diff.c

这是一个同样来自贝尔实验室的和UNIX有着渊源的操作系统, 其简洁的设计和实现易于我们学习和理解

"-s  --report-identical-files  Report when two files are the same.",
"-x PAT  --exclude=PAT  Exclude files that match PAT.",
"-X FILE  --exclude-from=FILE  Exclude files that match any pattern in FILE.",
"-S FILE  --starting-file=FILE  Start with FILE when comparing directories.\n",
"--horizon-lines=NUM  Keep NUM lines of the common prefix and suffix.",
"-d  --minimal  Try hard to find a smaller set of changes.",
"-H  --speed-large-files  Assume large files and many scattered small changes.\n",
"-v  --version  Output version info.",
"--help  Output this help.",
0
};

static void
usage ()
{
  char const * const *p;

  printf ("Usage: %s [OPTION]... FILE1 FILE2\n\n", program_name);
  for (p = option_help;  *p;  p++)
    printf ("  %s\n", *p);
  printf ("\nIf FILE1 or FILE2 is `-', read standard input.\n");
}

static int
specify_format (var, value)
     char **var;
     char *value;
{
  int err = *var ? strcmp (*var, value) : 0;
  *var = value;
  return err;
}

static void
specify_style (style)
     enum output_style style;
{
  if (output_style != OUTPUT_NORMAL
      && output_style != style)
    error ("conflicting specifications of output style", 0, 0);
  output_style = style;
}

static char const *
filetype (st)
     struct stat const *st;
{
  /* See Posix.2 section 4.17.6.1.1 and Table 5-1 for these formats.
     To keep diagnostics grammatical, the returned string must start
     with a consonant.  */

  if (S_ISREG (st->st_mode))
    {
      if (st->st_size == 0)
	return "regular empty file";
      /* Posix.2 section 5.14.2 seems to suggest that we must read the file
	 and guess whether it's C, Fortran, etc., but this is somewhat useless
	 and doesn't reflect historical practice.  We're allowed to guess
	 wrong, so we don't bother to read the file.  */
      return "regular file";
    }
  if (S_ISDIR (st->st_mode)) return "directory";

  /* other Posix.1 file types */
#ifdef S_ISBLK
  if (S_ISBLK (st->st_mode)) return "block special file";
#endif
#ifdef S_ISCHR
  if (S_ISCHR (st->st_mode)) return "character special file";
#endif
#ifdef S_ISFIFO
  if (S_ISFIFO (st->st_mode)) return "fifo";
#endif

  /* other Posix.1b file types */
#ifdef S_TYPEISMQ
  if (S_TYPEISMQ (st)) return "message queue";
#endif
#ifdef S_TYPEISSEM
  if (S_TYPEISSEM (st)) return "semaphore";
#endif
#ifdef S_TYPEISSHM
  if (S_TYPEISSHM (st)) return "shared memory object";
#endif

  /* other popular file types */
  /* S_ISLNK is impossible with `fstat' and `stat'.  */
#ifdef S_ISSOCK
  if (S_ISSOCK (st->st_mode)) return "socket";
#endif

  return "weird file";
}

/* Compare two files (or dirs) with specified names
   DIR0/NAME0 and DIR1/NAME1, at level DEPTH in directory recursion.
   (if DIR0 is 0, then the name is just NAME0, etc.)
   This is self-contained; it opens the files and closes them.

   Value is 0 if files are the same, 1 if different,
   2 if there is a problem opening them.  */

static int
compare_files (dir0, name0, dir1, name1, depth)
     char const *dir0, *dir1;
     char const *name0, *name1;
     int depth;
{
  struct file_data inf[2];
  register int i;
  int val;
  int same_files;
  int failed = 0;
  char *free0 = 0, *free1 = 0;

  /* If this is directory comparison, perhaps we have a file
     that exists only in one of the directories.
     If so, just print a message to that effect.  */

  if (! ((name0 != 0 && name1 != 0)
	 || (unidirectional_new_file_flag && name1 != 0)
	 || entire_new_file_flag))
    {
      char const *name = name0 == 0 ? name1 : name0;
      char const *dir = name0 == 0 ? dir1 : dir0;
      message ("Only in %s: %s\n", dir, name);
      /* Return 1 so that diff_dirs will return 1 ("some files differ").  */
      return 1;
    }

  bzero (inf, sizeof (inf));

  /* Mark any nonexistent file with -1 in the desc field.  */
  /* Mark unopened files (e.g. directories) with -2. */

  inf[0].desc = name0 == 0 ? -1 : -2;
  inf[1].desc = name1 == 0 ? -1 : -2;

  /* Now record the full name of each file, including nonexistent ones.  */

  if (name0 == 0)
    name0 = name1;
  if (name1 == 0)
    name1 = name0;

  inf[0].name = dir0 == 0 ? name0 : (free0 = dir_file_pathname (dir0, name0));
  inf[1].name = dir1 == 0 ? name1 : (free1 = dir_file_pathname (dir1, name1));

  /* Stat the files.  Record whether they are directories.  */

  for (i = 0; i <= 1; i++)
    {
      if (inf[i].desc != -1)
	{
	  int stat_result;

	  if (i && filename_cmp (inf[i].name, inf[0].name) == 0)
	    {
	      inf[i].stat = inf[0].stat;
	      stat_result = 0;
	    }
	  else if (strcmp (inf[i].name, "-") == 0)
	    {
	      inf[i].desc = STDIN_FILENO;
	      stat_result = fstat (STDIN_FILENO, &inf[i].stat);
	      if (stat_result == 0 && S_ISREG (inf[i].stat.st_mode))
		{
		  off_t pos = lseek (STDIN_FILENO, (off_t) 0, SEEK_CUR);
		  if (pos == -1)
		    stat_result = -1;
		  else
		    {
		      if (pos <= inf[i].stat.st_size)
			inf[i].stat.st_size -= pos;
		      else
			inf[i].stat.st_size = 0;
		      /* Posix.2 4.17.6.1.4 requires current time for stdin.  */
		      time (&inf[i].stat.st_mtime);
		    }
		}
	    }
	  else
	    stat_result = stat (inf[i].name, &inf[i].stat);

	  if (stat_result != 0)
	    {
	      perror_with_name (inf[i].name);
	      failed = 1;
	    }
	  else
	    {
	      inf[i].dir_p = S_ISDIR (inf[i].stat.st_mode) && inf[i].desc != 0;
	      if (inf[1 - i].desc == -1)
		{
		  inf[1 - i].dir_p = inf[i].dir_p;
		  inf[1 - i].stat.st_mode = inf[i].stat.st_mode;
		}
	    }
	}
    }

  if (! failed && depth == 0 && inf[0].dir_p != inf[1].dir_p)
    {
      /* If one is a directory, and it was specified in the command line,
	 use the file in that dir with the other file's basename.  */

      int fnm_arg = inf[0].dir_p;
      int dir_arg = 1 - fnm_arg;
      char const *fnm = inf[fnm_arg].name;
      char const *dir = inf[dir_arg].name;
      char const *p = filename_lastdirchar (fnm);
      char const *filename = inf[dir_arg].name
	= dir_file_pathname (dir, p ? p + 1 : fnm);

      if (strcmp (fnm, "-") == 0)
	fatal ("can't compare - to a directory");

      if (stat (filename, &inf[dir_arg].stat) != 0)
	{
	  perror_with_name (filename);
	  failed = 1;
	}
      else
	inf[dir_arg].dir_p = S_ISDIR (inf[dir_arg].stat.st_mode);
    }

  if (failed)
    {

      /* If either file should exist but does not, return 2.  */

      val = 2;

    }
  else if ((same_files = inf[0].desc != -1 && inf[1].desc != -1
			 && 0 < same_file (&inf[0].stat, &inf[1].stat))
	   && no_diff_means_no_output)
    {
      /* The two named files are actually the same physical file.
	 We know they are identical without actually reading them.  */

      val = 0;
    }
  else if (inf[0].dir_p & inf[1].dir_p)
    {
      if (output_style == OUTPUT_IFDEF)
	fatal ("-D option not supported with directories");

      /* If both are directories, compare the files in them.  */

      if (depth > 0 && !recursive)
	{
	  /* But don't compare dir contents one level down
	     unless -r was specified.  */
	  message ("Common subdirectories: %s and %s\n",
		   inf[0].name, inf[1].name);
	  val = 0;
	}
      else
	{
	  val = diff_dirs (inf, compare_files, depth);
	}

    }
  else if ((inf[0].dir_p | inf[1].dir_p)
	   || (depth > 0
	       && (! S_ISREG (inf[0].stat.st_mode)
		   || ! S_ISREG (inf[1].stat.st_mode))))
    {
      /* Perhaps we have a subdirectory that exists only in one directory.
	 If so, just print a message to that effect.  */

      if (inf[0].desc == -1 || inf[1].desc == -1)
	{
	  if ((inf[0].dir_p | inf[1].dir_p)
	      && recursive
	      && (entire_new_file_flag
		  || (unidirectional_new_file_flag && inf[0].desc == -1)))
	    val = diff_dirs (inf, compare_files, depth);
	  else
	    {
	      char const *dir = (inf[0].desc == -1) ? dir1 : dir0;
	      /* See Posix.2 section 4.17.6.1.1 for this format.  */
	      message ("Only in %s: %s\n", dir, name0);
	      val = 1;
	    }
	}
      else
	{
	  /* We have two files that are not to be compared.  */

	  /* See Posix.2 section 4.17.6.1.1 for this format.  */
	  message5 ("File %s is a %s while file %s is a %s\n",
		    inf[0].name, filetype (&inf[0].stat),
		    inf[1].name, filetype (&inf[1].stat));

	  /* This is a difference.  */
	  val = 1;
	}
    }
  else if ((no_details_flag & ~ignore_some_changes)
	   && inf[0].stat.st_size != inf[1].stat.st_size
	   && (inf[0].desc == -1 || S_ISREG (inf[0].stat.st_mode))
	   && (inf[1].desc == -1 || S_ISREG (inf[1].stat.st_mode)))
    {
      message ("Files %s and %s differ\n", inf[0].name, inf[1].name);
      val = 1;
    }
  else
    {
      /* Both exist and neither is a directory.  */

      /* Open the files and record their descriptors.  */

      if (inf[0].desc == -2)
	if ((inf[0].desc = open (inf[0].name, O_RDONLY, 0)) < 0)
	  {
	    perror_with_name (inf[0].name);
	    failed = 1;
	  }
      if (inf[1].desc == -2)
	if (same_files)
	  inf[1].desc = inf[0].desc;
	else if ((inf[1].desc = open (inf[1].name, O_RDONLY, 0)) < 0)
	  {
	    perror_with_name (inf[1].name);
	    failed = 1;
	  }

#if HAVE_SETMODE
      if (binary_I_O)
	for (i = 0; i <= 1; i++)
	  if (0 <= inf[i].desc)
	    setmode (inf[i].desc, O_BINARY);
#endif

      /* Compare the files, if no error was found.  */

      val = failed ? 2 : diff_2_files (inf, depth);

      /* Close the file descriptors.  */

      if (inf[0].desc >= 0 && close (inf[0].desc) != 0)
	{
	  perror_with_name (inf[0].name);
	  val = 2;
	}
      if (inf[1].desc >= 0 && inf[0].desc != inf[1].desc
	  && close (inf[1].desc) != 0)
	{
	  perror_with_name (inf[1].name);
	  val = 2;
	}
    }

  /* Now the comparison has been done, if no error prevented it,
     and VAL is the value this function will return.  */

  if (val == 0 && !inf[0].dir_p)
    {
      if (print_file_same_flag)
	message ("Files %s and %s are identical\n",
		 inf[0].name, inf[1].name);
    }
  else
    fflush (stdout);

  if (free0)
    free (free0);
  if (free1)
    free (free1);

  return val;
}