src/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. Neither the name of the University nor the names of its contributors
  17  *    may be used to endorse or promote products derived from this software
  18  *    without specific prior written permission.
  19  *
  20  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  21  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  23  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  24  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  25  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  26  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  27  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  28  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  29  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  30  * SUCH DAMAGE.
  31  */
  32
  33 /* 2002-05-22   David A. Willis: modified license to the 'BSD no advert'
  34  *              license, as required by the Director of the Office of
  35  *              Technology Licensing of the University of California on
  36  *              July 22, 1999.
  37  *              See http://www.opensource.org/licenses/bsd-license.html
  38  * 2002-05-22   David A. Willis: "Ported" code for compilation under
  39  *              ix86-pc-cygwin (just added the #if HAVE_CONFIG_H block,
  40  *              really.)
  41  */
  42
  43 #ifndef lint
  44 static char copyright[] =
  45 "@(#) Copyright (c) 1990, 1993, 1994\n\
  46         The Regents of the University of California.  All rights reserved.\n";
  47 #endif /* not lint */
  48
  49 #ifndef lint
  50 static char sccsid[] = "@(#)col.c       8.3 (Berkeley) 4/2/94";
  51 #endif /* not lint */
  52
  53 #if HAVE_CONFIG_H
  54 #include "config.h"
  55 #endif
  56 #include "common.h"
  57
  58 #include "err.h"
  59 #include <locale.h>
  60
  61 /* These headers are included via common.h, except for <err.h>  */
  62 /* err.h is located in CWD, so include using " " instead of < > */
  63 #if 0
  64 #include <ctype.h>
  65 #include <err.h>
  66 #include <string.h>
  67 #include <stdio.h>
  68 #include <stdlib.h>
  69 #include <locale.h>
  70
  71 #ifdef HAVE_UNISTD_H
  72 #include <unistd.h>
  73 #endif /* HAVE_UNISTD_H */
  74
  75 #endif
  76
  77 #define BS      '\b'            /* backspace */
  78 #define TAB     '\t'            /* tab */
  79 #define SPACE   ' '             /* space */
  80 #define NL      '\n'            /* newline */
  81 #define CR      '\r'            /* carriage return */
  82 #define ESC     '\033'          /* escape */
  83 #define SI      '\017'          /* shift in to normal character set */
  84 #define SO      '\016'          /* shift out to alternate character set */
  85 #define VT      '\013'          /* vertical tab (aka reverse line feed) */
  86 #define RLF     '7'             /* ESC-7 reverse line feed */
  87 #define RHLF    '8'             /* ESC-8 reverse half-line feed */
  88 #define FHLF    '9'             /* ESC-9 forward half-line feed */
  89
  90 /* build up at least this many lines before flushing them out */
  91 #define BUFFER_MARGIN           32
  92
  93 typedef char CSET;
  94
  95 typedef struct char_str {
  96 #define CS_NORMAL       1
  97 #define CS_ALTERNATE    2
  98         short           c_column;       /* column character is in */
  99         CSET            c_set;          /* character set (currently only 2) */
 100         char            c_char;         /* character in question */
 101 } CHARACTER;
 102
 103 typedef struct line_str LINE;
 104 struct line_str {
 105         CHARACTER       *l_line;                /* characters on the line */
 106         LINE    *l_prev;                /* previous line */
 107         LINE    *l_next;                /* next line */
 108         int     l_lsize;                /* allocated sizeof l_line */
 109         int     l_line_len;             /* strlen(l_line) */
 110         int     l_needs_sort;           /* set if chars went in out of order */
 111         int     l_max_col;              /* max column in the line */
 112 };
 113
 114 LINE   *alloc_line __P((void));
 115 void    dowarn __P((int));
 116 void    flush_line __P((LINE *));
 117 void    flush_lines __P((int));
 118 void    flush_blanks __P((void));
 119 void    free_line __P((LINE *));
 120 void    usage __P((void));
 121 void    wrerr __P((void));
 122 void   *xmalloc __P((void *, size_t));
 123
 124 CSET    last_set;               /* char_set of last char printed */
 125 LINE   *lines;
 126 int     compress_spaces;        /* if doing space -> tab conversion */
 127 int     fine;                   /* if `fine' resolution (half lines) */
 128 int     max_bufd_lines;         /* max # lines to keep in memory */
 129 int     nblank_lines;           /* # blanks after last flushed line */
 130 int     no_backspaces;          /* if not to output any backspaces */
 131
 132 #define PUTC(ch) \
 133         if (putchar(ch) == EOF) \
 134                 wrerr();
 135
 136 int
 137 main(argc, argv)
 138         int argc;
 139         char **argv;
 140 {
 141         int ch;
 142         CHARACTER *c;
 143         CSET cur_set;                   /* current character set */
 144         LINE *l;                        /* current line */
 145         int extra_lines;                /* # of lines above first line */
 146         int cur_col;                    /* current column */
 147         int cur_line;                   /* line number of current position */
 148         int max_line;                   /* max value of cur_line */
 149         int this_line;                  /* line l points to */
 150         int nflushd_lines;              /* number of lines that were flushed */
 151         int adjust, opt, warned;
 152
 153         err_setprogname(argv[0]);
 154         setlocale(LC_ALL, "");
 155
 156         max_bufd_lines = 128;
 157         compress_spaces = 1;            /* compress spaces into tabs */
 158         while ((opt = getopt(argc, argv, "bfhl:x")) != EOF)
 159                 switch (opt) {
 160                 case 'b':               /* do not output backspaces */
 161                         no_backspaces = 1;
 162                         break;
 163                 case 'f':               /* allow half forward line feeds */
 164                         fine = 1;
 165                         break;
 166                 case 'h':               /* compress spaces into tabs */
 167                         compress_spaces = 1;
 168                         break;
 169                 case 'l':               /* buffered line count */
 170                         if ((max_bufd_lines = atoi(optarg)) <= 0) {
 171                                 (void)fprintf(stderr,
 172                                     "col: bad -l argument %s.\n", optarg);
 173                                 exit(1);
 174                         }
 175                         break;
 176                 case 'x':               /* do not compress spaces into tabs */
 177                         compress_spaces = 0;
 178                         break;
 179                 case '?':
 180                 default:
 181                         usage();
 182                 }
 183
 184         if (optind != argc)
 185                 usage();
 186
 187         /* this value is in half lines */
 188         max_bufd_lines *= 2;
 189
 190         adjust = cur_col = extra_lines = warned = 0;
 191         cur_line = max_line = nflushd_lines = this_line = 0;
 192         cur_set = last_set = CS_NORMAL;
 193         lines = l = alloc_line();
 194
 195         while ((ch = getchar()) != EOF) {
 196                 if (!isgraph(ch)) {
 197                         switch (ch) {
 198                         case BS:                /* can't go back further */
 199                                 if (cur_col == 0)
 200                                         continue;
 201                                 --cur_col;
 202                                 continue;
 203                         case CR:
 204                                 cur_col = 0;
 205                                 continue;
 206                         case ESC:               /* just ignore EOF */
 207                                 switch(getchar()) {
 208                                 case RLF:
 209                                         cur_line -= 2;
 210                                         break;
 211                                 case RHLF:
 212                                         cur_line--;
 213                                         break;
 214                                 case FHLF:
 215                                         cur_line++;
 216                                         if (cur_line > max_line)
 217                                                 max_line = cur_line;
 218                                 }
 219                                 continue;
 220                         case NL:
 221                                 cur_line += 2;
 222                                 if (cur_line > max_line)
 223                                         max_line = cur_line;
 224                                 cur_col = 0;
 225                                 continue;
 226                         case SPACE:
 227                                 ++cur_col;
 228                                 continue;
 229                         case SI:
 230                                 cur_set = CS_NORMAL;
 231                                 continue;
 232                         case SO:
 233                                 cur_set = CS_ALTERNATE;
 234                                 continue;
 235                         case TAB:               /* adjust column */
 236                                 cur_col |= 7;
 237                                 ++cur_col;
 238                                 continue;
 239                         case VT:
 240                                 cur_line -= 2;
 241                                 continue;
 242                         }
 243                         continue;
 244                 }
 245
 246                 /* Must stuff ch in a line - are we at the right one? */
 247                 if (cur_line != this_line - adjust) {
 248                         LINE *lnew;
 249                         int nmove;
 250
 251                         adjust = 0;
 252                         nmove = cur_line - this_line;
 253                         if (!fine) {
 254                                 /* round up to next line */
 255                                 if (cur_line & 1) {
 256                                         adjust = 1;
 257                                         nmove++;
 258                                 }
 259                         }
 260                         if (nmove < 0) {
 261                                 for (; nmove < 0 && l->l_prev; nmove++)
 262                                         l = l->l_prev;
 263                                 if (nmove) {
 264                                         if (nflushd_lines == 0) {
 265                                                 /*
 266                                                  * Allow backup past first
 267                                                  * line if nothing has been
 268                                                  * flushed yet.
 269                                                  */
 270                                                 for (; nmove < 0; nmove++) {
 271                                                         lnew = alloc_line();
 272                                                         l->l_prev = lnew;
 273                                                         lnew->l_next = l;
 274                                                         l = lines = lnew;
 275                                                         extra_lines++;
 276                                                 }
 277                                         } else {
 278                                                 if (!warned++)
 279                                                         dowarn(cur_line);
 280                                                 cur_line -= nmove;
 281                                         }
 282                                 }
 283                         } else {
 284                                 /* may need to allocate here */
 285                                 for (; nmove > 0 && l->l_next; nmove--)
 286                                         l = l->l_next;
 287                                 for (; nmove > 0; nmove--) {
 288                                         lnew = alloc_line();
 289                                         lnew->l_prev = l;
 290                                         l->l_next = lnew;
 291                                         l = lnew;
 292                                 }
 293                         }
 294                         this_line = cur_line + adjust;
 295                         nmove = this_line - nflushd_lines;
 296                         if (nmove >= max_bufd_lines + BUFFER_MARGIN) {
 297                                 nflushd_lines += nmove - max_bufd_lines;
 298                                 flush_lines(nmove - max_bufd_lines);
 299                         }
 300                 }
 301                 /* grow line's buffer? */
 302                 if (l->l_line_len + 1 >= l->l_lsize) {
 303                         int need;
 304
 305                         need = l->l_lsize ? l->l_lsize * 2 : 90;
 306                         l->l_line = (CHARACTER *)xmalloc((void *) l->l_line,
 307                             (unsigned) need * sizeof(CHARACTER));
 308                         l->l_lsize = need;
 309                 }
 310                 c = &l->l_line[l->l_line_len++];
 311                 c->c_char = ch;
 312                 c->c_set = cur_set;
 313                 c->c_column = cur_col;
 314                 /*
 315                  * If things are put in out of order, they will need sorting
 316                  * when it is flushed.
 317                  */
 318                 if (cur_col < l->l_max_col)
 319                         l->l_needs_sort = 1;
 320                 else
 321                         l->l_max_col = cur_col;
 322                 cur_col++;
 323         }
 324         /* goto the last line that had a character on it */
 325         for (; l->l_next; l = l->l_next)
 326                 this_line++;
 327         flush_lines(this_line - nflushd_lines + extra_lines + 1);
 328
 329         /* make sure we leave things in a sane state */
 330         if (last_set != CS_NORMAL)
 331                 PUTC('\017');
 332
 333         /* flush out the last few blank lines */
 334         nblank_lines = max_line - this_line;
 335         if (max_line & 1)
 336                 nblank_lines++;
 337         else if (!nblank_lines)
 338                 /* missing a \n on the last line? */
 339                 nblank_lines = 2;
 340         flush_blanks();
 341         exit(0);
 342 }
 343
 344 void
 345 flush_lines(nflush)
 346         int nflush;
 347 {
 348         LINE *l;
 349
 350         while (--nflush >= 0) {
 351                 l = lines;
 352                 lines = l->l_next;
 353                 if (l->l_line) {
 354                         flush_blanks();
 355                         flush_line(l);
 356                 }
 357                 nblank_lines++;
 358                 if (l->l_line)
 359                         (void)free((void *)l->l_line);
 360                 free_line(l);
 361         }
 362         if (lines)
 363                 lines->l_prev = NULL;
 364 }
 365
 366 /*
 367  * Print a number of newline/half newlines.  If fine flag is set, nblank_lines
 368  * is the number of half line feeds, otherwise it is the number of whole line
 369  * feeds.
 370  */
 371 void
 372 flush_blanks()
 373 {
 374         int half, i, nb;
 375
 376         half = 0;
 377         nb = nblank_lines;
 378         if (nb & 1) {
 379                 if (fine)
 380                         half = 1;
 381                 else
 382                         nb++;
 383         }
 384         nb /= 2;
 385         for (i = nb; --i >= 0;)
 386                 PUTC('\n');
 387         if (half) {
 388                 PUTC('\033');
 389                 PUTC('9');
 390                 if (!nb)
 391                         PUTC('\r');
 392         }
 393         nblank_lines = 0;
 394 }
 395
 396 /*
 397  * Write a line to stdout taking care of space to tab conversion (-h flag)
 398  * and character set shifts.
 399  */
 400 void
 401 flush_line(l)
 402         LINE *l;
 403 {
 404         CHARACTER *c, *endc;
 405         int nchars, last_col, this_col;
 406
 407         last_col = 0;
 408         nchars = l->l_line_len;
 409
 410         if (l->l_needs_sort) {
 411                 static CHARACTER *sorted;
 412                 static int count_size, *count, i, save, sorted_size, tot;
 413
 414                 /*
 415                  * Do an O(n) sort on l->l_line by column being careful to
 416                  * preserve the order of characters in the same column.
 417                  */
 418                 if (l->l_lsize > sorted_size) {
 419                         sorted_size = l->l_lsize;
 420                         sorted = (CHARACTER *)xmalloc((void *)sorted,
 421                             (unsigned)sizeof(CHARACTER) * sorted_size);
 422                 }
 423                 if (l->l_max_col >= count_size) {
 424                         count_size = l->l_max_col + 1;
 425                         count = (int *)xmalloc((void *)count,
 426                             (unsigned)sizeof(int) * count_size);
 427                 }
 428                 memset((char *)count, 0, sizeof(int) * l->l_max_col + 1);
 429                 for (i = nchars, c = l->l_line; --i >= 0; c++)
 430                         count[c->c_column]++;
 431
 432                 /*
 433                  * calculate running total (shifted down by 1) to use as
 434                  * indices into new line.
 435                  */
 436                 for (tot = 0, i = 0; i <= l->l_max_col; i++) {
 437                         save = count[i];
 438                         count[i] = tot;
 439                         tot += save;
 440                 }
 441
 442                 for (i = nchars, c = l->l_line; --i >= 0; c++)
 443                         sorted[count[c->c_column]++] = *c;
 444                 c = sorted;
 445         } else
 446                 c = l->l_line;
 447         while (nchars > 0) {
 448                 this_col = c->c_column;
 449                 endc = c;
 450                 do {
 451                         ++endc;
 452                 } while (--nchars > 0 && this_col == endc->c_column);
 453
 454                 /* if -b only print last character */
 455                 if (no_backspaces)
 456                         c = endc - 1;
 457
 458                 if (this_col > last_col) {
 459                         int nspace = this_col - last_col;
 460
 461                         if (compress_spaces && nspace > 1) {
 462                                 int ntabs;
 463
 464                                 ntabs = this_col / 8 - last_col / 8;
 465                                 nspace -= ntabs * 8;
 466                                 while (--ntabs >= 0)
 467                                         PUTC('\t');
 468                         }
 469                         while (--nspace >= 0)
 470                                 PUTC(' ');
 471                         last_col = this_col;
 472                 }
 473                 last_col++;
 474
 475                 for (;;) {
 476                         if (c->c_set != last_set) {
 477                                 switch (c->c_set) {
 478                                 case CS_NORMAL:
 479                                         PUTC('\017');
 480                                         break;
 481                                 case CS_ALTERNATE:
 482                                         PUTC('\016');
 483                                 }
 484                                 last_set = c->c_set;
 485                         }
 486                         PUTC(c->c_char);
 487                         if (++c >= endc)
 488                                 break;
 489                         PUTC('\b');
 490                 }
 491         }
 492 }
 493
 494 #define NALLOC 64
 495
 496 static LINE *line_freelist;
 497
 498 LINE *
 499 alloc_line()
 500 {
 501         LINE *l;
 502         int i;
 503
 504         if (!line_freelist) {
 505                 l = (LINE *)xmalloc((void *)NULL, sizeof(LINE) * NALLOC);
 506                 line_freelist = l;
 507                 for (i = 1; i < NALLOC; i++, l++)
 508                         l->l_next = l + 1;
 509                 l->l_next = NULL;
 510         }
 511         l = line_freelist;
 512         line_freelist = l->l_next;
 513
 514         memset(l, 0, sizeof(LINE));
 515         return (l);
 516 }
 517
 518 void
 519 free_line(l)
 520         LINE *l;
 521 {
 522
 523         l->l_next = line_freelist;
 524         line_freelist = l;
 525 }
 526
 527 void *
 528 xmalloc(p, size)
 529         void *p;
 530         size_t size;
 531 {
 532
 533         if (!(p = (void *)realloc(p, size)))
 534                 err(1, NULL);
 535         return (p);
 536 }
 537
 538 void
 539 usage()
 540 {
 541
 542         (void)fprintf(stderr, "usage: col [-bfx] [-l nline]\n");
 543         exit(1);
 544 }
 545
 546 void
 547 wrerr()
 548 {
 549
 550         (void)fprintf(stderr, "col: write error.\n");
 551         exit(1);
 552 }
 553
 554 void
 555 dowarn(line)
 556         int line;
 557 {
 558
 559         warnx("warning: can't back up %s",
 560                 line < 0 ? "past first line" : "-- line already flushed");
 561 }