usr.bin/col/col.c

   1 /*-
   2  * Copyright (c) 1990, 1993, 1994
   3  *      The Regents of the University of California.  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  * @(#) Copyright (c) 1990, 1993, 1994 The Regents of the University of California.  All rights reserved.
  37  * @(#)col.c    8.5 (Berkeley) 5/4/95
  38  * $FreeBSD: src/usr.bin/col/col.c,v 1.6.6.4 2001/08/02 01:27:12 obrien Exp $
  39  * $DragonFly: src/usr.bin/col/col.c,v 1.5 2006/08/03 16:40:48 swildner Exp $
  40  */
  41
  42 #include <ctype.h>
  43 #include <err.h>
  44 #include <string.h>
  45 #include <stdio.h>
  46 #include <stdlib.h>
  47 #include <unistd.h>
  48 #include <locale.h>
  49
  50 #define BS      '\b'            /* backspace */
  51 #define TAB     '\t'            /* tab */
  52 #define SPACE   ' '             /* space */
  53 #define NL      '\n'            /* newline */
  54 #define CR      '\r'            /* carriage return */
  55 #define ESC     '\033'          /* escape */
  56 #define SI      '\017'          /* shift in to normal character set */
  57 #define SO      '\016'          /* shift out to alternate character set */
  58 #define VT      '\013'          /* vertical tab (aka reverse line feed) */
  59 #define RLF     '\007'          /* ESC-07 reverse line feed */
  60 #define RHLF    '\010'          /* ESC-010 reverse half-line feed */
  61 #define FHLF    '\011'          /* ESC-011 forward half-line feed */
  62
  63 /* build up at least this many lines before flushing them out */
  64 #define BUFFER_MARGIN           32
  65
  66 typedef char CSET;
  67
  68 typedef struct char_str {
  69 #define CS_NORMAL       1
  70 #define CS_ALTERNATE    2
  71         short           c_column;       /* column character is in */
  72         CSET            c_set;          /* character set (currently only 2) */
  73         char            c_char;         /* character in question */
  74 } CHAR;
  75
  76 typedef struct line_str LINE;
  77 struct line_str {
  78         CHAR    *l_line;                /* characters on the line */
  79         LINE    *l_prev;                /* previous line */
  80         LINE    *l_next;                /* next line */
  81         int     l_lsize;                /* allocated sizeof l_line */
  82         int     l_line_len;             /* strlen(l_line) */
  83         int     l_needs_sort;           /* set if chars went in out of order */
  84         int     l_max_col;              /* max column in the line */
  85 };
  86
  87 LINE   *alloc_line(void);
  88 void    dowarn(int);
  89 void    flush_line(LINE *);
  90 void    flush_lines(int);
  91 void    flush_blanks(void);
  92 void    free_line(LINE *);
  93 int     main(int, char **);
  94 void    usage(void);
  95
  96 CSET    last_set;               /* char_set of last char printed */
  97 LINE   *lines;
  98 int     compress_spaces;        /* if doing space -> tab conversion */
  99 int     fine;                   /* if `fine' resolution (half lines) */
 100 int     max_bufd_lines;         /* max # lines to keep in memory */
 101 int     nblank_lines;           /* # blanks after last flushed line */
 102 int     no_backspaces;          /* if not to output any backspaces */
 103 int     pass_unknown_seqs;      /* pass unknown control sequences */
 104
 105 #define PUTC(ch) \
 106         do {                                    \
 107                 if (putchar(ch) == EOF)         \
 108                         errx(1, "write error"); \
 109         } while (0)
 110
 111 int
 112 main(int argc, char **argv)
 113 {
 114         int ch;
 115         CHAR *c;
 116         CSET cur_set;                   /* current character set */
 117         LINE *l;                        /* current line */
 118         int extra_lines;                /* # of lines above first line */
 119         int cur_col;                    /* current column */
 120         int cur_line;                   /* line number of current position */
 121         int max_line;                   /* max value of cur_line */
 122         int this_line;                  /* line l points to */
 123         int nflushd_lines;              /* number of lines that were flushed */
 124         int adjust, opt, warned;
 125
 126         (void)setlocale(LC_CTYPE, "");
 127
 128         max_bufd_lines = 128;
 129         compress_spaces = 1;            /* compress spaces into tabs */
 130         while ((opt = getopt(argc, argv, "bfhl:px")) != -1)
 131                 switch (opt) {
 132                 case 'b':               /* do not output backspaces */
 133                         no_backspaces = 1;
 134                         break;
 135                 case 'f':               /* allow half forward line feeds */
 136                         fine = 1;
 137                         break;
 138                 case 'h':               /* compress spaces into tabs */
 139                         compress_spaces = 1;
 140                         break;
 141                 case 'l':               /* buffered line count */
 142                         if ((max_bufd_lines = atoi(optarg)) <= 0)
 143                                 errx(1, "bad -l argument %s", optarg);
 144                         break;
 145                 case 'p':               /* pass unknown control sequences */
 146                         pass_unknown_seqs = 1;
 147                         break;
 148                 case 'x':               /* do not compress spaces into tabs */
 149                         compress_spaces = 0;
 150                         break;
 151                 case '?':
 152                 default:
 153                         usage();
 154                 }
 155
 156         if (optind != argc)
 157                 usage();
 158
 159         /* this value is in half lines */
 160         max_bufd_lines *= 2;
 161
 162         adjust = cur_col = extra_lines = warned = 0;
 163         cur_line = max_line = nflushd_lines = this_line = 0;
 164         cur_set = last_set = CS_NORMAL;
 165         lines = l = alloc_line();
 166
 167         while ((ch = getchar()) != EOF) {
 168                 if (!isgraph(ch)) {
 169                         switch (ch) {
 170                         case BS:                /* can't go back further */
 171                                 if (cur_col == 0)
 172                                         continue;
 173                                 --cur_col;
 174                                 continue;
 175                         case CR:
 176                                 cur_col = 0;
 177                                 continue;
 178                         case ESC:               /* just ignore EOF */
 179                                 switch(getchar()) {
 180                                 case RLF:
 181                                         cur_line -= 2;
 182                                         break;
 183                                 case RHLF:
 184                                         cur_line--;
 185                                         break;
 186                                 case FHLF:
 187                                         cur_line++;
 188                                         if (cur_line > max_line)
 189                                                 max_line = cur_line;
 190                                 }
 191                                 continue;
 192                         case NL:
 193                                 cur_line += 2;
 194                                 if (cur_line > max_line)
 195                                         max_line = cur_line;
 196                                 cur_col = 0;
 197                                 continue;
 198                         case SPACE:
 199                                 ++cur_col;
 200                                 continue;
 201                         case SI:
 202                                 cur_set = CS_NORMAL;
 203                                 continue;
 204                         case SO:
 205                                 cur_set = CS_ALTERNATE;
 206                                 continue;
 207                         case TAB:               /* adjust column */
 208                                 cur_col |= 7;
 209                                 ++cur_col;
 210                                 continue;
 211                         case VT:
 212                                 cur_line -= 2;
 213                                 continue;
 214                         }
 215                         if (!pass_unknown_seqs)
 216                                 continue;
 217                 }
 218
 219                 /* Must stuff ch in a line - are we at the right one? */
 220                 if (cur_line != this_line - adjust) {
 221                         LINE *lnew;
 222                         int nmove;
 223
 224                         adjust = 0;
 225                         nmove = cur_line - this_line;
 226                         if (!fine) {
 227                                 /* round up to next line */
 228                                 if (cur_line & 1) {
 229                                         adjust = 1;
 230                                         nmove++;
 231                                 }
 232                         }
 233                         if (nmove < 0) {
 234                                 for (; nmove < 0 && l->l_prev; nmove++)
 235                                         l = l->l_prev;
 236                                 if (nmove) {
 237                                         if (nflushd_lines == 0) {
 238                                                 /*
 239                                                  * Allow backup past first
 240                                                  * line if nothing has been
 241                                                  * flushed yet.
 242                                                  */
 243                                                 for (; nmove < 0; nmove++) {
 244                                                         lnew = alloc_line();
 245                                                         l->l_prev = lnew;
 246                                                         lnew->l_next = l;
 247                                                         l = lines = lnew;
 248                                                         extra_lines++;
 249                                                 }
 250                                         } else {
 251                                                 if (!warned++)
 252                                                         dowarn(cur_line);
 253                                                 cur_line -= nmove;
 254                                         }
 255                                 }
 256                         } else {
 257                                 /* may need to allocate here */
 258                                 for (; nmove > 0 && l->l_next; nmove--)
 259                                         l = l->l_next;
 260                                 for (; nmove > 0; nmove--) {
 261                                         lnew = alloc_line();
 262                                         lnew->l_prev = l;
 263                                         l->l_next = lnew;
 264                                         l = lnew;
 265                                 }
 266                         }
 267                         this_line = cur_line + adjust;
 268                         nmove = this_line - nflushd_lines;
 269                         if (nmove >= max_bufd_lines + BUFFER_MARGIN) {
 270                                 nflushd_lines += nmove - max_bufd_lines;
 271                                 flush_lines(nmove - max_bufd_lines);
 272                         }
 273                 }
 274                 /* grow line's buffer? */
 275                 if (l->l_line_len + 1 >= l->l_lsize) {
 276                         int need;
 277
 278                         need = l->l_lsize ? l->l_lsize * 2 : 90;
 279                         if ((l->l_line = realloc(l->l_line,
 280                             (unsigned)need * sizeof(CHAR))) == NULL)
 281                                 err(1, NULL);
 282                         l->l_lsize = need;
 283                 }
 284                 c = &l->l_line[l->l_line_len++];
 285                 c->c_char = ch;
 286                 c->c_set = cur_set;
 287                 c->c_column = cur_col;
 288                 /*
 289                  * If things are put in out of order, they will need sorting
 290                  * when it is flushed.
 291                  */
 292                 if (cur_col < l->l_max_col)
 293                         l->l_needs_sort = 1;
 294                 else
 295                         l->l_max_col = cur_col;
 296                 cur_col++;
 297         }
 298         if (max_line == 0)
 299                 exit(0);        /* no lines, so just exit */
 300
 301         /* goto the last line that had a character on it */
 302         for (; l->l_next; l = l->l_next)
 303                 this_line++;
 304         flush_lines(this_line - nflushd_lines + extra_lines + 1);
 305
 306         /* make sure we leave things in a sane state */
 307         if (last_set != CS_NORMAL)
 308                 PUTC('\017');
 309
 310         /* flush out the last few blank lines */
 311         nblank_lines = max_line - this_line;
 312         if (max_line & 1)
 313                 nblank_lines++;
 314         else if (!nblank_lines)
 315                 /* missing a \n on the last line? */
 316                 nblank_lines = 2;
 317         flush_blanks();
 318         exit(0);
 319 }
 320
 321 void
 322 flush_lines(int nflush)
 323 {
 324         LINE *l;
 325
 326         while (--nflush >= 0) {
 327                 l = lines;
 328                 lines = l->l_next;
 329                 if (l->l_line) {
 330                         flush_blanks();
 331                         flush_line(l);
 332                 }
 333                 nblank_lines++;
 334                 if (l->l_line)
 335                         (void)free(l->l_line);
 336                 free_line(l);
 337         }
 338         if (lines)
 339                 lines->l_prev = NULL;
 340 }
 341
 342 /*
 343  * Print a number of newline/half newlines.  If fine flag is set, nblank_lines
 344  * is the number of half line feeds, otherwise it is the number of whole line
 345  * feeds.
 346  */
 347 void
 348 flush_blanks(void)
 349 {
 350         int half, i, nb;
 351
 352         half = 0;
 353         nb = nblank_lines;
 354         if (nb & 1) {
 355                 if (fine)
 356                         half = 1;
 357                 else
 358                         nb++;
 359         }
 360         nb /= 2;
 361         for (i = nb; --i >= 0;)
 362                 PUTC('\n');
 363         if (half) {
 364                 PUTC('\033');
 365                 PUTC('9');
 366                 if (!nb)
 367                         PUTC('\r');
 368         }
 369         nblank_lines = 0;
 370 }
 371
 372 /*
 373  * Write a line to stdout taking care of space to tab conversion (-h flag)
 374  * and character set shifts.
 375  */
 376 void
 377 flush_line(LINE *l)
 378 {
 379         CHAR *c, *endc;
 380         int nchars, last_col, this_col;
 381
 382         last_col = 0;
 383         nchars = l->l_line_len;
 384
 385         if (l->l_needs_sort) {
 386                 static CHAR *sorted;
 387                 static int count_size, *count, i, save, sorted_size, tot;
 388
 389                 /*
 390                  * Do an O(n) sort on l->l_line by column being careful to
 391                  * preserve the order of characters in the same column.
 392                  */
 393                 if (l->l_lsize > sorted_size) {
 394                         sorted_size = l->l_lsize;
 395                         if ((sorted = realloc(sorted,
 396                             (unsigned)sizeof(CHAR) * sorted_size)) == NULL)
 397                                 err(1, NULL);
 398                 }
 399                 if (l->l_max_col >= count_size) {
 400                         count_size = l->l_max_col + 1;
 401                         if ((count = realloc(count,
 402                             (unsigned)sizeof(int) * count_size)) == NULL)
 403                                 err(1, NULL);
 404                 }
 405                 memset(count, 0, sizeof(int) * l->l_max_col + 1);
 406                 for (i = nchars, c = l->l_line; --i >= 0; c++)
 407                         count[c->c_column]++;
 408
 409                 /*
 410                  * calculate running total (shifted down by 1) to use as
 411                  * indices into new line.
 412                  */
 413                 for (tot = 0, i = 0; i <= l->l_max_col; i++) {
 414                         save = count[i];
 415                         count[i] = tot;
 416                         tot += save;
 417                 }
 418
 419                 for (i = nchars, c = l->l_line; --i >= 0; c++)
 420                         sorted[count[c->c_column]++] = *c;
 421                 c = sorted;
 422         } else
 423                 c = l->l_line;
 424         while (nchars > 0) {
 425                 this_col = c->c_column;
 426                 endc = c;
 427                 do {
 428                         ++endc;
 429                 } while (--nchars > 0 && this_col == endc->c_column);
 430
 431                 /* if -b only print last character */
 432                 if (no_backspaces)
 433                         c = endc - 1;
 434
 435                 if (this_col > last_col) {
 436                         int nspace = this_col - last_col;
 437
 438                         if (compress_spaces && nspace > 1) {
 439                                 while (1) {
 440                                         int tab_col, tab_size;
 441
 442                                         tab_col = (last_col + 8) & ~7;
 443                                         if (tab_col > this_col)
 444                                                 break;
 445                                         tab_size = tab_col - last_col;
 446                                         if (tab_size == 1)
 447                                                 PUTC(' ');
 448                                         else
 449                                                 PUTC('\t');
 450                                         nspace -= tab_size;
 451                                         last_col = tab_col;
 452                                 }
 453                         }
 454                         while (--nspace >= 0)
 455                                 PUTC(' ');
 456                         last_col = this_col;
 457                 }
 458                 last_col++;
 459
 460                 for (;;) {
 461                         if (c->c_set != last_set) {
 462                                 switch (c->c_set) {
 463                                 case CS_NORMAL:
 464                                         PUTC('\017');
 465                                         break;
 466                                 case CS_ALTERNATE:
 467                                         PUTC('\016');
 468                                 }
 469                                 last_set = c->c_set;
 470                         }
 471                         PUTC(c->c_char);
 472                         if (++c >= endc)
 473                                 break;
 474                         PUTC('\b');
 475                 }
 476         }
 477 }
 478
 479 #define NALLOC 64
 480
 481 static LINE *line_freelist;
 482
 483 LINE *
 484 alloc_line(void)
 485 {
 486         LINE *l;
 487         int i;
 488
 489         if (!line_freelist) {
 490                 if ((l = realloc(NULL, sizeof(LINE) * NALLOC)) == NULL)
 491                         err(1, NULL);
 492                 line_freelist = l;
 493                 for (i = 1; i < NALLOC; i++, l++)
 494                         l->l_next = l + 1;
 495                 l->l_next = NULL;
 496         }
 497         l = line_freelist;
 498         line_freelist = l->l_next;
 499
 500         memset(l, 0, sizeof(LINE));
 501         return (l);
 502 }
 503
 504 void
 505 free_line(LINE *l)
 506 {
 507
 508         l->l_next = line_freelist;
 509         line_freelist = l;
 510 }
 511
 512 void
 513 usage(void)
 514 {
 515
 516         (void)fprintf(stderr, "usage: col [-bfhpx] [-l nline]\n");
 517         exit(1);
 518 }
 519
 520 void
 521 dowarn(int line)
 522 {
 523
 524         warnx("warning: can't back up %s",
 525                 line < 0 ? "past first line" : "-- line already flushed");
 526 }