src/md5sum.c

   1 /* Compute MD5 or SHA1 checksum of files or strings
   2    Copyright (C) 1995-2002 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 N_ ("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 "),
 131               program_name, program_name,
 132               DIGEST_TYPE_STRING (algorithm),
 133               DIGEST_BITS (algorithm));
 134       printf (_("\
 135 \n\
 136   -b, --binary            read files in binary mode (default on DOS/Windows)\n\
 137   -c, --check             check %s sums against given list\n\
 138   -t, --text              read files in text mode (default)\n\
 139 \n\
 140 "),
 141               DIGEST_TYPE_STRING (algorithm));
 142       fputs (_("\
 143 The following two options are useful only when verifying checksums:\n\
 144       --status            don't output anything, status code shows success\n\
 145   -w, --warn              warn about improperly formated checksum lines\n\
 146 \n\
 147 "), stdout);
 148       fputs (HELP_OPTION_DESCRIPTION, stdout);
 149       fputs (VERSION_OPTION_DESCRIPTION, stdout);
 150       printf (_("\
 151 \n\
 152 The sums are computed as described in %s.  When checking, the input\n\
 153 should be a former output of this program.  The default mode is to print\n\
 154 a line with checksum, a character indicating type (`*' for binary, ` ' for\n\
 155 text), and name for each FILE.\n"),
 156               (algorithm == ALG_MD5 ? "RFC 1321" : "FIPS-180-1"));
 157       printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
 158     }
 159
 160   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 161 }
 162
 163 static int
 164 split_3 (char *s, size_t s_len, unsigned char **u, int *binary, char **w)
 165 {
 166   size_t i;
 167   int escaped_filename = 0;
 168
 169 #define ISWHITE(c) ((c) == ' ' || (c) == '\t')
 170
 171   i = 0;
 172   while (ISWHITE (s[i]))
 173     ++i;
 174
 175   /* The line must have at least `min_digest_line_length - 1' (or one more, if
 176      the first is a backslash) more characters to contain correct message digest
 177      information.  Ignore this line if it is too short.  */
 178   if (!(s_len - i >= min_digest_line_length
 179         || (s[i] == '\\' && s_len - i >= 1 + min_digest_line_length)))
 180     return 1;
 181
 182   if (s[i] == '\\')
 183     {
 184       ++i;
 185       escaped_filename = 1;
 186     }
 187   *u = (unsigned char *) &s[i];
 188
 189   /* The first field has to be the n-character hexadecimal
 190      representation of the message digest.  If it is not followed
 191      immediately by a white space it's an error.  */
 192   i += digest_hex_bytes;
 193   if (!ISWHITE (s[i]))
 194     return 1;
 195
 196   s[i++] = '\0';
 197
 198   if (s[i] != ' ' && s[i] != '*')
 199     return 1;
 200   *binary = (s[i++] == '*');
 201
 202   /* All characters between the type indicator and end of line are
 203      significant -- that includes leading and trailing white space.  */
 204   *w = &s[i];
 205
 206   if (escaped_filename)
 207     {
 208       /* Translate each `\n' string in the file name to a NEWLINE,
 209          and each `\\' string to a backslash.  */
 210
 211       char *dst = &s[i];
 212
 213       while (i < s_len)
 214         {
 215           switch (s[i])
 216             {
 217             case '\\':
 218               if (i == s_len - 1)
 219                 {
 220                   /* A valid line does not end with a backslash.  */
 221                   return 1;
 222                 }
 223               ++i;
 224               switch (s[i++])
 225                 {
 226                 case 'n':
 227                   *dst++ = '\n';
 228                   break;
 229                 case '\\':
 230                   *dst++ = '\\';
 231                   break;
 232                 default:
 233                   /* Only `\' or `n' may follow a backslash.  */
 234                   return 1;
 235                 }
 236               break;
 237
 238             case '\0':
 239               /* The file name may not contain a NUL.  */
 240               return 1;
 241               break;
 242
 243             default:
 244               *dst++ = s[i++];
 245               break;
 246             }
 247         }
 248       *dst = '\0';
 249     }
 250   return 0;
 251 }
 252
 253 static int
 254 hex_digits (unsigned char const *s)
 255 {
 256   while (*s)
 257     {
 258       if (!ISXDIGIT (*s))
 259         return 0;
 260       ++s;
 261     }
 262   return 1;
 263 }
 264
 265 /* An interface to the function, DIGEST_STREAM, (either md5_stream or sha_stream).
 266    Operate on FILENAME (it may be "-") and put the result in *BIN_RESULT.
 267    Return non-zero upon failure, zero to indicate success.  */
 268
 269 static int
 270 digest_file (const char *filename, int binary, unsigned char *bin_result,
 271            int (*digest_stream)(FILE *, void *))
 272 {
 273   FILE *fp;
 274   int err;
 275
 276   if (STREQ (filename, "-"))
 277     {
 278       have_read_stdin = 1;
 279       fp = stdin;
 280 #if O_BINARY
 281       /* If we need binary reads from a pipe or redirected stdin, we need
 282          to switch it to BINARY mode here, since stdin is already open.  */
 283       if (binary)
 284         SET_BINARY (fileno (stdin));
 285 #endif
 286     }
 287   else
 288     {
 289       /* OPENOPTS is a macro.  It varies with the system.
 290          Some systems distinguish between internal and
 291          external text representations.  */
 292
 293       fp = fopen (filename, OPENOPTS (binary));
 294       if (fp == NULL)
 295         {
 296           error (0, errno, "%s", filename);
 297           return 1;
 298         }
 299     }
 300
 301   err = (*digest_stream) (fp, bin_result);
 302   if (err)
 303     {
 304       error (0, errno, "%s", filename);
 305       if (fp != stdin)
 306         fclose (fp);
 307       return 1;
 308     }
 309
 310   if (fp != stdin && fclose (fp) == EOF)
 311     {
 312       error (0, errno, "%s", filename);
 313       return 1;
 314     }
 315
 316   return 0;
 317 }
 318
 319 static int
 320 digest_check (const char *checkfile_name, int (*digest_stream)(FILE *, void *))
 321 {
 322   FILE *checkfile_stream;
 323   int n_properly_formated_lines = 0;
 324   int n_mismatched_checksums = 0;
 325   int n_open_or_read_failures = 0;
 326   unsigned char bin_buffer[MAX_DIGEST_BIN_BYTES];
 327   size_t line_number;
 328   char *line;
 329   size_t line_chars_allocated;
 330
 331   if (STREQ (checkfile_name, "-"))
 332     {
 333       have_read_stdin = 1;
 334       checkfile_name = _("standard input");
 335       checkfile_stream = stdin;
 336     }
 337   else
 338     {
 339       checkfile_stream = fopen (checkfile_name, "r");
 340       if (checkfile_stream == NULL)
 341         {
 342           error (0, errno, "%s", checkfile_name);
 343           return 1;
 344         }
 345     }
 346
 347   SET_MODE (fileno (checkfile_stream), O_TEXT);
 348   line_number = 0;
 349   line = NULL;
 350   line_chars_allocated = 0;
 351   do
 352     {
 353       char *filename;
 354       int binary;
 355       unsigned char *hex_digest;
 356       int err;
 357       int line_length;
 358
 359       ++line_number;
 360
 361       line_length = getline (&line, &line_chars_allocated, checkfile_stream);
 362       if (line_length <= 0)
 363         break;
 364
 365       /* Ignore comment lines, which begin with a '#' character.  */
 366       if (line[0] == '#')
 367         continue;
 368
 369       /* Remove any trailing newline.  */
 370       if (line[line_length - 1] == '\n')
 371         line[--line_length] = '\0';
 372
 373       err = split_3 (line, line_length, &hex_digest, &binary, &filename);
 374       if (err || !hex_digits (hex_digest))
 375         {
 376           if (warn)
 377             {
 378               error (0, 0,
 379                      _("%s: %lu: improperly formatted %s checksum line"),
 380                      checkfile_name, (unsigned long) line_number,
 381                      DIGEST_TYPE_STRING (algorithm));
 382             }
 383         }
 384       else
 385         {
 386           static const char bin2hex[] = { '0', '1', '2', '3',
 387                                           '4', '5', '6', '7',
 388                                           '8', '9', 'a', 'b',
 389                                           'c', 'd', 'e', 'f' };
 390           int fail;
 391
 392           ++n_properly_formated_lines;
 393
 394           fail = digest_file (filename, binary, bin_buffer, digest_stream);
 395
 396           if (fail)
 397             {
 398               ++n_open_or_read_failures;
 399               if (!status_only)
 400                 {
 401                   printf (_("%s: FAILED open or read\n"), filename);
 402                   fflush (stdout);
 403                 }
 404             }
 405           else
 406             {
 407               size_t digest_bin_bytes = digest_hex_bytes / 2;
 408               size_t cnt;
 409               /* Compare generated binary number with text representation
 410                  in check file.  Ignore case of hex digits.  */
 411               for (cnt = 0; cnt < digest_bin_bytes; ++cnt)
 412                 {
 413                   if (TOLOWER (hex_digest[2 * cnt])
 414                       != bin2hex[bin_buffer[cnt] >> 4]
 415                       || (TOLOWER (hex_digest[2 * cnt + 1])
 416                           != (bin2hex[bin_buffer[cnt] & 0xf])))
 417                     break;
 418                 }
 419               if (cnt != digest_bin_bytes)
 420                 ++n_mismatched_checksums;
 421
 422               if (!status_only)
 423                 {
 424                   printf ("%s: %s\n", filename,
 425                           (cnt != digest_bin_bytes ? _("FAILED") : _("OK")));
 426                   fflush (stdout);
 427                 }
 428             }
 429         }
 430     }
 431   while (!feof (checkfile_stream) && !ferror (checkfile_stream));
 432
 433   if (line)
 434     free (line);
 435
 436   if (ferror (checkfile_stream))
 437     {
 438       error (0, 0, _("%s: read error"), checkfile_name);
 439       return 1;
 440     }
 441
 442   if (checkfile_stream != stdin && fclose (checkfile_stream) == EOF)
 443     {
 444       error (0, errno, "%s", checkfile_name);
 445       return 1;
 446     }
 447
 448   if (n_properly_formated_lines == 0)
 449     {
 450       /* Warn if no tests are found.  */
 451       error (0, 0, _("%s: no properly formatted %s checksum lines found"),
 452              checkfile_name, DIGEST_TYPE_STRING (algorithm));
 453     }
 454   else
 455     {
 456       if (!status_only)
 457         {
 458           int n_computed_checkums = (n_properly_formated_lines
 459                                      - n_open_or_read_failures);
 460
 461           if (n_open_or_read_failures > 0)
 462             {
 463               error (0, 0,
 464                      _("WARNING: %d of %d listed %s could not be read"),
 465                      n_open_or_read_failures, n_properly_formated_lines,
 466                      (n_properly_formated_lines == 1
 467                       ? _("file") : _("files")));
 468             }
 469
 470           if (n_mismatched_checksums > 0)
 471             {
 472               error (0, 0,
 473                    _("WARNING: %d of %d computed %s did NOT match"),
 474                      n_mismatched_checksums, n_computed_checkums,
 475                      (n_computed_checkums == 1
 476                       ? _("checksum") : _("checksums")));
 477             }
 478         }
 479     }
 480
 481   return ((n_properly_formated_lines > 0 && n_mismatched_checksums == 0
 482            && n_open_or_read_failures == 0) ? 0 : 1);
 483 }
 484
 485 int
 486 main (int argc, char **argv)
 487 {
 488   unsigned char bin_buffer[MAX_DIGEST_BIN_BYTES];
 489   int do_check = 0;
 490   int opt;
 491   char **string = NULL;
 492   size_t n_strings = 0;
 493   size_t err = 0;
 494   int file_type_specified = 0;
 495
 496 #if O_BINARY
 497   /* Binary is default on MSDOS, so the actual file contents
 498      are used in computation.  */
 499   int binary = 1;
 500 #else
 501   /* Text is default of the Plumb/Lankester format.  */
 502   int binary = 0;
 503 #endif
 504
 505   /* Setting values of global variables.  */
 506   program_name = argv[0];
 507   setlocale (LC_ALL, "");
 508   bindtextdomain (PACKAGE, LOCALEDIR);
 509   textdomain (PACKAGE);
 510
 511   atexit (close_stdout);
 512
 513   while ((opt = getopt_long (argc, argv, "bctw", long_options, NULL)) != -1)
 514     switch (opt)
 515       {
 516       case 0:                   /* long option */
 517         break;
 518       case 1: /* --string */
 519         {
 520           if (string == NULL)
 521             string = (char **) xmalloc ((argc - 1) * sizeof (char *));
 522
 523           if (optarg == NULL)
 524             optarg = "";
 525           string[n_strings++] = optarg;
 526         }
 527         break;
 528       case 'b':
 529         file_type_specified = 1;
 530         binary = 1;
 531         break;
 532       case 'c':
 533         do_check = 1;
 534         break;
 535       case 2:
 536         status_only = 1;
 537         warn = 0;
 538         break;
 539       case 't':
 540         file_type_specified = 1;
 541         binary = 0;
 542         break;
 543       case 'w':
 544         status_only = 0;
 545         warn = 1;
 546         break;
 547       case_GETOPT_HELP_CHAR;
 548       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 549       default:
 550         usage (EXIT_FAILURE);
 551       }
 552
 553   min_digest_line_length = MIN_DIGEST_LINE_LENGTH (algorithm);
 554   digest_hex_bytes = DIGEST_HEX_BYTES (algorithm);
 555
 556   if (file_type_specified && do_check)
 557     {
 558       error (0, 0, _("the --binary and --text options are meaningless when \
 559 verifying checksums"));
 560       usage (EXIT_FAILURE);
 561     }
 562
 563   if (n_strings > 0 && do_check)
 564     {
 565       error (0, 0,
 566              _("the --string and --check options are mutually exclusive"));
 567       usage (EXIT_FAILURE);
 568     }
 569
 570   if (status_only && !do_check)
 571     {
 572       error (0, 0,
 573        _("the --status option is meaningful only when verifying checksums"));
 574       usage (EXIT_FAILURE);
 575     }
 576
 577   if (warn && !do_check)
 578     {
 579       error (0, 0,
 580        _("the --warn option is meaningful only when verifying checksums"));
 581       usage (EXIT_FAILURE);
 582     }
 583
 584   if (n_strings > 0)
 585     {
 586       size_t i;
 587
 588       if (optind < argc)
 589         {
 590           error (0, 0, _("no files may be specified when using --string"));
 591           usage (EXIT_FAILURE);
 592         }
 593       for (i = 0; i < n_strings; ++i)
 594         {
 595           size_t cnt;
 596           if (algorithm == ALG_MD5)
 597             md5_buffer (string[i], strlen (string[i]), bin_buffer);
 598           else
 599             sha_buffer (string[i], strlen (string[i]), bin_buffer);
 600
 601           for (cnt = 0; cnt < (digest_hex_bytes / 2); ++cnt)
 602             printf ("%02x", bin_buffer[cnt]);
 603
 604           printf ("  \"%s\"\n", string[i]);
 605         }
 606     }
 607   else if (do_check)
 608     {
 609       if (optind + 1 < argc)
 610         {
 611           error (0, 0,
 612                  _("only one argument may be specified when using --check"));
 613           usage (EXIT_FAILURE);
 614         }
 615
 616       err = digest_check ((optind == argc) ? "-" : argv[optind],
 617                           DIGEST_STREAM (algorithm));
 618     }
 619   else
 620     {
 621       if (optind == argc)
 622         argv[argc++] = "-";
 623
 624       for (; optind < argc; ++optind)
 625         {
 626           int fail;
 627           char *file = argv[optind];
 628
 629           fail = digest_file (file, binary, bin_buffer,
 630                               DIGEST_STREAM (algorithm));
 631           err |= fail;
 632           if (!fail)
 633             {
 634               size_t i;
 635
 636               /* Output a leading backslash if the file name contains
 637                  a newline or backslash.  */
 638               if (strchr (file, '\n') || strchr (file, '\\'))
 639                 putchar ('\\');
 640
 641               for (i = 0; i < (digest_hex_bytes / 2); ++i)
 642                 printf ("%02x", bin_buffer[i]);
 643
 644               putchar (' ');
 645               if (binary)
 646                 putchar ('*');
 647               else
 648                 putchar (' ');
 649
 650               /* Translate each NEWLINE byte to the string, "\\n",
 651                  and each backslash to "\\\\".  */
 652               for (i = 0; i < strlen (file); ++i)
 653                 {
 654                   switch (file[i])
 655                     {
 656                     case '\n':
 657                       fputs ("\\n", stdout);
 658                       break;
 659
 660                     case '\\':
 661                       fputs ("\\\\", stdout);
 662                       break;
 663
 664                     default:
 665                       putchar (file[i]);
 666                       break;
 667                     }
 668                 }
 669               putchar ('\n');
 670             }
 671         }
 672     }
 673
 674   if (have_read_stdin && fclose (stdin) == EOF)
 675     error (EXIT_FAILURE, errno, _("standard input"));
 676
 677   exit (err == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 678 }