))
 327     {
 328       have_read_stdin = 1;
 329       checkfile_name = _("standard input");
 330       checkfile_stream = stdin;
 331     }
 332   else
 333     {
 334       checkfile_stream = fopen (checkfile_name, "r");
 335       if (checkfile_stream == NULL)
 336         {
 337           error (0, errno, "%s", checkfile_name);
 338           return 1;
 339         }
 340     }
 341
 342   line_number = 0;
 343   line = NULL;
 344   line_chars_allocated = 0;
 345   do
 346     {
 347       char *filename;
 348       int binary;
 349       unsigned char *hex_digest;
 350       int err;
 351       int line_length;
 352
 353       ++line_number;
 354
 355       line_length = getline (&line, &line_chars_allocated, checkfile_stream);
 356       if (line_length <= 0)
 357         break;
 358
 359       /* Ignore comment lines, which begin with a '#' character.  */
 360       if (line[0] == '#')
 361         continue;
 362
 363       /* Remove any trailing newline.  */
 364       if (line[line_length - 1] == '\n')
 365         line[--line_length] = '\0';
 366
 367       err = split_3 (line, line_length, &hex_digest, &binary, &filename);
 368       if (err || !hex_digits (hex_digest))
 369         {
 370           if (warn)
 371             {
 372               error (0, 0,
 373                      _("%s: %lu: improperly formatted %s checksum line"),
 374                      checkfile_name, (unsigned long) line_number,
 375                      DIGEST_TYPE_STRING (algorithm));
 376             }
 377         }
 378       else
 379         {
 380           static const char bin2hex[] = { '0', '1', '2', '3',
 381                                           '4', '5', '6', '7',
 382                                           '8', '9', 'a', 'b',
 383                                           'c', 'd', 'e', 'f' };
 384           int fail;
 385
 386           ++n_properly_formated_lines;
 387
 388           fail = digest_file (filename, binary, bin_buffer, digest_stream);
 389
 390           if (fail)
 391             {
 392               ++n_open_or_read_failures;
 393               if (!status_only)
 394                 {
 395                   printf (_("%s: FAILED open or read\n"), filename);
 396                   fflush (stdout);
 397                 }
 398             }
 399           else
 400             {
 401               size_t digest_bin_bytes = digest_hex_bytes / 2;
 402               size_t cnt;
 403               /* Compare generated binary number with text representation
 404                  in check file.  Ignore case of hex digits.  */
 405               for (cnt = 0; cnt < digest_bin_bytes; ++cnt)
 406                 {
 407                   if (TOLOWER (hex_digest[2 * cnt]) != bin2hex[bin_buffer[cnt] >> 4]
 408                       || (TOLOWER (hex_digest[2 * cnt + 1])
 409                           != (bin2hex[bin_buffer[cnt] & 0xf])))
 410                     break;
 411                 }
 412               if (cnt != digest_bin_bytes)
 413                 ++n_mismatched_checksums;
 414
 415               if (!status_only)
 416                 {
 417                   printf ("%s: %s\n", filename,
 418                           (cnt != digest_bin_bytes ? _("FAILED") : _("OK")));
 419                   fflush (stdout);
 420                 }
 421             }
 422         }
 423     }
 424   while (!feof (checkfile_stream) && !ferror (checkfile_stream));
 425
 426   if (line)
 427     free (line);
 428
 429   if (ferror (checkfile_stream))
 430     {
 431       error (0, 0, _("%s: read error"), checkfile_name);
 432       return 1;
 433     }
 434
 435   if (checkfile_stream != stdin && fclose (checkfile_stream) == EOF)
 436     {
 437       error (0, errno, "%s", checkfile_name);
 438       return 1;
 439     }
 440
 441   if (n_properly_formated_lines == 0)
 442     {
 443       /* Warn if no tests are found.  */
 444       error (0, 0, _("%s: no properly formatted %s checksum lines found"),
 445              checkfile_name, DIGEST_TYPE_STRING (algorithm));
 446     }
 447   else
 448     {
 449       if (!status_only)
 450         {
 451           int n_computed_checkums = (n_properly_formated_lines
 452                                      - n_open_or_read_failures);
 453
 454           if (n_open_or_read_failures > 0)
 455             {
 456               error (0, 0,
 457                      _("WARNING: %d of %d listed %s could not be read"),
 458                      n_open_or_read_failures, n_properly_formated_lines,
 459                      (n_properly_formated_lines == 1
 460                       ? _("file") : _("files")));
 461             }
 462
 463           if (n_mismatched_checksums > 0)
 464             {
 465               error (0, 0,
 466                    _("WARNING: %d of %d computed %s did NOT match"),
 467                      n_mismatched_checksums, n_computed_checkums,
 468                      (n_computed_checkums == 1
 469                       ? _("checksum") : _("checksums")));
 470             }
 471         }
 472     }
 473
 474   return ((n_properly_formated_lines > 0 && n_mismatched_checksums == 0
 475            && n_open_or_read_failures == 0) ? 0 : 1);
 476 }
 477
 478 int
 479 main (int argc, char **argv)
 480 {
 481   unsigned char bin_buffer[MAX_DIGEST_BIN_BYTES];
 482   int do_check = 0;
 483   int opt;
 484   char **string = NULL;
 485   size_t n_strings = 0;
 486   size_t err = 0;
 487   int file_type_specified = 0;
 488
 489 #if O_BINARY
 490   /* Binary is default on MSDOS, so the actual file contents
 491      are used in computation.  */
 492   int binary = 1;
 493 #else
 494   /* Text is default of the Plumb/Lankester format.  */
 495   int binary = 0;
 496 #endif
 497
 498   /* Setting values of global variables.  */
 499   program_name = argv[0];
 500   setlocale (LC_ALL, "");
 501   bindtextdomain (PACKAGE, LOCALEDIR);
 502   textdomain (PACKAGE);
 503
 504   atexit (close_stdout);
 505
 506   while ((opt = getopt_long (argc, argv, "bctw", long_options, NULL)) != -1)
 507     switch (opt)
 508       {
 509       case 0:                   /* long option */
 510         break;
 511       case 1: /* --string */
 512         {
 513           if (string == NULL)
 514             string = (char **) xmalloc ((argc - 1) * sizeof (char *));
 515
 516           if (optarg == NULL)
 517             optarg = "";
 518           string[n_strings++] = optarg;
 519         }
 520         break;
 521       case 'b':
 522         file_type_specified = 1;
 523         binary = 1;
 524         break;
 525       case 'c':
 526         do_check = 1;
 527         break;
 528       case 2:
 529         status_only = 1;
 530         warn = 0;
 531         break;
 532       case 't':
 533         file_type_specified = 1;
 534         binary = 0;
 535         break;
 536       case 'w':
 537         status_only = 0;
 538         warn = 1;
 539         break;
 540       case_GETOPT_HELP_CHAR;
 541       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 542       default:
 543         usage (EXIT_FAILURE);
 544       }
 545
 546   min_digest_line_length = MIN_DIGEST_LINE_LENGTH (algorithm);
 547   digest_hex_bytes = DIGEST_HEX_BYTES (algorithm);
 548
 549   if (file_type_specified && do_check)
 550     {
 551       error (0, 0, _("the --binary and --text options are meaningless when \
 552 verifying checksums"));
 553       usage (EXIT_FAILURE);
 554     }
 555
 556   if (n_strings > 0 && do_check)
 557     {
 558       error (0, 0,
 559              _("the --string and --check options are mutually exclusive"));
 560       usage (EXIT_FAILURE);
 561     }
 562
 563   if (status_only && !do_check)
 564     {
 565       error (0, 0,
 566        _("the --status option is meaningful only when verifying checksums"));
 567       usage (EXIT_FAILURE);
 568     }
 569
 570   if (warn && !do_check)
 571     {
 572       error (0, 0,
 573        _("the --warn option is meaningful only when verifying checksums"));
 574       usage (EXIT_FAILURE);
 575     }
 576
 577   if (n_strings > 0)
 578     {
 579       size_t i;
 580
 581       if (optind < argc)
 582         {
 583           error (0, 0, _("no files may be specified when using --string"));
 584           usage (EXIT_FAILURE);
 585         }
 586       for (i = 0; i < n_strings; ++i)
 587         {
 588           size_t cnt;
 589           if (algorithm == ALG_MD5)
 590             md5_buffer (string[i], strlen (string[i]), bin_buffer);
 591           else
 592             sha_buffer (string[i], strlen (string[i]), bin_buffer);
 593
 594           for (cnt = 0; cnt < (digest_hex_bytes / 2); ++cnt)
 595             printf ("%02x", bin_buffer[cnt]);
 596
 597           printf ("  \"%s\"\n", string[i]);
 598         }
 599     }
 600   else if (do_check)
 601     {
 602       if (optind + 1 < argc)
 603         {
 604           error (0, 0,
 605                  _("only one argument may be specified when using --check"));
 606           usage (EXIT_FAILURE);
 607         }
 608
 609       err = digest_check ((optind == argc) ? "-" : argv[optind],
 610                           DIGEST_STREAM (algorithm));
 611     }
 612   else
 613     {
 614       if (optind == argc)
 615         argv[argc++] = "-";
 616
 617       for (; optind < argc; ++optind)
 618         {
 619           int fail;
 620           char *file = argv[optind];
 621
 622           fail = digest_file (file, binary, bin_buffer,
 623                               DIGEST_STREAM (algorithm));
 624           err |= fail;
 625           if (!fail)
 626             {
 627               size_t i;
 628
 629               /* Output a leading backslash if the file name contains
 630                  a newline or backslash.  */
 631               if (strchr (file, '\n') || strchr (file, '\\'))
 632                 putchar ('\\');
 633
 634               for (i = 0; i < (digest_hex_bytes / 2); ++i)
 635                 printf ("%02x", bin_buffer[i]);
 636
 637               putchar (' ');
 638               if (binary)
 639                 putchar ('*');
 640               else
 641                 putchar (' ');
 642
 643               /* Translate each NEWLINE byte to the string, "\\n",
 644                  and each backslash to "\\\\".  */
 645               for (i = 0; i < strlen (file); ++i)
 646                 {
 647                   switch (file[i])
 648                     {
 649                     case '\n':
 650                       fputs ("\\n", stdout);
 651                       break;
 652
 653                     case '\\':
 654                       fputs ("\\\\", stdout);
 655                       break;
 656
 657                     default:
 658                       putchar (file[i]);
 659                       break;
 660                     }
 661                 }
 662               putchar ('\n');
 663             }
 664         }
 665     }
 666
 667   if (have_read_stdin && fclose (stdin) == EOF)
 668     error (EXIT_FAILURE, errno, _("standard input"));
 669
 670   exit (err == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 671 }