src/fold.c

   1 /* fold -- wrap each input line to fit in specified width.
   2    Copyright (C) 1991-2013 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 <http://www.gnu.org/licenses/>.  */
  16
  17 /* Written by David MacKenzie, djm@gnu.ai.mit.edu. */
  18
  19 #include <config.h>
  20
  21 #include <stdio.h>
  22 #include <getopt.h>
  23 #include <sys/types.h>
  24
  25 #include "system.h"
  26 #include "error.h"
  27 #include "fadvise.h"
  28 #include "quote.h"
  29 #include "xstrtol.h"
  30
  31 #define TAB_WIDTH 8
  32
  33 /* The official name of this program (e.g., no 'g' prefix).  */
  34 #define PROGRAM_NAME "fold"
  35
  36 #define AUTHORS proper_name ("David MacKenzie")
  37
  38 /* If nonzero, try to break on whitespace. */
  39 static bool break_spaces;
  40
  41 /* If nonzero, count bytes, not column positions. */
  42 static bool count_bytes;
  43
  44 /* If nonzero, at least one of the files we read was standard input. */
  45 static bool have_read_stdin;
  46
  47 static char const shortopts[] = "bsw:0::1::2::3::4::5::6::7::8::9::";
  48
  49 static struct option const longopts[] =
  50 {
  51   {"bytes", no_argument, NULL, 'b'},
  52   {"spaces", no_argument, NULL, 's'},
  53   {"width", required_argument, NULL, 'w'},
  54   {GETOPT_HELP_OPTION_DECL},
  55   {GETOPT_VERSION_OPTION_DECL},
  56   {NULL, 0, NULL, 0}
  57 };
  58
  59 void
  60 usage (int status)
  61 {
  62   if (status != EXIT_SUCCESS)
  63     emit_try_help ();
  64   else
  65     {
  66       printf (_("\
  67 Usage: %s [OPTION]... [FILE]...\n\
  68 "),
  69               program_name);
  70       fputs (_("\
  71 Wrap input lines in each FILE (standard input by default), writing to\n\
  72 standard output.\n\
  73 "), stdout);
  74
  75       emit_mandatory_arg_note ();
  76
  77       fputs (_("\
  78   -b, --bytes         count bytes rather than columns\n\
  79   -s, --spaces        break at spaces\n\
  80   -w, --width=WIDTH   use WIDTH columns instead of 80\n\
  81 "), stdout);
  82       fputs (HELP_OPTION_DESCRIPTION, stdout);
  83       fputs (VERSION_OPTION_DESCRIPTION, stdout);
  84       emit_ancillary_info ();
  85     }
  86   exit (status);
  87 }
  88
  89 /* Assuming the current column is COLUMN, return the column that
  90    printing C will move the cursor to.
  91    The first column is 0. */
  92
  93 static size_t
  94 adjust_column (size_t column, char c)
  95 {
  96   if (!count_bytes)
  97     {
  98       if (c == '\b')
  99         {
 100           if (column > 0)
 101             column--;
 102         }
 103       else if (c == '\r')
 104         column = 0;
 105       else if (c == '\t')
 106         column += TAB_WIDTH - column % TAB_WIDTH;
 107       else /* if (isprint (c)) */
 108         column++;
 109     }
 110   else
 111     column++;
 112   return column;
 113 }
 114
 115 /* Fold file FILENAME, or standard input if FILENAME is "-",
 116    to stdout, with maximum line length WIDTH.
 117    Return true if successful.  */
 118
 119 static bool
 120 fold_file (char const *filename, size_t width)
 121 {
 122   FILE *istream;
 123   int c;
 124   size_t column = 0;            /* Screen column where next char will go. */
 125   size_t offset_out = 0;        /* Index in 'line_out' for next char. */
 126   static char *line_out = NULL;
 127   static size_t allocated_out = 0;
 128   int saved_errno;
 129
 130   if (STREQ (filename, "-"))
 131     {
 132       istream = stdin;
 133       have_read_stdin = true;
 134     }
 135   else
 136     istream = fopen (filename, "r");
 137
 138   if (istream == NULL)
 139     {
 140       error (0, errno, "%s", filename);
 141       return false;
 142     }
 143
 144   fadvise (istream, FADVISE_SEQUENTIAL);
 145
 146   while ((c = getc (istream)) != EOF)
 147     {
 148       if (offset_out + 1 >= allocated_out)
 149         line_out = X2REALLOC (line_out, &allocated_out);
 150
 151       if (c == '\n')
 152         {
 153           line_out[offset_out++] = c;
 154           fwrite (line_out, sizeof (char), offset_out, stdout);
 155           column = offset_out = 0;
 156           continue;
 157         }
 158
 159     rescan:
 160       column = adjust_column (column, c);
 161
 162       if (column > width)
 163         {
 164           /* This character would make the line too long.
 165              Print the line plus a newline, and make this character
 166              start the next line. */
 167           if (break_spaces)
 168             {
 169               bool found_blank = false;
 170               size_t logical_end = offset_out;
 171
 172               /* Look for the last blank. */
 173               while (logical_end)
 174                 {
 175                   --logical_end;
 176                   if (isblank (to_uchar (line_out[logical_end])))
 177                     {
 178                       found_blank = true;
 179                       break;
 180                     }
 181                 }
 182
 183               if (found_blank)
 184                 {
 185                   size_t i;
 186
 187                   /* Found a blank.  Don't output the part after it. */
 188                   logical_end++;
 189                   fwrite (line_out, sizeof (char), (size_t) logical_end,
 190                           stdout);
 191                   putchar ('\n');
 192                   /* Move the remainder to the beginning of the next line.
 193                      The areas being copied here might overlap. */
 194                   memmove (line_out, line_out + logical_end,
 195                            offset_out - logical_end);
 196                   offset_out -= logical_end;
 197                   for (column = i = 0; i < offset_out; i++)
 198                     column = adjust_column (column, line_out[i]);
 199                   goto rescan;
 200                 }
 201             }
 202
 203           if (offset_out == 0)
 204             {
 205               line_out[offset_out++] = c;
 206               continue;
 207             }
 208
 209           line_out[offset_out++] = '\n';
 210           fwrite (line_out, sizeof (char), (size_t) offset_out, stdout);
 211           column = offset_out = 0;
 212           goto rescan;
 213         }
 214
 215       line_out[offset_out++] = c;
 216     }
 217
 218   saved_errno = errno;
 219
 220   if (offset_out)
 221     fwrite (line_out, sizeof (char), (size_t) offset_out, stdout);
 222
 223   if (ferror (istream))
 224     {
 225       error (0, saved_errno, "%s", filename);
 226       if (!STREQ (filename, "-"))
 227         fclose (istream);
 228       return false;
 229     }
 230   if (!STREQ (filename, "-") && fclose (istream) == EOF)
 231     {
 232       error (0, errno, "%s", filename);
 233       return false;
 234     }
 235
 236   return true;
 237 }
 238
 239 int
 240 main (int argc, char **argv)
 241 {
 242   size_t width = 80;
 243   int i;
 244   int optc;
 245   bool ok;
 246
 247   initialize_main (&argc, &argv);
 248   set_program_name (argv[0]);
 249   setlocale (LC_ALL, "");
 250   bindtextdomain (PACKAGE, LOCALEDIR);
 251   textdomain (PACKAGE);
 252
 253   atexit (close_stdout);
 254
 255   break_spaces = count_bytes = have_read_stdin = false;
 256
 257   while ((optc = getopt_long (argc, argv, shortopts, longopts, NULL)) != -1)
 258     {
 259       char optargbuf[2];
 260
 261       switch (optc)
 262         {
 263         case 'b':               /* Count bytes rather than columns. */
 264           count_bytes = true;
 265           break;
 266
 267         case 's':               /* Break at word boundaries. */
 268           break_spaces = true;
 269           break;
 270
 271         case '0': case '1': case '2': case '3': case '4':
 272         case '5': case '6': case '7': case '8': case '9':
 273           if (optarg)
 274             optarg--;
 275           else
 276             {
 277               optargbuf[0] = optc;
 278               optargbuf[1] = '\0';
 279               optarg = optargbuf;
 280             }
 281           /* Fall through.  */
 282         case 'w':               /* Line width. */
 283           {
 284             unsigned long int tmp_ulong;
 285             if (! (xstrtoul (optarg, NULL, 10, &tmp_ulong, "") == LONGINT_OK
 286                    && 0 < tmp_ulong && tmp_ulong < SIZE_MAX - TAB_WIDTH))
 287               error (EXIT_FAILURE, 0,
 288                      _("invalid number of columns: %s"), quote (optarg));
 289             width = tmp_ulong;
 290           }
 291           break;
 292
 293         case_GETOPT_HELP_CHAR;
 294
 295         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 296
 297         default:
 298           usage (EXIT_FAILURE);
 299         }
 300     }
 301
 302   if (argc == optind)
 303     ok = fold_file ("-", width);
 304   else
 305     {
 306       ok = true;
 307       for (i = optind; i < argc; i++)
 308         ok &= fold_file (argv[i], width);
 309     }
 310
 311   if (have_read_stdin && fclose (stdin) == EOF)
 312     error (EXIT_FAILURE, errno, "-");
 313
 314   exit (ok ? EXIT_SUCCESS : EXIT_FAILURE);
 315 }