1/*
2 * Copyright (c) 1989, 1993
3 *	The Regents of the University of California.  All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Adam S. Moskowitz of Menlo Consulting and Marciano Pitargue.
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) 1989, 1993\n\
36	The Regents of the University of California.  All rights reserved.\n";
37static const char sccsid[] = "@(#)cut.c	8.3 (Berkeley) 5/4/95";
38#endif /* not lint */
39#include <sys/cdefs.h>
40__FBSDID("$FreeBSD$");
41
42#include <ctype.h>
43#include <err.h>
44#include <errno.h>
45#include <limits.h>
46#include <locale.h>
47#include <stdio.h>
48#include <stdlib.h>
49#include <string.h>
50#include <unistd.h>
51#include <wchar.h>
52
53static int	bflag;
54static int	cflag;
55static wchar_t	dchar;
56static char	dcharmb[MB_LEN_MAX + 1];
57static int	dflag;
58static int	fflag;
59static int	nflag;
60static int	sflag;
61static int	wflag;
62
63static size_t	autostart, autostop, maxval;
64static char *	positions;
65
66static int	b_cut(FILE *, const char *);
67static int	b_n_cut(FILE *, const char *);
68static int	c_cut(FILE *, const char *);
69static int	f_cut(FILE *, const char *);
70static void	get_list(char *);
71static int	is_delim(wchar_t);
72static void	needpos(size_t);
73static void	usage(void);
74
75int
76main(int argc, char *argv[])
77{
78	FILE *fp;
79	int (*fcn)(FILE *, const char *);
80	int ch, rval;
81	size_t n;
82
83	setlocale(LC_ALL, "");
84
85	fcn = NULL;
86	dchar = '\t';			/* default delimiter is \t */
87	strcpy(dcharmb, "\t");
88
89	while ((ch = getopt(argc, argv, "b:c:d:f:snw")) != -1)
90		switch(ch) {
91		case 'b':
92			get_list(optarg);
93			bflag = 1;
94			break;
95		case 'c':
96			get_list(optarg);
97			cflag = 1;
98			break;
99		case 'd':
100			n = mbrtowc(&dchar, optarg, MB_LEN_MAX, NULL);
101			if (dchar == '\0' || n != strlen(optarg))
102				errx(1, "bad delimiter");
103			strcpy(dcharmb, optarg);
104			dflag = 1;
105			break;
106		case 'f':
107			get_list(optarg);
108			fflag = 1;
109			break;
110		case 's':
111			sflag = 1;
112			break;
113		case 'n':
114			nflag = 1;
115			break;
116		case 'w':
117			wflag = 1;
118			break;
119		case '?':
120		default:
121			usage();
122		}
123	argc -= optind;
124	argv += optind;
125
126	if (fflag) {
127		if (bflag || cflag || nflag || (wflag && dflag))
128			usage();
129	} else if (!(bflag || cflag) || dflag || sflag || wflag)
130		usage();
131	else if (!bflag && nflag)
132		usage();
133
134	if (fflag)
135		fcn = f_cut;
136	else if (cflag)
137		fcn = MB_CUR_MAX > 1 ? c_cut : b_cut;
138	else if (bflag)
139		fcn = nflag && MB_CUR_MAX > 1 ? b_n_cut : b_cut;
140
141	rval = 0;
142	if (*argv)
143		for (; *argv; ++argv) {
144			if (strcmp(*argv, "-") == 0)
145				rval |= fcn(stdin, "stdin");
146			else {
147				if (!(fp = fopen(*argv, "r"))) {
148					warn("%s", *argv);
149					rval = 1;
150					continue;
151				}
152				fcn(fp, *argv);
153				(void)fclose(fp);
154			}
155		}
156	else
157		rval = fcn(stdin, "stdin");
158	exit(rval);
159}
160
161static void
162get_list(char *list)
163{
164	size_t setautostart, start, stop;
165	char *pos;
166	char *p;
167
168	/*
169	 * set a byte in the positions array to indicate if a field or
170	 * column is to be selected; use +1, it's 1-based, not 0-based.
171	 * Numbers and number ranges may be overlapping, repeated, and in
172	 * any order. We handle "-3-5" although there's no real reason to.
173	 */
174	for (; (p = strsep(&list, ", \t")) != NULL;) {
175		setautostart = start = stop = 0;
176		if (*p == '-') {
177			++p;
178			setautostart = 1;
179		}
180		if (isdigit((unsigned char)*p)) {
181			start = stop = strtol(p, &p, 10);
182			if (setautostart && start > autostart)
183				autostart = start;
184		}
185		if (*p == '-') {
186			if (isdigit((unsigned char)p[1]))
187				stop = strtol(p + 1, &p, 10);
188			if (*p == '-') {
189				++p;
190				if (!autostop || autostop > stop)
191					autostop = stop;
192			}
193		}
194		if (*p)
195			errx(1, "[-bcf] list: illegal list value");
196		if (!stop || !start)
197			errx(1, "[-bcf] list: values may not include zero");
198		if (maxval < stop) {
199			maxval = stop;
200			needpos(maxval + 1);
201		}
202		for (pos = positions + start; start++ <= stop; *pos++ = 1);
203	}
204
205	/* overlapping ranges */
206	if (autostop && maxval > autostop) {
207		maxval = autostop;
208		needpos(maxval + 1);
209	}
210
211	/* set autostart */
212	if (autostart)
213		memset(positions + 1, '1', autostart);
214}
215
216static void
217needpos(size_t n)
218{
219	static size_t npos;
220	size_t oldnpos;
221
222	/* Grow the positions array to at least the specified size. */
223	if (n > npos) {
224		oldnpos = npos;
225		if (npos == 0)
226			npos = n;
227		while (n > npos)
228			npos *= 2;
229		if ((positions = realloc(positions, npos)) == NULL)
230			err(1, "realloc");
231		memset((char *)positions + oldnpos, 0, npos - oldnpos);
232	}
233}
234
235static int
236b_cut(FILE *fp, const char *fname __unused)
237{
238	int ch, col;
239	char *pos;
240
241	ch = 0;
242	for (;;) {
243		pos = positions + 1;
244		for (col = maxval; col; --col) {
245			if ((ch = getc(fp)) == EOF)
246				return (0);
247			if (ch == '\n')
248				break;
249			if (*pos++)
250				(void)putchar(ch);
251		}
252		if (ch != '\n') {
253			if (autostop)
254				while ((ch = getc(fp)) != EOF && ch != '\n')
255					(void)putchar(ch);
256			else
257				while ((ch = getc(fp)) != EOF && ch != '\n');
258		}
259		(void)putchar('\n');
260	}
261	return (0);
262}
263
264/*
265 * Cut based on byte positions, taking care not to split multibyte characters.
266 * Although this function also handles the case where -n is not specified,
267 * b_cut() ought to be much faster.
268 */
269static int
270b_n_cut(FILE *fp, const char *fname)
271{
272	size_t col, i, lbuflen;
273	char *lbuf;
274	int canwrite, clen, warned;
275	mbstate_t mbs;
276
277	memset(&mbs, 0, sizeof(mbs));
278	warned = 0;
279	while ((lbuf = fgetln(fp, &lbuflen)) != NULL) {
280		for (col = 0; lbuflen > 0; col += clen) {
281			if ((clen = mbrlen(lbuf, lbuflen, &mbs)) < 0) {
282				if (!warned) {
283					warn("%s", fname);
284					warned = 1;
285				}
286				memset(&mbs, 0, sizeof(mbs));
287				clen = 1;
288			}
289			if (clen == 0 || *lbuf == '\n')
290				break;
291			if (col < maxval && !positions[1 + col]) {
292				/*
293				 * Print the character if (1) after an initial
294				 * segment of un-selected bytes, the rest of
295				 * it is selected, and (2) the last byte is
296				 * selected.
297				 */
298				i = col;
299				while (i < col + clen && i < maxval &&
300				    !positions[1 + i])
301					i++;
302				canwrite = i < col + clen;
303				for (; i < col + clen && i < maxval; i++)
304					canwrite &= positions[1 + i];
305				if (canwrite)
306					fwrite(lbuf, 1, clen, stdout);
307			} else {
308				/*
309				 * Print the character if all of it has
310				 * been selected.
311				 */
312				canwrite = 1;
313				for (i = col; i < col + clen; i++)
314					if ((i >= maxval && !autostop) ||
315					    (i < maxval && !positions[1 + i])) {
316						canwrite = 0;
317						break;
318					}
319				if (canwrite)
320					fwrite(lbuf, 1, clen, stdout);
321			}
322			lbuf += clen;
323			lbuflen -= clen;
324		}
325		if (lbuflen > 0)
326			putchar('\n');
327	}
328	return (warned);
329}
330
331static int
332c_cut(FILE *fp, const char *fname)
333{
334	wint_t ch;
335	int col;
336	char *pos;
337
338	ch = 0;
339	for (;;) {
340		pos = positions + 1;
341		for (col = maxval; col; --col) {
342			if ((ch = getwc(fp)) == WEOF)
343				goto out;
344			if (ch == '\n')
345				break;
346			if (*pos++)
347				(void)putwchar(ch);
348		}
349		if (ch != '\n') {
350			if (autostop)
351				while ((ch = getwc(fp)) != WEOF && ch != '\n')
352					(void)putwchar(ch);
353			else
354				while ((ch = getwc(fp)) != WEOF && ch != '\n');
355		}
356		(void)putwchar('\n');
357	}
358out:
359	if (ferror(fp)) {
360		warn("%s", fname);
361		return (1);
362	}
363	return (0);
364}
365
366static int
367is_delim(wchar_t ch)
368{
369	if (wflag) {
370		if (ch == ' ' || ch == '\t')
371			return 1;
372	} else {
373		if (ch == dchar)
374			return 1;
375	}
376	return 0;
377}
378
379static int
380f_cut(FILE *fp, const char *fname)
381{
382	wchar_t ch;
383	int field, i, isdelim;
384	char *pos, *p;
385	int output;
386	char *lbuf, *mlbuf;
387	size_t clen, lbuflen, reallen;
388
389	mlbuf = NULL;
390	while ((lbuf = fgetln(fp, &lbuflen)) != NULL) {
391		reallen = lbuflen;
392		/* Assert EOL has a newline. */
393		if (*(lbuf + lbuflen - 1) != '\n') {
394			/* Can't have > 1 line with no trailing newline. */
395			mlbuf = malloc(lbuflen + 1);
396			if (mlbuf == NULL)
397				err(1, "malloc");
398			memcpy(mlbuf, lbuf, lbuflen);
399			*(mlbuf + lbuflen) = '\n';
400			lbuf = mlbuf;
401			reallen++;
402		}
403		output = 0;
404		for (isdelim = 0, p = lbuf;; p += clen) {
405			clen = mbrtowc(&ch, p, lbuf + reallen - p, NULL);
406			if (clen == (size_t)-1 || clen == (size_t)-2) {
407				warnc(EILSEQ, "%s", fname);
408				free(mlbuf);
409				return (1);
410			}
411			if (clen == 0)
412				clen = 1;
413			/* this should work if newline is delimiter */
414			if (is_delim(ch))
415				isdelim = 1;
416			if (ch == '\n') {
417				if (!isdelim && !sflag)
418					(void)fwrite(lbuf, lbuflen, 1, stdout);
419				break;
420			}
421		}
422		if (!isdelim)
423			continue;
424
425		pos = positions + 1;
426		for (field = maxval, p = lbuf; field; --field, ++pos) {
427			if (*pos && output++)
428				for (i = 0; dcharmb[i] != '\0'; i++)
429					putchar(dcharmb[i]);
430			for (;;) {
431				clen = mbrtowc(&ch, p, lbuf + reallen - p,
432				    NULL);
433				if (clen == (size_t)-1 || clen == (size_t)-2) {
434					warnc(EILSEQ, "%s", fname);
435					free(mlbuf);
436					return (1);
437				}
438				if (clen == 0)
439					clen = 1;
440				p += clen;
441				if (ch == '\n' || is_delim(ch)) {
442					/* compress whitespace */
443					if (wflag && ch != '\n')
444						while (is_delim(*p))
445							p++;
446					break;
447				}
448				if (*pos)
449					for (i = 0; i < (int)clen; i++)
450						putchar(p[i - clen]);
451			}
452			if (ch == '\n')
453				break;
454		}
455		if (ch != '\n') {
456			if (autostop) {
457				if (output)
458					for (i = 0; dcharmb[i] != '\0'; i++)
459						putchar(dcharmb[i]);
460				for (; (ch = *p) != '\n'; ++p)
461					(void)putchar(ch);
462			} else
463				for (; (ch = *p) != '\n'; ++p);
464		}
465		(void)putchar('\n');
466	}
467	free(mlbuf);
468	return (0);
469}
470
471static void
472usage(void)
473{
474	(void)fprintf(stderr, "%s\n%s\n%s\n",
475		"usage: cut -b list [-n] [file ...]",
476		"       cut -c list [file ...]",
477		"       cut -f list [-s] [-w | -d delim] [file ...]");
478	exit(1);
479}
480