cut.c revision 131196 
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 131196 2004-06-27 15:27:15Z 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	 * Numbers and number ranges may be overlapping, repeated, and in
165	 * any order. We handle "-3-5" although there's no real reason too.
166	 */
167	for (; (p = strsep(&list, ", \t")) != NULL;) {
168		setautostart = start = stop = 0;
169		if (*p == '-') {
170			++p;
171			setautostart = 1;
172		}
173		if (isdigit((unsigned char)*p)) {
174			start = stop = strtol(p, &p, 10);
175			if (setautostart && start > autostart)
176				autostart = start;
177		}
178		if (*p == '-') {
179			if (isdigit((unsigned char)p[1]))
180				stop = strtol(p + 1, &p, 10);
181			if (*p == '-') {
182				++p;
183				if (!autostop || autostop > stop)
184					autostop = stop;
185			}
186		}
187		if (*p)
188			errx(1, "[-cf] list: illegal list value");
189		if (!stop || !start)
190			errx(1, "[-cf] list: values may not include zero");
191		if (maxval < stop) {
192			maxval = stop;
193			needpos(maxval + 1);
194		}
195		for (pos = positions + start; start++ <= stop; *pos++ = 1);
196	}
197
198	/* overlapping ranges */
199	if (autostop && maxval > autostop) {
200		maxval = autostop;
201		needpos(maxval + 1);
202	}
203
204	/* set autostart */
205	if (autostart)
206		memset(positions + 1, '1', autostart);
207}
208
209void
210needpos(size_t n)
211{
212	static size_t npos;
213	size_t oldnpos;
214
215	/* Grow the positions array to at least the specified size. */
216	if (n > npos) {
217		oldnpos = npos;
218		if (npos == 0)
219			npos = n;
220		while (n > npos)
221			npos *= 2;
222		if ((positions = realloc(positions, npos)) == NULL)
223			err(1, "realloc");
224		memset((char *)positions + oldnpos, 0, npos - oldnpos);
225	}
226}
227
228int
229b_cut(FILE *fp, const char *fname)
230{
231	int ch, col;
232	char *pos;
233
234	ch = 0;
235	for (;;) {
236		pos = positions + 1;
237		for (col = maxval; col; --col) {
238			if ((ch = getc(fp)) == EOF)
239				return (0);
240			if (ch == '\n')
241				break;
242			if (*pos++)
243				(void)putchar(ch);
244		}
245		if (ch != '\n') {
246			if (autostop)
247				while ((ch = getc(fp)) != EOF && ch != '\n')
248					(void)putchar(ch);
249			else
250				while ((ch = getc(fp)) != EOF && ch != '\n');
251		}
252		(void)putchar('\n');
253	}
254	return (0);
255}
256
257/*
258 * Cut based on byte positions, taking care not to split multibyte characters.
259 * Although this function also handles the case where -n is not specified,
260 * b_cut() ought to be much faster.
261 */
262int
263b_n_cut(FILE *fp, const char *fname)
264{
265	size_t col, i, lbuflen;
266	char *lbuf;
267	int canwrite, clen, warned;
268	mbstate_t mbs;
269
270	memset(&mbs, 0, sizeof(mbs));
271	warned = 0;
272	while ((lbuf = fgetln(fp, &lbuflen)) != NULL) {
273		for (col = 0; lbuflen > 0; col += clen) {
274			if ((clen = mbrlen(lbuf, lbuflen, &mbs)) < 0) {
275				if (!warned) {
276					warn("%s", fname);
277					warned = 1;
278				}
279				memset(&mbs, 0, sizeof(mbs));
280				clen = 1;
281			}
282			if (clen == 0 || *lbuf == '\n')
283				break;
284			if (col < maxval && !positions[1 + col]) {
285				/*
286				 * Print the character if (1) after an initial
287				 * segment of un-selected bytes, the rest of
288				 * it is selected, and (2) the last byte is
289				 * selected.
290				 */
291				i = col;
292				while (i < col + clen && i < maxval &&
293				    !positions[1 + i])
294					i++;
295				canwrite = i < col + clen;
296				for (; i < col + clen && i < maxval; i++)
297					canwrite &= positions[1 + i];
298				if (canwrite)
299					fwrite(lbuf, 1, clen, stdout);
300			} else {
301				/*
302				 * Print the character if all of it has
303				 * been selected.
304				 */
305				canwrite = 1;
306				for (i = col; i < col + clen; i++)
307					if ((i >= maxval && !autostop) ||
308					    (i < maxval && !positions[1 + i])) {
309						canwrite = 0;
310						break;
311					}
312				if (canwrite)
313					fwrite(lbuf, 1, clen, stdout);
314			}
315			lbuf += clen;
316			lbuflen -= clen;
317		}
318		if (lbuflen > 0)
319			putchar('\n');
320	}
321	return (warned);
322}
323
324int
325c_cut(FILE *fp, const char *fname)
326{
327	wint_t ch;
328	int col;
329	char *pos;
330
331	ch = 0;
332	for (;;) {
333		pos = positions + 1;
334		for (col = maxval; col; --col) {
335			if ((ch = getwc(fp)) == WEOF)
336				goto out;
337			if (ch == '\n')
338				break;
339			if (*pos++)
340				(void)putwchar(ch);
341		}
342		if (ch != '\n') {
343			if (autostop)
344				while ((ch = getwc(fp)) != WEOF && ch != '\n')
345					(void)putwchar(ch);
346			else
347				while ((ch = getwc(fp)) != WEOF && ch != '\n');
348		}
349		(void)putwchar('\n');
350	}
351out:
352	if (ferror(fp)) {
353		warn("%s", fname);
354		return (1);
355	}
356	return (0);
357}
358
359int
360f_cut(FILE *fp, const char *fname __unused)
361{
362	int ch, field, isdelim;
363	char *pos, *p, sep;
364	int output;
365	char *lbuf, *mlbuf;
366	size_t lbuflen;
367
368	mlbuf = NULL;
369	for (sep = dchar; (lbuf = fgetln(fp, &lbuflen)) != NULL;) {
370		/* Assert EOL has a newline. */
371		if (*(lbuf + lbuflen - 1) != '\n') {
372			/* Can't have > 1 line with no trailing newline. */
373			mlbuf = malloc(lbuflen + 1);
374			if (mlbuf == NULL)
375				err(1, "malloc");
376			memcpy(mlbuf, lbuf, lbuflen);
377			*(mlbuf + lbuflen) = '\n';
378			lbuf = mlbuf;
379		}
380		output = 0;
381		for (isdelim = 0, p = lbuf;; ++p) {
382			ch = *p;
383			/* this should work if newline is delimiter */
384			if (ch == sep)
385				isdelim = 1;
386			if (ch == '\n') {
387				if (!isdelim && !sflag)
388					(void)fwrite(lbuf, lbuflen, 1, stdout);
389				break;
390			}
391		}
392		if (!isdelim)
393			continue;
394
395		pos = positions + 1;
396		for (field = maxval, p = lbuf; field; --field, ++pos) {
397			if (*pos) {
398				if (output++)
399					(void)putchar(sep);
400				while ((ch = *p++) != '\n' && ch != sep)
401					(void)putchar(ch);
402			} else {
403				while ((ch = *p++) != '\n' && ch != sep)
404					continue;
405			}
406			if (ch == '\n')
407				break;
408		}
409		if (ch != '\n') {
410			if (autostop) {
411				if (output)
412					(void)putchar(sep);
413				for (; (ch = *p) != '\n'; ++p)
414					(void)putchar(ch);
415			} else
416				for (; (ch = *p) != '\n'; ++p);
417		}
418		(void)putchar('\n');
419	}
420	if (mlbuf != NULL)
421		free(mlbuf);
422	return (0);
423}
424
425static void
426usage(void)
427{
428	(void)fprintf(stderr, "%s\n%s\n%s\n",
429		"usage: cut -b list [-n] [file ...]",
430		"       cut -c list [file ...]",
431		"       cut -f list [-s] [-d delim] [file ...]");
432	exit(1);
433}
434