col.c revision 282342 
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 * 4. 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#ifndef lint
34static const char copyright[] =
35"@(#) Copyright (c) 1990, 1993, 1994\n\
36	The Regents of the University of California.  All rights reserved.\n";
37#endif
38
39#if 0
40#ifndef lint
41static char sccsid[] = "@(#)col.c	8.5 (Berkeley) 5/4/95";
42#endif
43#endif
44
45#include <sys/cdefs.h>
46__FBSDID("$FreeBSD: head/usr.bin/col/col.c 282342 2015-05-02 12:22:24Z bapt $");
47
48#include <sys/capsicum.h>
49
50#include <err.h>
51#include <errno.h>
52#include <locale.h>
53#include <stdio.h>
54#include <stdlib.h>
55#include <string.h>
56#include <termios.h>
57#include <unistd.h>
58#include <wchar.h>
59#include <wctype.h>
60
61#define	BS	'\b'		/* backspace */
62#define	TAB	'\t'		/* tab */
63#define	SPACE	' '		/* space */
64#define	NL	'\n'		/* newline */
65#define	CR	'\r'		/* carriage return */
66#define	ESC	'\033'		/* escape */
67#define	SI	'\017'		/* shift in to normal character set */
68#define	SO	'\016'		/* shift out to alternate character set */
69#define	VT	'\013'		/* vertical tab (aka reverse line feed) */
70#define	RLF	'7'		/* ESC-7 reverse line feed */
71#define	RHLF	'8'		/* ESC-8 reverse half-line feed */
72#define	FHLF	'9'		/* ESC-9 forward half-line feed */
73
74/* build up at least this many lines before flushing them out */
75#define	BUFFER_MARGIN		32
76
77typedef char CSET;
78
79typedef struct char_str {
80#define	CS_NORMAL	1
81#define	CS_ALTERNATE	2
82	short		c_column;	/* column character is in */
83	CSET		c_set;		/* character set (currently only 2) */
84	wchar_t		c_char;		/* character in question */
85	int		c_width;	/* character width */
86} CHAR;
87
88typedef struct line_str LINE;
89struct line_str {
90	CHAR	*l_line;		/* characters on the line */
91	LINE	*l_prev;		/* previous line */
92	LINE	*l_next;		/* next line */
93	int	l_lsize;		/* allocated sizeof l_line */
94	int	l_line_len;		/* strlen(l_line) */
95	int	l_needs_sort;		/* set if chars went in out of order */
96	int	l_max_col;		/* max column in the line */
97};
98
99static LINE   *alloc_line(void);
100static void	dowarn(int);
101static void	flush_line(LINE *);
102static void	flush_lines(int);
103static void	flush_blanks(void);
104static void	free_line(LINE *);
105static void	usage(void);
106
107static CSET	last_set;		/* char_set of last char printed */
108static LINE    *lines;
109static int	compress_spaces;	/* if doing space -> tab conversion */
110static int	fine;			/* if `fine' resolution (half lines) */
111static int	max_bufd_lines;		/* max # lines to keep in memory */
112static int	nblank_lines;		/* # blanks after last flushed line */
113static int	no_backspaces;		/* if not to output any backspaces */
114static int	pass_unknown_seqs;	/* pass unknown control sequences */
115
116#define	PUTC(ch) \
117	do {					\
118		if (putwchar(ch) == WEOF)	\
119			errx(1, "write error");	\
120	} while (0)
121
122int
123main(int argc, char **argv)
124{
125	wint_t ch;
126	CHAR *c;
127	CSET cur_set;			/* current character set */
128	LINE *l;			/* current line */
129	int extra_lines;		/* # of lines above first line */
130	int cur_col;			/* current column */
131	int cur_line;			/* line number of current position */
132	int max_line;			/* max value of cur_line */
133	int this_line;			/* line l points to */
134	int nflushd_lines;		/* number of lines that were flushed */
135	int adjust, opt, warned, width;
136	cap_rights_t rights;
137	unsigned long cmd;
138
139	(void)setlocale(LC_CTYPE, "");
140
141	cap_rights_init(&rights, CAP_FSTAT, CAP_READ);
142	if (cap_rights_limit(STDIN_FILENO, &rights) < 0 && errno != ENOSYS)
143		err(1, "unable to limit rights for stdin");
144	cap_rights_init(&rights, CAP_FSTAT, CAP_WRITE, CAP_IOCTL);
145	if (cap_rights_limit(STDOUT_FILENO, &rights) < 0 && errno != ENOSYS)
146		err(1, "unable to limit rights for stdout");
147	cmd = TIOCGETA; /* required by isatty(3) in printf(3) */
148	if (cap_ioctls_limit(STDOUT_FILENO, &cmd, 1) < 0 && errno != ENOSYS)
149		err(1, "unable to limit ioctls for stdout");
150
151	if (cap_enter() < 0 && errno != ENOSYS)
152		err(1, "unable to enter capability mode");
153
154	max_bufd_lines = 128;
155	compress_spaces = 1;		/* compress spaces into tabs */
156	while ((opt = getopt(argc, argv, "bfhl:px")) != -1)
157		switch (opt) {
158		case 'b':		/* do not output backspaces */
159			no_backspaces = 1;
160			break;
161		case 'f':		/* allow half forward line feeds */
162			fine = 1;
163			break;
164		case 'h':		/* compress spaces into tabs */
165			compress_spaces = 1;
166			break;
167		case 'l':		/* buffered line count */
168			if ((max_bufd_lines = atoi(optarg)) <= 0)
169				errx(1, "bad -l argument %s", optarg);
170			break;
171		case 'p':		/* pass unknown control sequences */
172			pass_unknown_seqs = 1;
173			break;
174		case 'x':		/* do not compress spaces into tabs */
175			compress_spaces = 0;
176			break;
177		case '?':
178		default:
179			usage();
180		}
181
182	if (optind != argc)
183		usage();
184
185	/* this value is in half lines */
186	max_bufd_lines *= 2;
187
188	adjust = cur_col = extra_lines = warned = 0;
189	cur_line = max_line = nflushd_lines = this_line = 0;
190	cur_set = last_set = CS_NORMAL;
191	lines = l = alloc_line();
192
193	while ((ch = getwchar()) != WEOF) {
194		if (!iswgraph(ch)) {
195			switch (ch) {
196			case BS:		/* can't go back further */
197				if (cur_col == 0)
198					continue;
199				--cur_col;
200				continue;
201			case CR:
202				cur_col = 0;
203				continue;
204			case ESC:		/* just ignore EOF */
205				switch(getwchar()) {
206				case RLF:
207					cur_line -= 2;
208					break;
209				case RHLF:
210					cur_line--;
211					break;
212				case FHLF:
213					cur_line++;
214					if (cur_line > max_line)
215						max_line = cur_line;
216				}
217				continue;
218			case NL:
219				cur_line += 2;
220				if (cur_line > max_line)
221					max_line = cur_line;
222				cur_col = 0;
223				continue;
224			case SPACE:
225				++cur_col;
226				continue;
227			case SI:
228				cur_set = CS_NORMAL;
229				continue;
230			case SO:
231				cur_set = CS_ALTERNATE;
232				continue;
233			case TAB:		/* adjust column */
234				cur_col |= 7;
235				++cur_col;
236				continue;
237			case VT:
238				cur_line -= 2;
239				continue;
240			}
241			if (iswspace(ch)) {
242				if ((width = wcwidth(ch)) > 0)
243					cur_col += width;
244				continue;
245			}
246			if (!pass_unknown_seqs)
247				continue;
248		}
249
250		/* Must stuff ch in a line - are we at the right one? */
251		if (cur_line != this_line - adjust) {
252			LINE *lnew;
253			int nmove;
254
255			adjust = 0;
256			nmove = cur_line - this_line;
257			if (!fine) {
258				/* round up to next line */
259				if (cur_line & 1) {
260					adjust = 1;
261					nmove++;
262				}
263			}
264			if (nmove < 0) {
265				for (; nmove < 0 && l->l_prev; nmove++)
266					l = l->l_prev;
267				if (nmove) {
268					if (nflushd_lines == 0) {
269						/*
270						 * Allow backup past first
271						 * line if nothing has been
272						 * flushed yet.
273						 */
274						for (; nmove < 0; nmove++) {
275							lnew = alloc_line();
276							l->l_prev = lnew;
277							lnew->l_next = l;
278							l = lines = lnew;
279							extra_lines++;
280						}
281					} else {
282						if (!warned++)
283							dowarn(cur_line);
284						cur_line -= nmove;
285					}
286				}
287			} else {
288				/* may need to allocate here */
289				for (; nmove > 0 && l->l_next; nmove--)
290					l = l->l_next;
291				for (; nmove > 0; nmove--) {
292					lnew = alloc_line();
293					lnew->l_prev = l;
294					l->l_next = lnew;
295					l = lnew;
296				}
297			}
298			this_line = cur_line + adjust;
299			nmove = this_line - nflushd_lines;
300			if (nmove >= max_bufd_lines + BUFFER_MARGIN) {
301				nflushd_lines += nmove - max_bufd_lines;
302				flush_lines(nmove - max_bufd_lines);
303			}
304		}
305		/* grow line's buffer? */
306		if (l->l_line_len + 1 >= l->l_lsize) {
307			int need;
308
309			need = l->l_lsize ? l->l_lsize * 2 : 90;
310			if ((l->l_line = realloc(l->l_line,
311			    (unsigned)need * sizeof(CHAR))) == NULL)
312				err(1, NULL);
313			l->l_lsize = need;
314		}
315		c = &l->l_line[l->l_line_len++];
316		c->c_char = ch;
317		c->c_set = cur_set;
318		c->c_column = cur_col;
319		c->c_width = wcwidth(ch);
320		/*
321		 * If things are put in out of order, they will need sorting
322		 * when it is flushed.
323		 */
324		if (cur_col < l->l_max_col)
325			l->l_needs_sort = 1;
326		else
327			l->l_max_col = cur_col;
328		if (c->c_width > 0)
329			cur_col += c->c_width;
330	}
331	if (ferror(stdin))
332		err(1, NULL);
333	if (max_line == 0)
334		exit(0);	/* no lines, so just exit */
335
336	/* goto the last line that had a character on it */
337	for (; l->l_next; l = l->l_next)
338		this_line++;
339	flush_lines(this_line - nflushd_lines + extra_lines + 1);
340
341	/* make sure we leave things in a sane state */
342	if (last_set != CS_NORMAL)
343		PUTC(SI);
344
345	/* flush out the last few blank lines */
346	nblank_lines = max_line - this_line;
347	if (max_line & 1)
348		nblank_lines++;
349	else if (!nblank_lines)
350		/* missing a \n on the last line? */
351		nblank_lines = 2;
352	flush_blanks();
353	exit(0);
354}
355
356static void
357flush_lines(int nflush)
358{
359	LINE *l;
360
361	while (--nflush >= 0) {
362		l = lines;
363		lines = l->l_next;
364		if (l->l_line) {
365			flush_blanks();
366			flush_line(l);
367		}
368		nblank_lines++;
369		if (l->l_line)
370			(void)free(l->l_line);
371		free_line(l);
372	}
373	if (lines)
374		lines->l_prev = NULL;
375}
376
377/*
378 * Print a number of newline/half newlines.  If fine flag is set, nblank_lines
379 * is the number of half line feeds, otherwise it is the number of whole line
380 * feeds.
381 */
382static void
383flush_blanks(void)
384{
385	int half, i, nb;
386
387	half = 0;
388	nb = nblank_lines;
389	if (nb & 1) {
390		if (fine)
391			half = 1;
392		else
393			nb++;
394	}
395	nb /= 2;
396	for (i = nb; --i >= 0;)
397		PUTC('\n');
398	if (half) {
399		PUTC(ESC);
400		PUTC(FHLF);
401		if (!nb)
402			PUTC('\r');
403	}
404	nblank_lines = 0;
405}
406
407/*
408 * Write a line to stdout taking care of space to tab conversion (-h flag)
409 * and character set shifts.
410 */
411static void
412flush_line(LINE *l)
413{
414	CHAR *c, *endc;
415	int i, j, nchars, last_col, save, this_col, tot;
416
417	last_col = 0;
418	nchars = l->l_line_len;
419
420	if (l->l_needs_sort) {
421		static CHAR *sorted;
422		static int count_size, *count, sorted_size;
423
424		/*
425		 * Do an O(n) sort on l->l_line by column being careful to
426		 * preserve the order of characters in the same column.
427		 */
428		if (l->l_lsize > sorted_size) {
429			sorted_size = l->l_lsize;
430			if ((sorted = realloc(sorted,
431			    (unsigned)sizeof(CHAR) * sorted_size)) == NULL)
432				err(1, NULL);
433		}
434		if (l->l_max_col >= count_size) {
435			count_size = l->l_max_col + 1;
436			if ((count = realloc(count,
437			    (unsigned)sizeof(int) * count_size)) == NULL)
438				err(1, NULL);
439		}
440		memset(count, 0, sizeof(int) * l->l_max_col + 1);
441		for (i = nchars, c = l->l_line; --i >= 0; c++)
442			count[c->c_column]++;
443
444		/*
445		 * calculate running total (shifted down by 1) to use as
446		 * indices into new line.
447		 */
448		for (tot = 0, i = 0; i <= l->l_max_col; i++) {
449			save = count[i];
450			count[i] = tot;
451			tot += save;
452		}
453
454		for (i = nchars, c = l->l_line; --i >= 0; c++)
455			sorted[count[c->c_column]++] = *c;
456		c = sorted;
457	} else
458		c = l->l_line;
459	while (nchars > 0) {
460		this_col = c->c_column;
461		endc = c;
462		do {
463			++endc;
464		} while (--nchars > 0 && this_col == endc->c_column);
465
466		/* if -b only print last character */
467		if (no_backspaces) {
468			c = endc - 1;
469			if (nchars > 0 &&
470			    this_col + c->c_width > endc->c_column)
471				continue;
472		}
473
474		if (this_col > last_col) {
475			int nspace = this_col - last_col;
476
477			if (compress_spaces && nspace > 1) {
478				while (1) {
479					int tab_col, tab_size;
480
481					tab_col = (last_col + 8) & ~7;
482					if (tab_col > this_col)
483						break;
484					tab_size = tab_col - last_col;
485					if (tab_size == 1)
486						PUTC(' ');
487					else
488						PUTC('\t');
489					nspace -= tab_size;
490					last_col = tab_col;
491				}
492			}
493			while (--nspace >= 0)
494				PUTC(' ');
495			last_col = this_col;
496		}
497
498		for (;;) {
499			if (c->c_set != last_set) {
500				switch (c->c_set) {
501				case CS_NORMAL:
502					PUTC(SI);
503					break;
504				case CS_ALTERNATE:
505					PUTC(SO);
506				}
507				last_set = c->c_set;
508			}
509			PUTC(c->c_char);
510			if ((c + 1) < endc)
511				for (j = 0; j < c->c_width; j++)
512					PUTC('\b');
513			if (++c >= endc)
514				break;
515		}
516		last_col += (c - 1)->c_width;
517	}
518}
519
520#define	NALLOC 64
521
522static LINE *line_freelist;
523
524static LINE *
525alloc_line(void)
526{
527	LINE *l;
528	int i;
529
530	if (!line_freelist) {
531		if ((l = realloc(NULL, sizeof(LINE) * NALLOC)) == NULL)
532			err(1, NULL);
533		line_freelist = l;
534		for (i = 1; i < NALLOC; i++, l++)
535			l->l_next = l + 1;
536		l->l_next = NULL;
537	}
538	l = line_freelist;
539	line_freelist = l->l_next;
540
541	memset(l, 0, sizeof(LINE));
542	return (l);
543}
544
545static void
546free_line(LINE *l)
547{
548
549	l->l_next = line_freelist;
550	line_freelist = l;
551}
552
553static void
554usage(void)
555{
556
557	(void)fprintf(stderr, "usage: col [-bfhpx] [-l nline]\n");
558	exit(1);
559}
560
561static void
562dowarn(int line)
563{
564
565	warnx("warning: can't back up %s",
566		line < 0 ? "past first line" : "-- line already flushed");
567}
568