text-utils/col.c

   1 /*-
   2  * Copyright (c) 1990 The Regents of the University of California.
   3  * All rights reserved.
   4  *
   5  * This code is derived from software contributed to Berkeley by
   6  * Michael Rendell of the Memorial University of Newfoundland.
   7  *
   8  * Redistribution and use in source and binary forms, with or without
   9  * modification, are permitted provided that the following conditions
  10  * are met:
  11  * 1. Redistributions of source code must retain the above copyright
  12  *    notice, this list of conditions and the following disclaimer.
  13  * 2. Redistributions in binary form must reproduce the above copyright
  14  *    notice, this list of conditions and the following disclaimer in the
  15  *    documentation and/or other materials provided with the distribution.
  16  * 3. All advertising materials mentioning features or use of this software
  17  *    must display the following acknowledgement:
  18  *      This product includes software developed by the University of
  19  *      California, Berkeley and its contributors.
  20  * 4. Neither the name of the University nor the names of its contributors
  21  *    may be used to endorse or promote products derived from this software
  22  *    without specific prior written permission.
  23  *
  24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  34  * SUCH DAMAGE.
  35  *
  36  * Wed Jun 22 22:15:41 1994, faith@cs.unc.edu: Added internationalization
  37  *                           patches from Andries.Brouwer@cwi.nl
  38  * Wed Sep 14 22:31:17 1994: patches from Carl Christofferson
  39  *                           (cchris@connected.com)
  40  * 1999-02-22 Arkadiusz Miśkiewicz <misiek@pld.ORG.PL>
  41  *      added Native Language Support
  42  * 1999-09-19 Bruno Haible <haible@clisp.cons.org>
  43  *      modified to work correctly in multi-byte locales
  44  *
  45  */
  46
  47 /*
  48  * This command is deprecated.  The utility is in maintenance mode,
  49  * meaning we keep them in source tree for backward compatibility
  50  * only.  Do not waste time making this command better, unless the
  51  * fix is about security or other very critical issue.
  52  *
  53  * See Documentation/deprecated.txt for more information.
  54  */
  55
  56 #include <stdlib.h>
  57 #include <errno.h>
  58 #include <ctype.h>
  59 #include <string.h>
  60 #include <stdio.h>
  61 #include <unistd.h>
  62 #include <getopt.h>
  63
  64 #include "nls.h"
  65 #include "xalloc.h"
  66 #include "widechar.h"
  67 #include "strutils.h"
  68 #include "closestream.h"
  69
  70 #define BS      '\b'            /* backspace */
  71 #define TAB     '\t'            /* tab */
  72 #define SPACE   ' '             /* space */
  73 #define NL      '\n'            /* newline */
  74 #define CR      '\r'            /* carriage return */
  75 #define ESC     '\033'          /* escape */
  76 #define SI      '\017'          /* shift in to normal character set */
  77 #define SO      '\016'          /* shift out to alternate character set */
  78 #define VT      '\013'          /* vertical tab (aka reverse line feed) */
  79 #define RLF     '\007'          /* ESC-07 reverse line feed */
  80 #define RHLF    '\010'          /* ESC-010 reverse half-line feed */
  81 #define FHLF    '\011'          /* ESC-011 forward half-line feed */
  82
  83 /* build up at least this many lines before flushing them out */
  84 #define BUFFER_MARGIN           32
  85
  86 typedef char CSET;
  87
  88 typedef struct char_str {
  89 #define CS_NORMAL       1
  90 #define CS_ALTERNATE    2
  91         short           c_column;       /* column character is in */
  92         CSET            c_set;          /* character set (currently only 2) */
  93         wchar_t         c_char;         /* character in question */
  94         int             c_width;        /* character width */
  95 } CHAR;
  96
  97 typedef struct line_str LINE;
  98 struct line_str {
  99         CHAR    *l_line;                /* characters on the line */
 100         LINE    *l_prev;                /* previous line */
 101         LINE    *l_next;                /* next line */
 102         int     l_lsize;                /* allocated sizeof l_line */
 103         int     l_line_len;             /* strlen(l_line) */
 104         int     l_needs_sort;           /* set if chars went in out of order */
 105         int     l_max_col;              /* max column in the line */
 106 };
 107
 108 void free_line(LINE *l);
 109 void flush_line(LINE *l);
 110 void flush_lines(int);
 111 void flush_blanks(void);
 112 LINE *alloc_line(void);
 113
 114 static CSET last_set;                   /* char_set of last char printed */
 115 static LINE *lines;
 116 static int compress_spaces;             /* if doing space -> tab conversion */
 117 static int fine;                        /* if `fine' resolution (half lines) */
 118 static unsigned max_bufd_lines;         /* max # lines to keep in memory */
 119 static int nblank_lines;                /* # blanks after last flushed line */
 120 static int no_backspaces;               /* if not to output any backspaces */
 121 static int pass_unknown_seqs;           /* whether to pass unknown control sequences */
 122
 123 #define PUTC(ch) \
 124         if (putwchar(ch) == WEOF) \
 125                 wrerr();
 126
 127 static void __attribute__((__noreturn__)) usage(FILE *out)
 128 {
 129         fprintf(out, _(
 130                 "\nUsage:\n"
 131                 " %s [options]\n"), program_invocation_short_name);
 132
 133         fputs(USAGE_SEPARATOR, out);
 134         fputs(_("Filter out reverse line feeds.\n"), out);
 135
 136         fprintf(out, _(
 137                 "\nOptions:\n"
 138                 " -b, --no-backspaces    do not output backspaces\n"
 139                 " -f, --fine             permit forward half line feeds\n"
 140                 " -p, --pass             pass unknown control sequences\n"
 141                 " -h, --tabs             convert spaces to tabs\n"
 142                 " -x, --spaces           convert tabs to spaces\n"
 143                 " -l, --lines NUM        buffer at least NUM lines\n"
 144                 " -V, --version          output version information and exit\n"
 145                 " -H, --help             display this help and exit\n\n"));
 146
 147         fprintf(out, _(
 148                 "%s reads from standard input and writes to standard output\n\n"),
 149                 program_invocation_short_name);
 150
 151         fprintf(out, USAGE_MAN_TAIL("col(1)"));
 152         exit(out == stderr ? EXIT_FAILURE : EXIT_SUCCESS);
 153 }
 154
 155 static void __attribute__((__noreturn__)) wrerr(void)
 156 {
 157         errx(EXIT_FAILURE, _("write error"));
 158 }
 159
 160 int main(int argc, char **argv)
 161 {
 162         register wint_t ch;
 163         CHAR *c = NULL;
 164         CSET cur_set;                   /* current character set */
 165         LINE *l;                        /* current line */
 166         int extra_lines;                /* # of lines above first line */
 167         int cur_col;                    /* current column */
 168         int cur_line;                   /* line number of current position */
 169         int max_line;                   /* max value of cur_line */
 170         int this_line;                  /* line l points to */
 171         int nflushd_lines;              /* number of lines that were flushed */
 172         int adjust, opt, warned;
 173         int ret = EXIT_SUCCESS;
 174
 175         static const struct option longopts[] = {
 176                 { "no-backspaces", no_argument,         NULL, 'b' },
 177                 { "fine",          no_argument,         NULL, 'f' },
 178                 { "pass",          no_argument,         NULL, 'p' },
 179                 { "tabs",          no_argument,         NULL, 'h' },
 180                 { "spaces",        no_argument,         NULL, 'x' },
 181                 { "lines",         required_argument,   NULL, 'l' },
 182                 { "version",       no_argument,         NULL, 'V' },
 183                 { "help",          no_argument,         NULL, 'H' },
 184                 { NULL, 0, NULL, 0 }
 185         };
 186
 187         setlocale(LC_ALL, "");
 188         bindtextdomain(PACKAGE, LOCALEDIR);
 189         textdomain(PACKAGE);
 190         atexit(close_stdout);
 191
 192         max_bufd_lines = 128 * 2;
 193         compress_spaces = 1;            /* compress spaces into tabs */
 194         pass_unknown_seqs = 0;          /* remove unknown escape sequences */
 195
 196         while ((opt = getopt_long(argc, argv, "bfhl:pxVH", longopts, NULL)) != -1)
 197                 switch (opt) {
 198                 case 'b':               /* do not output backspaces */
 199                         no_backspaces = 1;
 200                         break;
 201                 case 'f':               /* allow half forward line feeds */
 202                         fine = 1;
 203                         break;
 204                 case 'h':               /* compress spaces into tabs */
 205                         compress_spaces = 1;
 206                         break;
 207                 case 'l':
 208                         /*
 209                          * Buffered line count, which is a value in half
 210                          * lines e.g. twice the amount specified.
 211                          */
 212                         max_bufd_lines = strtou32_or_err(optarg, _("bad -l argument")) * 2;
 213                         break;
 214                 case 'p':
 215                         pass_unknown_seqs = 1;
 216                         break;
 217                 case 'x':               /* do not compress spaces into tabs */
 218                         compress_spaces = 0;
 219                         break;
 220                 case 'V':
 221                         printf(UTIL_LINUX_VERSION);
 222                         return EXIT_SUCCESS;
 223                 case 'H':
 224                         usage(stdout);
 225                 default:
 226                         errtryhelp(EXIT_FAILURE);
 227                 }
 228
 229         if (optind != argc)
 230                 usage(stderr);
 231
 232         adjust = cur_col = extra_lines = warned = 0;
 233         cur_line = max_line = nflushd_lines = this_line = 0;
 234         cur_set = last_set = CS_NORMAL;
 235         lines = l = alloc_line();
 236
 237         while (feof(stdin) == 0) {
 238                 errno = 0;
 239                 if ((ch = getwchar()) == WEOF) {
 240                         if (errno == EILSEQ) {
 241                                 warn(NULL);
 242                                 ret = EXIT_FAILURE;
 243                         }
 244                         break;
 245                 }
 246                 if (!iswgraph(ch)) {
 247                         switch (ch) {
 248                         case BS:                /* can't go back further */
 249                                 if (cur_col == 0)
 250                                         continue;
 251                                 if (c)
 252                                         cur_col -= c->c_width;
 253                                 else
 254                                         cur_col--;
 255                                 continue;
 256                         case CR:
 257                                 cur_col = 0;
 258                                 continue;
 259                         case ESC:               /* just ignore EOF */
 260                                 switch(getwchar()) {
 261                                 case RLF:
 262                                         cur_line -= 2;
 263                                         break;
 264                                 case RHLF:
 265                                         cur_line--;
 266                                         break;
 267                                 case FHLF:
 268                                         cur_line++;
 269                                         if (cur_line > max_line)
 270                                                 max_line = cur_line;
 271                                 }
 272                                 continue;
 273                         case NL:
 274                                 cur_line += 2;
 275                                 if (cur_line > max_line)
 276                                         max_line = cur_line;
 277                                 cur_col = 0;
 278                                 continue;
 279                         case SPACE:
 280                                 ++cur_col;
 281                                 continue;
 282                         case SI:
 283                                 cur_set = CS_NORMAL;
 284                                 continue;
 285                         case SO:
 286                                 cur_set = CS_ALTERNATE;
 287                                 continue;
 288                         case TAB:               /* adjust column */
 289                                 cur_col |= 7;
 290                                 ++cur_col;
 291                                 continue;
 292                         case VT:
 293                                 cur_line -= 2;
 294                                 continue;
 295                         }
 296                         if (iswspace(ch)) {
 297                                 if (wcwidth(ch) > 0)
 298                                         cur_col += wcwidth(ch);
 299                                 continue;
 300                         }
 301                         if (!pass_unknown_seqs)
 302                                 continue;
 303                 }
 304
 305                 /* Must stuff ch in a line - are we at the right one? */
 306                 if (cur_line != this_line - adjust) {
 307                         LINE *lnew;
 308                         int nmove;
 309
 310                         adjust = 0;
 311                         nmove = cur_line - this_line;
 312                         if (!fine) {
 313                                 /* round up to next line */
 314                                 if (cur_line & 1) {
 315                                         adjust = 1;
 316                                         nmove++;
 317                                 }
 318                         }
 319                         if (nmove < 0) {
 320                                 for (; nmove < 0 && l->l_prev; nmove++)
 321                                         l = l->l_prev;
 322                                 if (nmove) {
 323                                         if (nflushd_lines == 0) {
 324                                                 /*
 325                                                  * Allow backup past first
 326                                                  * line if nothing has been
 327                                                  * flushed yet.
 328                                                  */
 329                                                 for (; nmove < 0; nmove++) {
 330                                                         lnew = alloc_line();
 331                                                         l->l_prev = lnew;
 332                                                         lnew->l_next = l;
 333                                                         l = lines = lnew;
 334                                                         extra_lines++;
 335                                                 }
 336                                         } else {
 337                                                 if (!warned++)
 338                                                         warnx(
 339                         _("warning: can't back up %s."), cur_line < 0 ?
 340                         _("past first line") : _("-- line already flushed"));
 341                                                 cur_line -= nmove;
 342                                         }
 343                                 }
 344                         } else {
 345                                 /* may need to allocate here */
 346                                 for (; nmove > 0 && l->l_next; nmove--)
 347                                         l = l->l_next;
 348                                 for (; nmove > 0; nmove--) {
 349                                         lnew = alloc_line();
 350                                         lnew->l_prev = l;
 351                                         l->l_next = lnew;
 352                                         l = lnew;
 353                                 }
 354                         }
 355                         this_line = cur_line + adjust;
 356                         nmove = this_line - nflushd_lines;
 357                         if (nmove > 0
 358                             && (unsigned) nmove >= max_bufd_lines + BUFFER_MARGIN) {
 359                                 nflushd_lines += nmove - max_bufd_lines;
 360                                 flush_lines(nmove - max_bufd_lines);
 361                         }
 362                 }
 363                 /* grow line's buffer? */
 364                 if (l->l_line_len + 1 >= l->l_lsize) {
 365                         int need;
 366
 367                         need = l->l_lsize ? l->l_lsize * 2 : 90;
 368                         l->l_line = xrealloc((void *) l->l_line,
 369                                                     (unsigned) need * sizeof(CHAR));
 370                         l->l_lsize = need;
 371                 }
 372                 c = &l->l_line[l->l_line_len++];
 373                 c->c_char = ch;
 374                 c->c_set = cur_set;
 375                 if (0 < cur_col)
 376                         c->c_column = cur_col;
 377                 else
 378                         c->c_column = 0;
 379                 c->c_width = wcwidth(ch);
 380                 /*
 381                  * If things are put in out of order, they will need sorting
 382                  * when it is flushed.
 383                  */
 384                 if (cur_col < l->l_max_col)
 385                         l->l_needs_sort = 1;
 386                 else
 387                         l->l_max_col = cur_col;
 388                 if (c->c_width > 0)
 389                         cur_col += c->c_width;
 390         }
 391         /* goto the last line that had a character on it */
 392         for (; l->l_next; l = l->l_next)
 393                 this_line++;
 394         if (max_line == 0)
 395                 return EXIT_SUCCESS;    /* no lines, so just exit */
 396         flush_lines(this_line - nflushd_lines + extra_lines + 1);
 397
 398         /* make sure we leave things in a sane state */
 399         if (last_set != CS_NORMAL)
 400                 PUTC('\017');
 401
 402         /* flush out the last few blank lines */
 403         nblank_lines = max_line - this_line;
 404         if (max_line & 1)
 405                 nblank_lines++;
 406         else if (!nblank_lines)
 407                 /* missing a \n on the last line? */
 408                 nblank_lines = 2;
 409         flush_blanks();
 410         return ret;
 411 }
 412
 413 void flush_lines(int nflush)
 414 {
 415         LINE *l;
 416
 417         while (--nflush >= 0) {
 418                 l = lines;
 419                 lines = l->l_next;
 420                 if (l->l_line) {
 421                         flush_blanks();
 422                         flush_line(l);
 423                 }
 424                 nblank_lines++;
 425                 free((void *)l->l_line);
 426                 free_line(l);
 427         }
 428         if (lines)
 429                 lines->l_prev = NULL;
 430 }
 431
 432 /*
 433  * Print a number of newline/half newlines.  If fine flag is set, nblank_lines
 434  * is the number of half line feeds, otherwise it is the number of whole line
 435  * feeds.
 436  */
 437 void flush_blanks(void)
 438 {
 439         int half, i, nb;
 440
 441         half = 0;
 442         nb = nblank_lines;
 443         if (nb & 1) {
 444                 if (fine)
 445                         half = 1;
 446                 else
 447                         nb++;
 448         }
 449         nb /= 2;
 450         for (i = nb; --i >= 0;)
 451                 PUTC('\n');
 452         if (half) {
 453                 PUTC('\033');
 454                 PUTC('9');
 455                 if (!nb)
 456                         PUTC('\r');
 457         }
 458         nblank_lines = 0;
 459 }
 460
 461 /*
 462  * Write a line to stdout taking care of space to tab conversion (-h flag)
 463  * and character set shifts.
 464  */
 465 void flush_line(LINE *l)
 466 {
 467         CHAR *c, *endc;
 468         int nchars, last_col, this_col;
 469
 470         last_col = 0;
 471         nchars = l->l_line_len;
 472
 473         if (l->l_needs_sort) {
 474                 static CHAR *sorted;
 475                 static int count_size, *count, i, save, sorted_size, tot;
 476
 477                 /*
 478                  * Do an O(n) sort on l->l_line by column being careful to
 479                  * preserve the order of characters in the same column.
 480                  */
 481                 if (l->l_lsize > sorted_size) {
 482                         sorted_size = l->l_lsize;
 483                         sorted = xrealloc((void *)sorted,
 484                                                   (unsigned)sizeof(CHAR) * sorted_size);
 485                 }
 486                 if (l->l_max_col >= count_size) {
 487                         count_size = l->l_max_col + 1;
 488                         count = (int *)xrealloc((void *)count,
 489                             (unsigned)sizeof(int) * count_size);
 490                 }
 491                 memset(count, 0, sizeof(int) * l->l_max_col + 1);
 492                 for (i = nchars, c = l->l_line; --i >= 0; c++)
 493                         count[c->c_column]++;
 494
 495                 /*
 496                  * calculate running total (shifted down by 1) to use as
 497                  * indices into new line.
 498                  */
 499                 for (tot = 0, i = 0; i <= l->l_max_col; i++) {
 500                         save = count[i];
 501                         count[i] = tot;
 502                         tot += save;
 503                 }
 504
 505                 for (i = nchars, c = l->l_line; --i >= 0; c++)
 506                         sorted[count[c->c_column]++] = *c;
 507                 c = sorted;
 508         } else
 509                 c = l->l_line;
 510         while (nchars > 0) {
 511                 this_col = c->c_column;
 512                 endc = c;
 513                 do {
 514                         ++endc;
 515                 } while (--nchars > 0 && this_col == endc->c_column);
 516
 517                 /* if -b only print last character */
 518                 if (no_backspaces) {
 519                         c = endc - 1;
 520                         if (nchars > 0 &&
 521                             this_col + c->c_width > endc->c_column)
 522                                 continue;
 523                 }
 524
 525                 if (this_col > last_col) {
 526                         int nspace = this_col - last_col;
 527
 528                         if (compress_spaces && nspace > 1) {
 529                                 int ntabs;
 530
 531                                 ntabs = this_col / 8 - last_col / 8;
 532                                 if (ntabs > 0) {
 533                                         nspace = this_col & 7;
 534                                         while (--ntabs >= 0)
 535                                                 PUTC('\t');
 536                                 }
 537                         }
 538                         while (--nspace >= 0)
 539                                 PUTC(' ');
 540                         last_col = this_col;
 541                 }
 542
 543                 for (;;) {
 544                         if (c->c_set != last_set) {
 545                                 switch (c->c_set) {
 546                                 case CS_NORMAL:
 547                                         PUTC('\017');
 548                                         break;
 549                                 case CS_ALTERNATE:
 550                                         PUTC('\016');
 551                                 }
 552                                 last_set = c->c_set;
 553                         }
 554                         PUTC(c->c_char);
 555                         if ((c + 1) < endc) {
 556                                 int i;
 557                                 for (i=0; i < c->c_width; i++)
 558                                         PUTC('\b');
 559                         }
 560                         if (++c >= endc)
 561                                 break;
 562                 }
 563                 last_col += (c - 1)->c_width;
 564         }
 565 }
 566
 567 #define NALLOC 64
 568
 569 static LINE *line_freelist;
 570
 571 LINE *
 572 alloc_line(void)
 573 {
 574         LINE *l;
 575         int i;
 576
 577         if (!line_freelist) {
 578                 l = xmalloc(sizeof(LINE) * NALLOC);
 579                 line_freelist = l;
 580                 for (i = 1; i < NALLOC; i++, l++)
 581                         l->l_next = l + 1;
 582                 l->l_next = NULL;
 583         }
 584         l = line_freelist;
 585         line_freelist = l->l_next;
 586
 587         memset(l, 0, sizeof(LINE));
 588         return l;
 589 }
 590
 591 void free_line(LINE *l)
 592 {
 593         l->l_next = line_freelist;
 594         line_freelist = l;
 595 }