src/wc.c

   1 /* wc - print the number of bytes, words, and lines in files
   2    Copyright (C) 1985, 1991, 1995 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
  16    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  17
  18 /* Written by Paul Rubin, phr@ocf.berkeley.edu
  19    and David MacKenzie, djm@gnu.ai.mit.edu. */
  20 \f
  21 #include <config.h>
  22
  23 #include <stdio.h>
  24 #include <getopt.h>
  25 #include <sys/types.h>
  26 #include "system.h"
  27 #include "version.h"
  28 #include "error.h"
  29
  30 /* Size of atomic reads. */
  31 #define BUFFER_SIZE (16 * 1024)
  32
  33 int safe_read ();
  34
  35 static void wc ();
  36 static void wc_file ();
  37 static void write_counts ();
  38
  39 /* The name this program was run with. */
  40 char *program_name;
  41
  42 /* Cumulative number of lines, words, and chars in all files so far. */
  43 static unsigned long total_lines, total_words, total_chars;
  44
  45 /* Which counts to print. */
  46 static int print_lines, print_words, print_chars;
  47
  48 /* Nonzero if we have ever read the standard input. */
  49 static int have_read_stdin;
  50
  51 /* The error code to return to the system. */
  52 static int exit_status;
  53
  54 /* If non-zero, display usage information and exit.  */
  55 static int show_help;
  56
  57 /* If non-zero, print the version on standard output then exits.  */
  58 static int show_version;
  59
  60 static struct option const longopts[] =
  61 {
  62   {"bytes", no_argument, NULL, 'c'},
  63   {"chars", no_argument, NULL, 'c'},
  64   {"lines", no_argument, NULL, 'l'},
  65   {"words", no_argument, NULL, 'w'},
  66   {"help", no_argument, &show_help, 1},
  67   {"version", no_argument, &show_version, 1},
  68   {NULL, 0, NULL, 0}
  69 };
  70
  71 static void
  72 usage (status)
  73      int status;
  74 {
  75   if (status != 0)
  76     fprintf (stderr, "Try `%s --help' for more information.\n",
  77              program_name);
  78   else
  79     {
  80       printf ("\
  81 Usage: %s [OPTION]... [FILE]...\n\
  82 ",
  83               program_name);
  84       printf ("\
  85 Print line, word, and byte counts for each FILE, and a total line if\n\
  86 more than one FILE is specified.  With no FILE, or when FILE is -,\n\
  87 read standard input.\n\
  88   -l, --lines            print the newline counts\n\
  89   -w, --words            print the word counts\n\
  90   -c, --bytes, --chars   print the byte counts\n\
  91       --help             display this help and exit\n\
  92       --version          output version information and exit\n\
  93 ");
  94     }
  95   exit (status);
  96 }
  97
  98 void
  99 main (argc, argv)
 100      int argc;
 101      char **argv;
 102 {
 103   int optc;
 104   int nfiles;
 105
 106   program_name = argv[0];
 107   exit_status = 0;
 108   print_lines = print_words = print_chars = 0;
 109   total_lines = total_words = total_chars = 0;
 110
 111   while ((optc = getopt_long (argc, argv, "clw", longopts, (int *) 0)) != EOF)
 112     switch (optc)
 113       {
 114       case 0:
 115         break;
 116
 117       case 'c':
 118         print_chars = 1;
 119         break;
 120
 121       case 'l':
 122         print_lines = 1;
 123         break;
 124
 125       case 'w':
 126         print_words = 1;
 127         break;
 128
 129       default:
 130         usage (1);
 131       }
 132
 133   if (show_version)
 134     {
 135       printf ("wc - %s\n", version_string);
 136       exit (0);
 137     }
 138
 139   if (show_help)
 140     usage (0);
 141
 142   if (print_lines + print_words + print_chars == 0)
 143     print_lines = print_words = print_chars = 1;
 144
 145   nfiles = argc - optind;
 146
 147   if (nfiles == 0)
 148     {
 149       have_read_stdin = 1;
 150       wc (0, "");
 151     }
 152   else
 153     {
 154       for (; optind < argc; ++optind)
 155         wc_file (argv[optind]);
 156
 157       if (nfiles > 1)
 158         write_counts (total_lines, total_words, total_chars, "total");
 159     }
 160
 161   if (have_read_stdin && close (0))
 162     error (1, errno, "-");
 163
 164   exit (exit_status);
 165 }
 166
 167 static void
 168 wc_file (file)
 169      char *file;
 170 {
 171   if (!strcmp (file, "-"))
 172     {
 173       have_read_stdin = 1;
 174       wc (0, file);
 175     }
 176   else
 177     {
 178       int fd = open (file, O_RDONLY);
 179       if (fd == -1)
 180         {
 181           error (0, errno, "%s", file);
 182           exit_status = 1;
 183           return;
 184         }
 185       wc (fd, file);
 186       if (close (fd))
 187         {
 188           error (0, errno, "%s", file);
 189           exit_status = 1;
 190         }
 191     }
 192 }
 193
 194 static void
 195 wc (fd, file)
 196      int fd;
 197      char *file;
 198 {
 199   char buf[BUFFER_SIZE + 1];
 200   register int bytes_read;
 201   register int in_word = 0;
 202   register unsigned long lines, words, chars;
 203
 204   lines = words = chars = 0;
 205
 206   /* When counting only bytes, save some line- and word-counting
 207      overhead.  If FD is a `regular' Unix file, using lseek is enough
 208      to get its `size' in bytes.  Otherwise, read blocks of BUFFER_SIZE
 209      bytes at a time until EOF.  Note that the `size' (number of bytes)
 210      that wc reports is smaller than stats.st_size when the file is not
 211      positioned at its beginning.  That's why the lseek calls below are
 212      necessary.  For example the command
 213      `(dd ibs=99k skip=1 count=0; ./wc -c) < /etc/group'
 214      should make wc report `0' bytes.  */
 215
 216   if (print_chars && !print_words && !print_lines)
 217     {
 218       off_t current_pos, end_pos;
 219       struct stat stats;
 220
 221       if (fstat (fd, &stats) == 0 && S_ISREG (stats.st_mode)
 222           && (current_pos = lseek (fd, (off_t) 0, SEEK_CUR)) != -1
 223           && (end_pos = lseek (fd, (off_t) 0, SEEK_END)) != -1)
 224         {
 225           off_t diff;
 226           /* Be careful here.  The current position may actually be
 227              beyond the end of the file.  As in the example above.  */
 228           chars = (diff = end_pos - current_pos) < 0 ? 0 : diff;
 229         }
 230       else
 231         {
 232           while ((bytes_read = safe_read (fd, buf, BUFFER_SIZE)) > 0)
 233             {
 234               chars += bytes_read;
 235             }
 236           if (bytes_read < 0)
 237             {
 238               error (0, errno, "%s", file);
 239               exit_status = 1;
 240             }
 241         }
 242     }
 243   else if (!print_words)
 244     {
 245       /* Use a separate loop when counting only lines or lines and bytes --
 246          but not words.  */
 247       while ((bytes_read = safe_read (fd, buf, BUFFER_SIZE)) > 0)
 248         {
 249           register char *p = buf;
 250
 251           while ((p = memchr (p, '\n', (buf + bytes_read) - p)))
 252             {
 253               ++p;
 254               ++lines;
 255             }
 256           chars += bytes_read;
 257         }
 258       if (bytes_read < 0)
 259         {
 260           error (0, errno, "%s", file);
 261           exit_status = 1;
 262         }
 263     }
 264   else
 265     {
 266       while ((bytes_read = safe_read (fd, buf, BUFFER_SIZE)) > 0)
 267         {
 268           register char *p = buf;
 269
 270           chars += bytes_read;
 271           do
 272             {
 273               switch (*p++)
 274                 {
 275                 case '\n':
 276                   lines++;
 277                   /* Fall through. */
 278                 case '\r':
 279                 case '\f':
 280                 case '\t':
 281                 case '\v':
 282                 case ' ':
 283                   if (in_word)
 284                     {
 285                       in_word = 0;
 286                       words++;
 287                     }
 288                   break;
 289                 default:
 290                   in_word = 1;
 291                   break;
 292                 }
 293             }
 294           while (--bytes_read);
 295         }
 296       if (bytes_read < 0)
 297         {
 298           error (0, errno, "%s", file);
 299           exit_status = 1;
 300         }
 301       if (in_word)
 302         words++;
 303     }
 304
 305   write_counts (lines, words, chars, file);
 306   total_lines += lines;
 307   total_words += words;
 308   total_chars += chars;
 309 }
 310
 311 static void
 312 write_counts (lines, words, chars, file)
 313      unsigned long lines, words, chars;
 314      char *file;
 315 {
 316   if (print_lines)
 317     printf ("%7lu", lines);
 318   if (print_words)
 319     {
 320       if (print_lines)
 321         putchar (' ');
 322       printf ("%7lu", words);
 323     }
 324   if (print_chars)
 325     {
 326       if (print_lines || print_words)
 327         putchar (' ');
 328       printf ("%7lu", chars);
 329     }
 330   if (*file)
 331     printf (" %s", file);
 332   putchar ('\n');
 333 }