cut.c revision 131194 
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 * 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 const char copyright[] =
39"@(#) Copyright (c) 1989, 1993\n\
40	The Regents of the University of California.  All rights reserved.\n";
41static const char sccsid[] = "@(#)cut.c	8.3 (Berkeley) 5/4/95";
42#endif /* not lint */
43#include <sys/cdefs.h>
44__FBSDID("$FreeBSD: head/usr.bin/cut/cut.c 131194 2004-06-27 14:55:07Z tjr $");
45
46#include <ctype.h>
47#include <err.h>
48#include <limits.h>
49#include <locale.h>
50#include <stdio.h>
51#include <stdlib.h>
52#include <string.h>
53#include <unistd.h>
54#include <wchar.h>
55
56int	bflag;
57int	cflag;
58char	dchar;
59int	dflag;
60int	fflag;
61int	nflag;
62int	sflag;
63
64int	b_cut(FILE *, const char *);
65int	b_n_cut(FILE *, const char *);
66int	c_cut(FILE *, const char *);
67int	f_cut(FILE *, const char *);
68void	get_list(char *);
69void	needpos(size_t);
70static 	void usage(void);
71
72int
73main(int argc, char *argv[])
74{
75	FILE *fp;
76	int (*fcn)(FILE *, const char *);
77	int ch, rval;
78
79	setlocale(LC_ALL, "");
80
81	fcn = NULL;
82	dchar = '\t';			/* default delimiter is \t */
83
84	while ((ch = getopt(argc, argv, "b:c:d:f:sn")) != -1)
85		switch(ch) {
86		case 'b':
87			get_list(optarg);
88			bflag = 1;
89			break;
90		case 'c':
91			get_list(optarg);
92			cflag = 1;
93			break;
94		case 'd':
95			dchar = *optarg;
96			dflag = 1;
97			break;
98		case 'f':
99			get_list(optarg);
100			fflag = 1;
101			break;
102		case 's':
103			sflag = 1;
104			break;
105		case 'n':
106			nflag = 1;
107			break;
108		case '?':
109		default:
110			usage();
111		}
112	argc -= optind;
113	argv += optind;
114
115	if (fflag) {
116		if (bflag || cflag || nflag)
117			usage();
118	} else if (!(bflag || cflag) || dflag || sflag)
119		usage();
120	else if (!bflag && nflag)
121		usage();
122
123	if (fflag)
124		fcn = f_cut;
125	else if (cflag)
126		fcn = MB_CUR_MAX > 1 ? c_cut : b_cut;
127	else if (bflag)
128		fcn = nflag && MB_CUR_MAX > 1 ? b_n_cut : b_cut;
129
130	rval = 0;
131	if (*argv)
132		for (; *argv; ++argv) {
133			if (strcmp(*argv, "-") == 0)
134				rval |= fcn(stdin, "stdin");
135			else {
136				if (!(fp = fopen(*argv, "r"))) {
137					warn("%s", *argv);
138					rval = 1;
139					continue;
140				}
141				fcn(fp, *argv);
142				(void)fclose(fp);
143			}
144		}
145	else
146		rval = fcn(stdin, "stdin");
147	exit(rval);
148}
149
150size_t autostart, autostop, maxval;
151
152char *positions;
153
154void
155get_list(char *list)
156{
157	size_t setautostart, start, stop;
158	char *pos;
159	char *p;
160
161	/*
162	 * set a byte in the positions array to indicate if a field or
163	 * column is to be selected; use +1, it's 1-based, not 0-based.
164	 * This parser is less restrictive than the Draft 9 POSIX spec.
165	 * POSIX doesn't allow lists that aren't in increasing order or
166	 * overlapping lists.  We also handle "-3-5" although there's no
167	 * real reason too.
168	 */
169	for (; (p = strsep(&list, ", \t")) != NULL;) {
170		setautostart = start = stop = 0;
171		if (*p == '-') {
172			++p;
173			setautostart = 1;
174		}
175		if (isdigit((unsigned char)*p)) {
176			start = stop = strtol(p, &p, 10);
177			if (setautostart && start > autostart)
178				autostart = start;
179		}
180		if (*p == '-') {
181			if (isdigit((unsigned char)p[1]))
182				stop = strtol(p + 1, &p, 10);
183			if (*p == '-') {
184				++p;
185				if (!autostop || autostop > stop)
186					autostop = stop;
187			}
188		}
189		if (*p)
190			errx(1, "[-cf] list: illegal list value");
191		if (!stop || !start)
192			errx(1, "[-cf] list: values may not include zero");
193		if (maxval < stop) {
194			maxval = stop;
195			needpos(maxval + 1);
196		}
197		for (pos = positions + start; start++ <= stop; *pos++ = 1);
198	}
199
200	/* overlapping ranges */
201	if (autostop && maxval > autostop) {
202		maxval = autostop;
203		needpos(maxval + 1);
204	}
205
206	/* set autostart */
207	if (autostart)
208		memset(positions + 1, '1', autostart);
209}
210
211void
212needpos(size_t n)
213{
214	static size_t npos;
215	size_t oldnpos;
216
217	/* Grow the positions array to at least the specified size. */
218	if (n > npos) {
219		oldnpos = npos;
220		if (npos == 0)
221			npos = n;
222		while (n > npos)
223			npos *= 2;
224		if ((positions = realloc(positions, npos)) == NULL)
225			err(1, "realloc");
226		memset((char *)positions + oldnpos, 0, npos - oldnpos);
227	}
228}
229
230int
231b_cut(FILE *fp, const char *fname)
232{
233	int ch, col;
234	char *pos;
235
236	ch = 0;
237	for (;;) {
238		pos = positions + 1;
239		for (col = maxval; col; --col) {
240			if ((ch = getc(fp)) == EOF)
241				return (0);
242			if (ch == '\n')
243				break;
244			if (*pos++)
245				(void)putchar(ch);
246		}
247		if (ch != '\n') {
248			if (autostop)
249				while ((ch = getc(fp)) != EOF && ch != '\n')
250					(void)putchar(ch);
251			else
252				while ((ch = getc(fp)) != EOF && ch != '\n');
253		}
254		(void)putchar('\n');
255	}
256	return (0);
257}
258
259/*
260 * Cut based on byte positions, taking care not to split multibyte characters.
261 * Although this function also handles the case where -n is not specified,
262 * b_cut() ought to be much faster.
263 */
264int
265b_n_cut(FILE *fp, const char *fname)
266{
267	size_t col, i, lbuflen;
268	char *lbuf;
269	int canwrite, clen, warned;
270	mbstate_t mbs;
271
272	memset(&mbs, 0, sizeof(mbs));
273	warned = 0;
274	while ((lbuf = fgetln(fp, &lbuflen)) != NULL) {
275		for (col = 0; lbuflen > 0; col += clen) {
276			if ((clen = mbrlen(lbuf, lbuflen, &mbs)) < 0) {
277				if (!warned) {
278					warn("%s", fname);
279					warned = 1;
280				}
281				memset(&mbs, 0, sizeof(mbs));
282				clen = 1;
283			}
284			if (clen == 0 || *lbuf == '\n')
285				break;
286			if (col < maxval && !positions[1 + col]) {
287				/*
288				 * Print the character if (1) after an initial
289				 * segment of un-selected bytes, the rest of
290				 * it is selected, and (2) the last byte is
291				 * selected.
292				 */
293				i = col;
294				while (i < col + clen && i < maxval &&
295				    !positions[1 + i])
296					i++;
297				canwrite = i < col + clen;
298				for (; i < col + clen && i < maxval; i++)
299					canwrite &= positions[1 + i];
300				if (canwrite)
301					fwrite(lbuf, 1, clen, stdout);
302			} else {
303				/*
304				 * Print the character if all of it has
305				 * been selected.
306				 */
307				canwrite = 1;
308				for (i = col; i < col + clen; i++)
309					if ((i >= maxval && !autostop) ||
310					    (i < maxval && !positions[1 + i])) {
311						canwrite = 0;
312						break;
313					}
314				if (canwrite)
315					fwrite(lbuf, 1, clen, stdout);
316			}
317			lbuf += clen;
318			lbuflen -= clen;
319		}
320		if (lbuflen > 0)
321			putchar('\n');
322	}
323	return (warned);
324}
325
326int
327c_cut(FILE *fp, const char *fname)
328{
329	wint_t ch;
330	int col;
331	char *pos;
332
333	ch = 0;
334	for (;;) {
335		pos = positions + 1;
336		for (col = maxval; col; --col) {
337			if ((ch = getwc(fp)) == WEOF)
338				goto out;
339			if (ch == '\n')
340				break;
341			if (*pos++)
342				(void)putwchar(ch);
343		}
344		if (ch != '\n') {
345			if (autostop)
346				while ((ch = getwc(fp)) != WEOF && ch != '\n')
347					(void)putwchar(ch);
348			else
349				while ((ch = getwc(fp)) != WEOF && ch != '\n');
350		}
351		(void)putwchar('\n');
352	}
353out:
354	if (ferror(fp)) {
355		warn("%s", fname);
356		return (1);
357	}
358	return (0);
359}
360
361int
362f_cut(FILE *fp, const char *fname __unused)
363{
364	int ch, field, isdelim;
365	char *pos, *p, sep;
366	int output;
367	char *lbuf, *mlbuf;
368	size_t lbuflen;
369
370	mlbuf = NULL;
371	for (sep = dchar; (lbuf = fgetln(fp, &lbuflen)) != NULL;) {
372		/* Assert EOL has a newline. */
373		if (*(lbuf + lbuflen - 1) != '\n') {
374			/* Can't have > 1 line with no trailing newline. */
375			mlbuf = malloc(lbuflen + 1);
376			if (mlbuf == NULL)
377				err(1, "malloc");
378			memcpy(mlbuf, lbuf, lbuflen);
379			*(mlbuf + lbuflen) = '\n';
380			lbuf = mlbuf;
381		}
382		output = 0;
383		for (isdelim = 0, p = lbuf;; ++p) {
384			ch = *p;
385			/* this should work if newline is delimiter */
386			if (ch == sep)
387				isdelim = 1;
388			if (ch == '\n') {
389				if (!isdelim && !sflag)
390					(void)fwrite(lbuf, lbuflen, 1, stdout);
391				break;
392			}
393		}
394		if (!isdelim)
395			continue;
396
397		pos = positions + 1;
398		for (field = maxval, p = lbuf; field; --field, ++pos) {
399			if (*pos) {
400				if (output++)
401					(void)putchar(sep);
402				while ((ch = *p++) != '\n' && ch != sep)
403					(void)putchar(ch);
404			} else {
405				while ((ch = *p++) != '\n' && ch != sep)
406					continue;
407			}
408			if (ch == '\n')
409				break;
410		}
411		if (ch != '\n') {
412			if (autostop) {
413				if (output)
414					(void)putchar(sep);
415				for (; (ch = *p) != '\n'; ++p)
416					(void)putchar(ch);
417			} else
418				for (; (ch = *p) != '\n'; ++p);
419		}
420		(void)putchar('\n');
421	}
422	if (mlbuf != NULL)
423		free(mlbuf);
424	return (0);
425}
426
427static void
428usage(void)
429{
430	(void)fprintf(stderr, "%s\n%s\n%s\n",
431		"usage: cut -b list [-n] [file ...]",
432		"       cut -c list [file ...]",
433		"       cut -f list [-s] [-d delim] [file ...]");
434	exit(1);
435}
436