src/unexpand.c

   1 /* unexpand - convert spaces to tabs
   2    Copyright (C) 89, 91, 95, 96, 1997, 1998 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 /* By default, convert only maximal strings of initial blanks and tabs
  19    into tabs.
  20    Preserves backspace characters in the output; they decrement the
  21    column count for tab calculations.
  22    The default action is equivalent to -8.
  23
  24    Options:
  25    --tabs=tab1[,tab2[,...]]
  26    -t tab1[,tab2[,...]]
  27    -tab1[,tab2[,...]]   If only one tab stop is given, set the tabs tab1
  28                         spaces apart instead of the default 8.  Otherwise,
  29                         set the tabs at columns tab1, tab2, etc. (numbered from
  30                         0); replace any tabs beyond the tabstops given with
  31                         single spaces.
  32    --all
  33    -a                   Use tabs wherever they would replace 2 or more spaces,
  34                         not just at the beginnings of lines.
  35
  36    David MacKenzie <djm@gnu.ai.mit.edu> */
  37
  38 #include <config.h>
  39
  40 #include <stdio.h>
  41 #include <getopt.h>
  42 #include <sys/types.h>
  43 #include "system.h"
  44
  45 #include "error.h"
  46
  47 /* The number of bytes added at a time to the amount of memory
  48    allocated for the output line. */
  49 #define OUTPUT_BLOCK 256
  50
  51 /* The number of bytes added at a time to the amount of memory
  52    allocated for the list of tabstops. */
  53 #define TABLIST_BLOCK 256
  54
  55 /* The name this program was run with. */
  56 char *program_name;
  57
  58 /* If nonzero, convert blanks even after nonblank characters have been
  59    read on the line. */
  60 static int convert_entire_line;
  61
  62 /* If nonzero, the size of all tab stops.  If zero, use `tab_list' instead. */
  63 static int tab_size;
  64
  65 /* Array of the explicit column numbers of the tab stops;
  66    after `tab_list' is exhausted, the rest of the line is printed
  67    unchanged.  The first column is column 0. */
  68 static int *tab_list;
  69
  70 /* The index of the first invalid element of `tab_list',
  71    where the next element can be added. */
  72 static int first_free_tab;
  73
  74 /* Null-terminated array of input filenames. */
  75 static char **file_list;
  76
  77 /* Default for `file_list' if no files are given on the command line. */
  78 static char *stdin_argv[] =
  79 {
  80   "-", NULL
  81 };
  82
  83 /* Nonzero if we have ever read standard input. */
  84 static int have_read_stdin;
  85
  86 /* Status to return to the system. */
  87 static int exit_status;
  88
  89 /* If nonzero, display usage information and exit.  */
  90 static int show_help;
  91
  92 /* If nonzero, print the version on standard output then exit.  */
  93 static int show_version;
  94
  95 static struct option const longopts[] =
  96 {
  97   {"tabs", required_argument, NULL, 't'},
  98   {"all", no_argument, NULL, 'a'},
  99   {"help", no_argument, &show_help, 1},
 100   {"version", no_argument, &show_version, 1},
 101   {NULL, 0, NULL, 0}
 102 };
 103
 104 /* Add tab stop TABVAL to the end of `tab_list', except
 105    if TABVAL is -1, do nothing. */
 106
 107 static void
 108 add_tabstop (int tabval)
 109 {
 110   if (tabval == -1)
 111     return;
 112   if (first_free_tab % TABLIST_BLOCK == 0)
 113     tab_list = (int *) xrealloc ((char *) tab_list,
 114                                  first_free_tab + TABLIST_BLOCK);
 115   tab_list[first_free_tab++] = tabval;
 116 }
 117
 118 /* Add the comma or blank separated list of tabstops STOPS
 119    to the list of tabstops. */
 120
 121 static void
 122 parse_tabstops (const char *stops)
 123 {
 124   int tabval = -1;
 125
 126   for (; *stops; stops++)
 127     {
 128       if (*stops == ',' || ISBLANK (*stops))
 129         {
 130           add_tabstop (tabval);
 131           tabval = -1;
 132         }
 133       else if (ISDIGIT (*stops))
 134         {
 135           if (tabval == -1)
 136             tabval = 0;
 137           tabval = tabval * 10 + *stops - '0';
 138         }
 139       else
 140         error (EXIT_FAILURE, 0, _("tab size contains an invalid character"));
 141     }
 142
 143   add_tabstop (tabval);
 144 }
 145
 146 /* Check that the list of tabstops TABS, with ENTRIES entries,
 147    contains only nonzero, ascending values. */
 148
 149 static void
 150 validate_tabstops (const int *tabs, int entries)
 151 {
 152   int prev_tab = 0;
 153   int i;
 154
 155   for (i = 0; i < entries; i++)
 156     {
 157       if (tabs[i] == 0)
 158         error (EXIT_FAILURE, 0, _("tab size cannot be 0"));
 159       if (tabs[i] <= prev_tab)
 160         error (EXIT_FAILURE, 0, _("tab sizes must be ascending"));
 161       prev_tab = tabs[i];
 162     }
 163 }
 164
 165 /* Close the old stream pointer FP if it is non-NULL,
 166    and return a new one opened to read the next input file.
 167    Open a filename of `-' as the standard input.
 168    Return NULL if there are no more input files.  */
 169
 170 static FILE *
 171 next_file (FILE *fp)
 172 {
 173   static char *prev_file;
 174   char *file;
 175
 176   if (fp)
 177     {
 178       if (ferror (fp))
 179         {
 180           error (0, errno, "%s", prev_file);
 181           exit_status = 1;
 182         }
 183       if (fp == stdin)
 184         clearerr (fp);          /* Also clear EOF. */
 185       else if (fclose (fp) == EOF)
 186         {
 187           error (0, errno, "%s", prev_file);
 188           exit_status = 1;
 189         }
 190     }
 191
 192   while ((file = *file_list++) != NULL)
 193     {
 194       if (file[0] == '-' && file[1] == '\0')
 195         {
 196           have_read_stdin = 1;
 197           prev_file = file;
 198           return stdin;
 199         }
 200       fp = fopen (file, "r");
 201       if (fp)
 202         {
 203           prev_file = file;
 204           return fp;
 205         }
 206       error (0, errno, "%s", file);
 207       exit_status = 1;
 208     }
 209   return NULL;
 210 }
 211
 212 /* Change spaces to tabs, writing to stdout.
 213    Read each file in `file_list', in order. */
 214
 215 static void
 216 unexpand (void)
 217 {
 218   FILE *fp;                     /* Input stream. */
 219   int c;                        /* Each input character. */
 220   /* Index in `tab_list' of next tabstop: */
 221   int tab_index = 0;            /* For calculating width of pending tabs. */
 222   int print_tab_index = 0;      /* For printing as many tabs as possible. */
 223   int column = 0;               /* Column on screen of next char. */
 224   int next_tab_column;          /* Column the next tab stop is on. */
 225   int convert = 1;              /* If nonzero, perform translations. */
 226   int pending = 0;              /* Pending columns of blanks. */
 227
 228   fp = next_file ((FILE *) NULL);
 229   if (fp == NULL)
 230     return;
 231
 232   for (;;)
 233     {
 234       c = getc (fp);
 235
 236       if (c == ' ' && convert)
 237         {
 238           ++pending;
 239           ++column;
 240         }
 241       else if (c == '\t' && convert)
 242         {
 243           if (tab_size == 0)
 244             {
 245               /* Do not let tab_index == first_free_tab;
 246                  stop when it is 1 less. */
 247               while (tab_index < first_free_tab - 1
 248                      && column >= tab_list[tab_index])
 249                 tab_index++;
 250               next_tab_column = tab_list[tab_index];
 251               if (tab_index < first_free_tab - 1)
 252                 tab_index++;
 253               if (column >= next_tab_column)
 254                 {
 255                   convert = 0;  /* Ran out of tab stops. */
 256                   goto flush_pend;
 257                 }
 258             }
 259           else
 260             {
 261               next_tab_column = column + tab_size - column % tab_size;
 262             }
 263           pending += next_tab_column - column;
 264           column = next_tab_column;
 265         }
 266       else
 267         {
 268         flush_pend:
 269           /* Flush pending spaces.  Print as many tabs as possible,
 270              then print the rest as spaces. */
 271           if (pending == 1)
 272             {
 273               putchar (' ');
 274               pending = 0;
 275             }
 276           column -= pending;
 277           while (pending != 0)
 278             {
 279               if (tab_size == 0)
 280                 {
 281                   /* Do not let print_tab_index == first_free_tab;
 282                      stop when it is 1 less. */
 283                   while (print_tab_index < first_free_tab - 1
 284                          && column >= tab_list[print_tab_index])
 285                     print_tab_index++;
 286                   next_tab_column = tab_list[print_tab_index];
 287                   if (print_tab_index < first_free_tab - 1)
 288                     print_tab_index++;
 289                 }
 290               else
 291                 {
 292                   next_tab_column = column + tab_size - column % tab_size;
 293                 }
 294               if (next_tab_column - column <= pending)
 295                 {
 296                   putchar ('\t');
 297                   pending -= next_tab_column - column;
 298                   column = next_tab_column;
 299                 }
 300               else
 301                 {
 302                   --print_tab_index;
 303                   column += pending;
 304                   while (pending != 0)
 305                     {
 306                       putchar (' ');
 307                       pending--;
 308                     }
 309                 }
 310             }
 311
 312           if (c == EOF)
 313             {
 314               fp = next_file (fp);
 315               if (fp == NULL)
 316                 break;          /* No more files. */
 317               else
 318                 continue;
 319             }
 320
 321           if (convert)
 322             {
 323               if (c == '\b')
 324                 {
 325                   if (column > 0)
 326                     --column;
 327                 }
 328               else
 329                 {
 330                   ++column;
 331                   if (convert_entire_line == 0)
 332                     convert = 0;
 333                 }
 334             }
 335
 336           putchar (c);
 337
 338           if (c == '\n')
 339             {
 340               tab_index = print_tab_index = 0;
 341               column = pending = 0;
 342               convert = 1;
 343             }
 344         }
 345     }
 346 }
 347
 348 static void
 349 usage (int status)
 350 {
 351   if (status != 0)
 352     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 353              program_name);
 354   else
 355     {
 356       printf (_("\
 357 Usage: %s [OPTION]... [FILE]...\n\
 358 "),
 359               program_name);
 360       printf (_("\
 361 Convert spaces in each FILE to tabs, writing to standard output.\n\
 362 With no FILE, or when FILE is -, read standard input.\n\
 363 \n\
 364   -a, --all           convert all whitespace, instead of initial whitespace\n\
 365   -t, --tabs=NUMBER   have tabs NUMBER characters apart instead of 8\n\
 366   -t, --tabs=LIST     use comma separated list of explicit tab positions\n\
 367       --help          display this help and exit\n\
 368       --version       output version information and exit\n\
 369 \n\
 370 Instead of -t NUMBER or -t LIST, -NUMBER or -LIST may be used.\n\
 371 "));
 372       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
 373     }
 374   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 375 }
 376
 377 int
 378 main (int argc, char **argv)
 379 {
 380   int tabval = -1;              /* Value of tabstop being read, or -1. */
 381   int c;                        /* Option character. */
 382
 383   program_name = argv[0];
 384   setlocale (LC_ALL, "");
 385   bindtextdomain (PACKAGE, LOCALEDIR);
 386   textdomain (PACKAGE);
 387
 388   have_read_stdin = 0;
 389   exit_status = 0;
 390   convert_entire_line = 0;
 391   tab_list = NULL;
 392   first_free_tab = 0;
 393
 394   while ((c = getopt_long (argc, argv, "at:,0123456789", longopts, NULL)) != -1)
 395     {
 396       switch (c)
 397         {
 398         case 0:
 399           break;
 400
 401         case '?':
 402           usage (1);
 403         case 'a':
 404           convert_entire_line = 1;
 405           break;
 406         case 't':
 407           convert_entire_line = 1;
 408           parse_tabstops (optarg);
 409           break;
 410         case ',':
 411           add_tabstop (tabval);
 412           tabval = -1;
 413           break;
 414         default:
 415           if (tabval == -1)
 416             tabval = 0;
 417           tabval = tabval * 10 + c - '0';
 418           break;
 419         }
 420     }
 421
 422   if (show_version)
 423     {
 424       printf ("unexpand (%s) %s\n", GNU_PACKAGE, VERSION);
 425       exit (EXIT_SUCCESS);
 426     }
 427
 428   if (show_help)
 429     usage (0);
 430
 431   add_tabstop (tabval);
 432
 433   validate_tabstops (tab_list, first_free_tab);
 434
 435   if (first_free_tab == 0)
 436     tab_size = 8;
 437   else if (first_free_tab == 1)
 438     tab_size = tab_list[0];
 439   else
 440     {
 441       /* Append a sentinel to the list of tab stop indices.  */
 442       add_tabstop (INT_MAX);
 443       tab_size = 0;
 444     }
 445
 446   if (optind == argc)
 447     file_list = stdin_argv;
 448   else
 449     file_list = &argv[optind];
 450
 451   unexpand ();
 452
 453   if (have_read_stdin && fclose (stdin) == EOF)
 454     error (EXIT_FAILURE, errno, "-");
 455   if (fclose (stdout) == EOF)
 456     error (EXIT_FAILURE, errno, _("write error"));
 457   exit (exit_status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 458 }