cut.c revision 131201 
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 131201 2004-06-27 16:58:56Z tjr $");
45
46#include <ctype.h>
47#include <err.h>
48#include <errno.h>
49#include <limits.h>
50#include <locale.h>
51#include <stdio.h>
52#include <stdlib.h>
53#include <string.h>
54#include <unistd.h>
55#include <wchar.h>
56
57int	bflag;
58int	cflag;
59wchar_t	dchar;
60char	dcharmb[MB_LEN_MAX + 1];
61int	dflag;
62int	fflag;
63int	nflag;
64int	sflag;
65
66int	b_cut(FILE *, const char *);
67int	b_n_cut(FILE *, const char *);
68int	c_cut(FILE *, const char *);
69int	f_cut(FILE *, const char *);
70void	get_list(char *);
71void	needpos(size_t);
72static 	void usage(void);
73
74int
75main(int argc, char *argv[])
76{
77	FILE *fp;
78	int (*fcn)(FILE *, const char *);
79	int ch, rval;
80	size_t n;
81
82	setlocale(LC_ALL, "");
83
84	fcn = NULL;
85	dchar = '\t';			/* default delimiter is \t */
86	strcpy(dcharmb, "\t");
87
88	while ((ch = getopt(argc, argv, "b:c:d:f:sn")) != -1)
89		switch(ch) {
90		case 'b':
91			get_list(optarg);
92			bflag = 1;
93			break;
94		case 'c':
95			get_list(optarg);
96			cflag = 1;
97			break;
98		case 'd':
99			n = mbrtowc(&dchar, optarg, MB_LEN_MAX, NULL);
100			if (dchar == '\0' || n != strlen(optarg))
101				errx(1, "bad delimiter");
102			strcpy(dcharmb, optarg);
103			dflag = 1;
104			break;
105		case 'f':
106			get_list(optarg);
107			fflag = 1;
108			break;
109		case 's':
110			sflag = 1;
111			break;
112		case 'n':
113			nflag = 1;
114			break;
115		case '?':
116		default:
117			usage();
118		}
119	argc -= optind;
120	argv += optind;
121
122	if (fflag) {
123		if (bflag || cflag || nflag)
124			usage();
125	} else if (!(bflag || cflag) || dflag || sflag)
126		usage();
127	else if (!bflag && nflag)
128		usage();
129
130	if (fflag)
131		fcn = f_cut;
132	else if (cflag)
133		fcn = MB_CUR_MAX > 1 ? c_cut : b_cut;
134	else if (bflag)
135		fcn = nflag && MB_CUR_MAX > 1 ? b_n_cut : b_cut;
136
137	rval = 0;
138	if (*argv)
139		for (; *argv; ++argv) {
140			if (strcmp(*argv, "-") == 0)
141				rval |= fcn(stdin, "stdin");
142			else {
143				if (!(fp = fopen(*argv, "r"))) {
144					warn("%s", *argv);
145					rval = 1;
146					continue;
147				}
148				fcn(fp, *argv);
149				(void)fclose(fp);
150			}
151		}
152	else
153		rval = fcn(stdin, "stdin");
154	exit(rval);
155}
156
157size_t autostart, autostop, maxval;
158
159char *positions;
160
161void
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 too.
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, "[-cf] list: illegal list value");
196		if (!stop || !start)
197			errx(1, "[-cf] 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
216void
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
235int
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 */
269int
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
331int
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
366int
367f_cut(FILE *fp, const char *fname)
368{
369	wchar_t ch;
370	int field, i, isdelim;
371	char *pos, *p;
372	wchar_t sep;
373	int output;
374	char *lbuf, *mlbuf;
375	size_t clen, lbuflen;
376
377	mlbuf = NULL;
378	for (sep = dchar; (lbuf = fgetln(fp, &lbuflen)) != NULL;) {
379		/* Assert EOL has a newline. */
380		if (*(lbuf + lbuflen - 1) != '\n') {
381			/* Can't have > 1 line with no trailing newline. */
382			mlbuf = malloc(lbuflen + 1);
383			if (mlbuf == NULL)
384				err(1, "malloc");
385			memcpy(mlbuf, lbuf, lbuflen);
386			*(mlbuf + lbuflen) = '\n';
387			lbuf = mlbuf;
388		}
389		output = 0;
390		for (isdelim = 0, p = lbuf;; p += clen) {
391			clen = mbrtowc(&ch, p, lbuf + lbuflen - p, NULL);
392			if (clen == (size_t)-1 || clen == (size_t)-2) {
393				warnc(EILSEQ, "%s", fname);
394				free(mlbuf);
395				return (1);
396			}
397			if (clen == 0)
398				clen = 1;
399			/* this should work if newline is delimiter */
400			if (ch == sep)
401				isdelim = 1;
402			if (ch == '\n') {
403				if (!isdelim && !sflag)
404					(void)fwrite(lbuf, lbuflen, 1, stdout);
405				break;
406			}
407		}
408		if (!isdelim)
409			continue;
410
411		pos = positions + 1;
412		for (field = maxval, p = lbuf; field; --field, ++pos) {
413			if (*pos && output++)
414				for (i = 0; dcharmb[i] != '\0'; i++)
415					putchar(dcharmb[i]);
416			for (;;) {
417				clen = mbrtowc(&ch, p, lbuf + lbuflen - p,
418				    NULL);
419				if (clen == (size_t)-1 || clen == (size_t)-2) {
420					warnc(EILSEQ, "%s", fname);
421					free(mlbuf);
422					return (1);
423				}
424				if (clen == 0)
425					clen = 1;
426				p += clen;
427				if (ch == '\n' || ch == sep)
428					break;
429				if (*pos)
430					for (i = 0; i < (int)clen; i++)
431						putchar(p[i - clen]);
432			}
433			if (ch == '\n')
434				break;
435		}
436		if (ch != '\n') {
437			if (autostop) {
438				if (output)
439					for (i = 0; dcharmb[i] != '\0'; i++)
440						putchar(dcharmb[i]);
441				for (; (ch = *p) != '\n'; ++p)
442					(void)putchar(ch);
443			} else
444				for (; (ch = *p) != '\n'; ++p);
445		}
446		(void)putchar('\n');
447	}
448	free(mlbuf);
449	return (0);
450}
451
452static void
453usage(void)
454{
455	(void)fprintf(stderr, "%s\n%s\n%s\n",
456		"usage: cut -b list [-n] [file ...]",
457		"       cut -c list [file ...]",
458		"       cut -f list [-s] [-d delim] [file ...]");
459	exit(1);
460}
461