col.c revision 24360 
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
37#ifndef lint
38static char copyright[] =
39"@(#) Copyright (c) 1990, 1993, 1994\n\
40	The Regents of the University of California.  All rights reserved.\n";
41#endif /* not lint */
42
43#ifndef lint
44static char sccsid[] = "@(#)col.c	8.5 (Berkeley) 5/4/95";
45#endif /* not lint */
46
47#include <ctype.h>
48#include <err.h>
49#include <string.h>
50#include <stdio.h>
51#include <stdlib.h>
52#include <unistd.h>
53#include <locale.h>
54
55#define	BS	'\b'		/* backspace */
56#define	TAB	'\t'		/* tab */
57#define	SPACE	' '		/* space */
58#define	NL	'\n'		/* newline */
59#define	CR	'\r'		/* carriage return */
60#define	ESC	'\033'		/* escape */
61#define	SI	'\017'		/* shift in to normal character set */
62#define	SO	'\016'		/* shift out to alternate character set */
63#define	VT	'\013'		/* vertical tab (aka reverse line feed) */
64#define	RLF	'\007'		/* ESC-07 reverse line feed */
65#define	RHLF	'\010'		/* ESC-010 reverse half-line feed */
66#define	FHLF	'\011'		/* ESC-011 forward half-line feed */
67
68/* build up at least this many lines before flushing them out */
69#define	BUFFER_MARGIN		32
70
71typedef char CSET;
72
73typedef struct char_str {
74#define	CS_NORMAL	1
75#define	CS_ALTERNATE	2
76	short		c_column;	/* column character is in */
77	CSET		c_set;		/* character set (currently only 2) */
78	char		c_char;		/* character in question */
79} CHAR;
80
81typedef struct line_str LINE;
82struct line_str {
83	CHAR	*l_line;		/* characters on the line */
84	LINE	*l_prev;		/* previous line */
85	LINE	*l_next;		/* next line */
86	int	l_lsize;		/* allocated sizeof l_line */
87	int	l_line_len;		/* strlen(l_line) */
88	int	l_needs_sort;		/* set if chars went in out of order */
89	int	l_max_col;		/* max column in the line */
90};
91
92LINE   *alloc_line __P((void));
93void	dowarn __P((int));
94void	flush_line __P((LINE *));
95void	flush_lines __P((int));
96void	flush_blanks __P((void));
97void	free_line __P((LINE *));
98void	usage __P((void));
99void	wrerr __P((void));
100void   *xmalloc __P((void *, size_t));
101
102CSET	last_set;		/* char_set of last char printed */
103LINE   *lines;
104int	compress_spaces;	/* if doing space -> tab conversion */
105int	fine;			/* if `fine' resolution (half lines) */
106int	max_bufd_lines;		/* max # lines to keep in memory */
107int	nblank_lines;		/* # blanks after last flushed line */
108int	no_backspaces;		/* if not to output any backspaces */
109
110#define	PUTC(ch) \
111	do {				\
112		if (putchar(ch) == EOF) \
113			wrerr();	\
114	} while (0)
115
116int
117main(argc, argv)
118	int argc;
119	char **argv;
120{
121	int ch;
122	CHAR *c;
123	CSET cur_set;			/* current character set */
124	LINE *l;			/* current line */
125	int extra_lines;		/* # of lines above first line */
126	int cur_col;			/* current column */
127	int cur_line;			/* line number of current position */
128	int max_line;			/* max value of cur_line */
129	int this_line;			/* line l points to */
130	int nflushd_lines;		/* number of lines that were flushed */
131	int adjust, opt, warned;
132
133	(void) setlocale(LC_CTYPE, "");
134
135	max_bufd_lines = 128;
136	compress_spaces = 1;		/* compress spaces into tabs */
137	while ((opt = getopt(argc, argv, "bfhl:x")) != -1)
138		switch (opt) {
139		case 'b':		/* do not output backspaces */
140			no_backspaces = 1;
141			break;
142		case 'f':		/* allow half forward line feeds */
143			fine = 1;
144			break;
145		case 'h':		/* compress spaces into tabs */
146			compress_spaces = 1;
147			break;
148		case 'l':		/* buffered line count */
149			if ((max_bufd_lines = atoi(optarg)) <= 0) {
150				(void)fprintf(stderr,
151				    "col: bad -l argument %s.\n", optarg);
152				exit(1);
153			}
154			break;
155		case 'x':		/* do not compress spaces into tabs */
156			compress_spaces = 0;
157			break;
158		case '?':
159		default:
160			usage();
161		}
162
163	if (optind != argc)
164		usage();
165
166	/* this value is in half lines */
167	max_bufd_lines *= 2;
168
169	adjust = cur_col = extra_lines = warned = 0;
170	cur_line = max_line = nflushd_lines = this_line = 0;
171	cur_set = last_set = CS_NORMAL;
172	lines = l = alloc_line();
173
174	while ((ch = getchar()) != EOF) {
175		if (!isgraph(ch)) {
176			switch (ch) {
177			case BS:		/* can't go back further */
178				if (cur_col == 0)
179					continue;
180				--cur_col;
181				continue;
182			case CR:
183				cur_col = 0;
184				continue;
185			case ESC:		/* just ignore EOF */
186				switch(getchar()) {
187				case RLF:
188					cur_line -= 2;
189					break;
190				case RHLF:
191					cur_line--;
192					break;
193				case FHLF:
194					cur_line++;
195					if (cur_line > max_line)
196						max_line = cur_line;
197				}
198				continue;
199			case NL:
200				cur_line += 2;
201				if (cur_line > max_line)
202					max_line = cur_line;
203				cur_col = 0;
204				continue;
205			case SPACE:
206				++cur_col;
207				continue;
208			case SI:
209				cur_set = CS_NORMAL;
210				continue;
211			case SO:
212				cur_set = CS_ALTERNATE;
213				continue;
214			case TAB:		/* adjust column */
215				cur_col |= 7;
216				++cur_col;
217				continue;
218			case VT:
219				cur_line -= 2;
220				continue;
221			}
222			continue;
223		}
224
225		/* Must stuff ch in a line - are we at the right one? */
226		if (cur_line != this_line - adjust) {
227			LINE *lnew;
228			int nmove;
229
230			adjust = 0;
231			nmove = cur_line - this_line;
232			if (!fine) {
233				/* round up to next line */
234				if (cur_line & 1) {
235					adjust = 1;
236					nmove++;
237				}
238			}
239			if (nmove < 0) {
240				for (; nmove < 0 && l->l_prev; nmove++)
241					l = l->l_prev;
242				if (nmove) {
243					if (nflushd_lines == 0) {
244						/*
245						 * Allow backup past first
246						 * line if nothing has been
247						 * flushed yet.
248						 */
249						for (; nmove < 0; nmove++) {
250							lnew = alloc_line();
251							l->l_prev = lnew;
252							lnew->l_next = l;
253							l = lines = lnew;
254							extra_lines++;
255						}
256					} else {
257						if (!warned++)
258							dowarn(cur_line);
259						cur_line -= nmove;
260					}
261				}
262			} else {
263				/* may need to allocate here */
264				for (; nmove > 0 && l->l_next; nmove--)
265					l = l->l_next;
266				for (; nmove > 0; nmove--) {
267					lnew = alloc_line();
268					lnew->l_prev = l;
269					l->l_next = lnew;
270					l = lnew;
271				}
272			}
273			this_line = cur_line + adjust;
274			nmove = this_line - nflushd_lines;
275			if (nmove >= max_bufd_lines + BUFFER_MARGIN) {
276				nflushd_lines += nmove - max_bufd_lines;
277				flush_lines(nmove - max_bufd_lines);
278			}
279		}
280		/* grow line's buffer? */
281		if (l->l_line_len + 1 >= l->l_lsize) {
282			int need;
283
284			need = l->l_lsize ? l->l_lsize * 2 : 90;
285			l->l_line = (CHAR *)xmalloc((void *) l->l_line,
286			    (unsigned) need * sizeof(CHAR));
287			l->l_lsize = need;
288		}
289		c = &l->l_line[l->l_line_len++];
290		c->c_char = ch;
291		c->c_set = cur_set;
292		c->c_column = cur_col;
293		/*
294		 * If things are put in out of order, they will need sorting
295		 * when it is flushed.
296		 */
297		if (cur_col < l->l_max_col)
298			l->l_needs_sort = 1;
299		else
300			l->l_max_col = cur_col;
301		cur_col++;
302	}
303	if (max_line == 0)
304		exit(0);	/* no lines, so just exit */
305
306	/* goto the last line that had a character on it */
307	for (; l->l_next; l = l->l_next)
308		this_line++;
309	flush_lines(this_line - nflushd_lines + extra_lines + 1);
310
311	/* make sure we leave things in a sane state */
312	if (last_set != CS_NORMAL)
313		PUTC('\017');
314
315	/* flush out the last few blank lines */
316	nblank_lines = max_line - this_line;
317	if (max_line & 1)
318		nblank_lines++;
319	else if (!nblank_lines)
320		/* missing a \n on the last line? */
321		nblank_lines = 2;
322	flush_blanks();
323	exit(0);
324}
325
326void
327flush_lines(nflush)
328	int nflush;
329{
330	LINE *l;
331
332	while (--nflush >= 0) {
333		l = lines;
334		lines = l->l_next;
335		if (l->l_line) {
336			flush_blanks();
337			flush_line(l);
338		}
339		nblank_lines++;
340		if (l->l_line)
341			(void)free((void *)l->l_line);
342		free_line(l);
343	}
344	if (lines)
345		lines->l_prev = NULL;
346}
347
348/*
349 * Print a number of newline/half newlines.  If fine flag is set, nblank_lines
350 * is the number of half line feeds, otherwise it is the number of whole line
351 * feeds.
352 */
353void
354flush_blanks()
355{
356	int half, i, nb;
357
358	half = 0;
359	nb = nblank_lines;
360	if (nb & 1) {
361		if (fine)
362			half = 1;
363		else
364			nb++;
365	}
366	nb /= 2;
367	for (i = nb; --i >= 0;)
368		PUTC('\n');
369	if (half) {
370		PUTC('\033');
371		PUTC('9');
372		if (!nb)
373			PUTC('\r');
374	}
375	nblank_lines = 0;
376}
377
378/*
379 * Write a line to stdout taking care of space to tab conversion (-h flag)
380 * and character set shifts.
381 */
382void
383flush_line(l)
384	LINE *l;
385{
386	CHAR *c, *endc;
387	int nchars, last_col, this_col;
388
389	last_col = 0;
390	nchars = l->l_line_len;
391
392	if (l->l_needs_sort) {
393		static CHAR *sorted;
394		static int count_size, *count, i, save, sorted_size, tot;
395
396		/*
397		 * Do an O(n) sort on l->l_line by column being careful to
398		 * preserve the order of characters in the same column.
399		 */
400		if (l->l_lsize > sorted_size) {
401			sorted_size = l->l_lsize;
402			sorted = (CHAR *)xmalloc((void *)sorted,
403			    (unsigned)sizeof(CHAR) * sorted_size);
404		}
405		if (l->l_max_col >= count_size) {
406			count_size = l->l_max_col + 1;
407			count = (int *)xmalloc((void *)count,
408			    (unsigned)sizeof(int) * count_size);
409		}
410		memset((char *)count, 0, sizeof(int) * l->l_max_col + 1);
411		for (i = nchars, c = l->l_line; --i >= 0; c++)
412			count[c->c_column]++;
413
414		/*
415		 * calculate running total (shifted down by 1) to use as
416		 * indices into new line.
417		 */
418		for (tot = 0, i = 0; i <= l->l_max_col; i++) {
419			save = count[i];
420			count[i] = tot;
421			tot += save;
422		}
423
424		for (i = nchars, c = l->l_line; --i >= 0; c++)
425			sorted[count[c->c_column]++] = *c;
426		c = sorted;
427	} else
428		c = l->l_line;
429	while (nchars > 0) {
430		this_col = c->c_column;
431		endc = c;
432		do {
433			++endc;
434		} while (--nchars > 0 && this_col == endc->c_column);
435
436		/* if -b only print last character */
437		if (no_backspaces)
438			c = endc - 1;
439
440		if (this_col > last_col) {
441			int nspace = this_col - last_col;
442
443			if (compress_spaces && nspace > 1) {
444				while (1) {
445					int tab_col, tab_size;;
446
447					tab_col = (last_col + 8) & ~7;
448					if (tab_col > this_col)
449						break;
450					tab_size = tab_col - last_col;
451					if (tab_size == 1)
452						PUTC(' ');
453					else
454						PUTC('\t');
455					nspace -= tab_size;
456					last_col = tab_col;
457				}
458			}
459			while (--nspace >= 0)
460				PUTC(' ');
461			last_col = this_col;
462		}
463		last_col++;
464
465		for (;;) {
466			if (c->c_set != last_set) {
467				switch (c->c_set) {
468				case CS_NORMAL:
469					PUTC('\017');
470					break;
471				case CS_ALTERNATE:
472					PUTC('\016');
473				}
474				last_set = c->c_set;
475			}
476			PUTC(c->c_char);
477			if (++c >= endc)
478				break;
479			PUTC('\b');
480		}
481	}
482}
483
484#define	NALLOC 64
485
486static LINE *line_freelist;
487
488LINE *
489alloc_line()
490{
491	LINE *l;
492	int i;
493
494	if (!line_freelist) {
495		l = (LINE *)xmalloc((void *)NULL, sizeof(LINE) * NALLOC);
496		line_freelist = l;
497		for (i = 1; i < NALLOC; i++, l++)
498			l->l_next = l + 1;
499		l->l_next = NULL;
500	}
501	l = line_freelist;
502	line_freelist = l->l_next;
503
504	memset(l, 0, sizeof(LINE));
505	return (l);
506}
507
508void
509free_line(l)
510	LINE *l;
511{
512
513	l->l_next = line_freelist;
514	line_freelist = l;
515}
516
517void *
518xmalloc(p, size)
519	void *p;
520	size_t size;
521{
522
523	if (!(p = (void *)realloc(p, size)))
524		err(1, NULL);
525	return (p);
526}
527
528void
529usage()
530{
531
532	(void)fprintf(stderr, "usage: col [-bfx] [-l nline]\n");
533	exit(1);
534}
535
536void
537wrerr()
538{
539
540	(void)fprintf(stderr, "col: write error.\n");
541	exit(1);
542}
543
544void
545dowarn(line)
546	int line;
547{
548
549	warnx("warning: can't back up %s",
550		line < 0 ? "past first line" : "-- line already flushed");
551}
552