1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/code/cut.c__netbsd.2014-02-03	Tue May 12 06:46:59 2015 +0200
     1.3 @@ -0,0 +1,306 @@
     1.4 +/*	$NetBSD: cut.c,v 1.29 2014/02/03 20:22:19 wiz Exp $	*/
     1.5 +
     1.6 +/*
     1.7 + * Copyright (c) 1989, 1993
     1.8 + *	The Regents of the University of California.  All rights reserved.
     1.9 + *
    1.10 + * This code is derived from software contributed to Berkeley by
    1.11 + * Adam S. Moskowitz of Menlo Consulting and Marciano Pitargue.
    1.12 + *
    1.13 + * Redistribution and use in source and binary forms, with or without
    1.14 + * modification, are permitted provided that the following conditions
    1.15 + * are met:
    1.16 + * 1. Redistributions of source code must retain the above copyright
    1.17 + *    notice, this list of conditions and the following disclaimer.
    1.18 + * 2. Redistributions in binary form must reproduce the above copyright
    1.19 + *    notice, this list of conditions and the following disclaimer in the
    1.20 + *    documentation and/or other materials provided with the distribution.
    1.21 + * 3. Neither the name of the University nor the names of its contributors
    1.22 + *    may be used to endorse or promote products derived from this software
    1.23 + *    without specific prior written permission.
    1.24 + *
    1.25 + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
    1.26 + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
    1.27 + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
    1.28 + * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
    1.29 + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
    1.30 + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
    1.31 + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
    1.32 + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
    1.33 + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
    1.34 + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
    1.35 + * SUCH DAMAGE.
    1.36 + */
    1.37 +
    1.38 +#include <sys/cdefs.h>
    1.39 +#ifndef lint
    1.40 +__COPYRIGHT("@(#) Copyright (c) 1989, 1993\
    1.41 + The Regents of the University of California.  All rights reserved.");
    1.42 +#endif /* not lint */
    1.43 +
    1.44 +#ifndef lint
    1.45 +#if 0
    1.46 +static char sccsid[] = "@(#)cut.c	8.3 (Berkeley) 5/4/95";
    1.47 +#endif
    1.48 +__RCSID("$NetBSD: cut.c,v 1.29 2014/02/03 20:22:19 wiz Exp $");
    1.49 +#endif /* not lint */
    1.50 +
    1.51 +#include <ctype.h>
    1.52 +#include <err.h>
    1.53 +#include <errno.h>
    1.54 +#include <limits.h>
    1.55 +#include <locale.h>
    1.56 +#include <stdio.h>
    1.57 +#include <stdlib.h>
    1.58 +#include <string.h>
    1.59 +#include <unistd.h>
    1.60 +#include <util.h>
    1.61 +#include <wchar.h>
    1.62 +#include <sys/param.h>
    1.63 +
    1.64 +static int bflag;
    1.65 +static int	cflag;
    1.66 +static char	dchar;
    1.67 +static int	dflag;
    1.68 +static int	fflag;
    1.69 +static int	sflag;
    1.70 +
    1.71 +static void	b_cut(FILE *, const char *);
    1.72 +static void	c_cut(FILE *, const char *);
    1.73 +static void	f_cut(FILE *, const char *);
    1.74 +static void	get_list(char *);
    1.75 +static void	usage(void) __dead;
    1.76 +
    1.77 +int
    1.78 +main(int argc, char *argv[])
    1.79 +{
    1.80 +	FILE *fp;
    1.81 +	void (*fcn)(FILE *, const char *);
    1.82 +	int ch, rval;
    1.83 +
    1.84 +	fcn = NULL;
    1.85 +	(void)setlocale(LC_ALL, "");
    1.86 +
    1.87 +	dchar = '\t';			/* default delimiter is \t */
    1.88 +
    1.89 +	/* Since we don't support multi-byte characters, the -c and -b
    1.90 +	   options are equivalent, and the -n option is meaningless. */
    1.91 +	while ((ch = getopt(argc, argv, "b:c:d:f:sn")) != -1)
    1.92 +		switch(ch) {
    1.93 +		case 'b':
    1.94 +			fcn = b_cut;
    1.95 +			get_list(optarg);
    1.96 +			bflag = 1;
    1.97 +			break;
    1.98 +		case 'c':
    1.99 +			fcn = c_cut;
   1.100 +			get_list(optarg);
   1.101 +			cflag = 1;
   1.102 +			break;
   1.103 +		case 'd':
   1.104 +			dchar = *optarg;
   1.105 +			dflag = 1;
   1.106 +			break;
   1.107 +		case 'f':
   1.108 +			get_list(optarg);
   1.109 +			fcn = f_cut;
   1.110 +			fflag = 1;
   1.111 +			break;
   1.112 +		case 's':
   1.113 +			sflag = 1;
   1.114 +			break;
   1.115 +		case 'n':
   1.116 +			break;
   1.117 +		case '?':
   1.118 +		default:
   1.119 +			usage();
   1.120 +		}
   1.121 +	argc -= optind;
   1.122 +	argv += optind;
   1.123 +
   1.124 +	if (fflag) {
   1.125 +		if (cflag || bflag)
   1.126 +			usage();
   1.127 +	} else if ((!cflag && !bflag) || dflag || sflag)
   1.128 +		usage();
   1.129 +	else if (bflag && cflag)
   1.130 +		usage();
   1.131 +
   1.132 +	rval = 0;
   1.133 +	if (*argv)
   1.134 +		for (; *argv; ++argv) {
   1.135 +			if (strcmp(*argv, "-") == 0)
   1.136 +				fcn(stdin, "stdin");
   1.137 +			else {
   1.138 +				if ((fp = fopen(*argv, "r"))) {
   1.139 +					fcn(fp, *argv);
   1.140 +					(void)fclose(fp);
   1.141 +				} else {
   1.142 +					rval = 1;
   1.143 +					warn("%s", *argv);
   1.144 +				}
   1.145 +			}
   1.146 +		}
   1.147 +	else
   1.148 +		fcn(stdin, "stdin");
   1.149 +	return(rval);
   1.150 +}
   1.151 +
   1.152 +static size_t autostart, autostop, maxval;
   1.153 +
   1.154 +static char *positions = NULL;
   1.155 +static size_t numpositions = 0;
   1.156 +#define ALLOC_CHUNK	_POSIX2_LINE_MAX	/* malloc granularity */
   1.157 +
   1.158 +static void
   1.159 +get_list(char *list)
   1.160 +{
   1.161 +	size_t setautostart, start, stop;
   1.162 +	char *pos;
   1.163 +	char *p;
   1.164 +
   1.165 +	if (positions == NULL) {
   1.166 +		numpositions = ALLOC_CHUNK;
   1.167 +		positions = ecalloc(numpositions, sizeof(*positions));
   1.168 +	}
   1.169 +
   1.170 +	/*
   1.171 +	 * set a byte in the positions array to indicate if a field or
   1.172 +	 * column is to be selected; use +1, it's 1-based, not 0-based.
   1.173 +	 * This parser is less restrictive than the Draft 9 POSIX spec.
   1.174 +	 * POSIX doesn't allow lists that aren't in increasing order or
   1.175 +	 * overlapping lists.  We also handle "-3-5" although there's no
   1.176 +	 * real reason to.
   1.177 +	 */
   1.178 +	for (; (p = strtok(list, ", \t")) != NULL; list = NULL) {
   1.179 +		setautostart = start = stop = 0;
   1.180 +		if (*p == '-') {
   1.181 +			++p;
   1.182 +			setautostart = 1;
   1.183 +		}
   1.184 +		if (isdigit((unsigned char)*p)) {
   1.185 +			start = stop = strtol(p, &p, 10);
   1.186 +			if (setautostart && start > autostart)
   1.187 +				autostart = start;
   1.188 +		}
   1.189 +		if (*p == '-') {
   1.190 +			if (isdigit((unsigned char)p[1]))
   1.191 +				stop = strtol(p + 1, &p, 10);
   1.192 +			if (*p == '-') {
   1.193 +				++p;
   1.194 +				if (!autostop || autostop > stop)
   1.195 +					autostop = stop;
   1.196 +			}
   1.197 +		}
   1.198 +		if (*p)
   1.199 +			errx(1, "[-bcf] list: illegal list value");
   1.200 +		if (!stop || !start)
   1.201 +			errx(1, "[-bcf] list: values may not include zero");
   1.202 +		if (stop + 1 > numpositions) {
   1.203 +			size_t newsize;
   1.204 +			newsize = roundup(stop + 1, ALLOC_CHUNK);
   1.205 +			positions = erealloc(positions, newsize);
   1.206 +			(void)memset(positions + numpositions, 0,
   1.207 +			    newsize - numpositions);
   1.208 +			numpositions = newsize;
   1.209 +		}
   1.210 +		if (maxval < stop)
   1.211 +			maxval = stop;
   1.212 +		for (pos = positions + start; start++ <= stop; pos++)
   1.213 +			*pos = 1;
   1.214 +	}
   1.215 +
   1.216 +	/* overlapping ranges */
   1.217 +	if (autostop && maxval > autostop)
   1.218 +		maxval = autostop;
   1.219 +
   1.220 +	/* set autostart */
   1.221 +	if (autostart)
   1.222 +		(void)memset(positions + 1, '1', autostart);
   1.223 +}
   1.224 +
   1.225 +static void
   1.226 +/*ARGSUSED*/
   1.227 +f_cut(FILE *fp, const char *fname __unused)
   1.228 +{
   1.229 +	int ch, field, isdelim;
   1.230 +	char *pos, *p, sep;
   1.231 +	int output;
   1.232 +	size_t len;
   1.233 +	char *lbuf, *tbuf;
   1.234 +
   1.235 +	for (sep = dchar, tbuf = NULL; (lbuf = fgetln(fp, &len)) != NULL;) {
   1.236 +		output = 0;
   1.237 +		if (lbuf[len - 1] != '\n') {
   1.238 +			/* no newline at the end of the last line so add one */
   1.239 +			if ((tbuf = (char *)malloc(len + 1)) == NULL)
   1.240 +				err(1, NULL);
   1.241 +			(void)memcpy(tbuf, lbuf, len);
   1.242 +			tbuf[len++] = '\n';
   1.243 +			lbuf = tbuf;
   1.244 +		}
   1.245 +		for (isdelim = 0, p = lbuf;; ++p) {
   1.246 +			ch = *p;
   1.247 +			/* this should work if newline is delimiter */
   1.248 +			if (ch == sep)
   1.249 +				isdelim = 1;
   1.250 +			if (ch == '\n') {
   1.251 +				if (!isdelim && !sflag)
   1.252 +					(void)fwrite(lbuf, len, 1, stdout);
   1.253 +				break;
   1.254 +			}
   1.255 +		}
   1.256 +		if (!isdelim)
   1.257 +			continue;
   1.258 +
   1.259 +		pos = positions + 1;
   1.260 +		for (field = maxval, p = lbuf; field; --field, ++pos) {
   1.261 +			if (*pos) {
   1.262 +				if (output++)
   1.263 +					(void)putchar(sep);
   1.264 +				while ((ch = *p++) != '\n' && ch != sep)
   1.265 +					(void)putchar(ch);
   1.266 +			} else {
   1.267 +				while ((ch = *p++) != '\n' && ch != sep)
   1.268 +					continue;
   1.269 +			}
   1.270 +			if (ch == '\n')
   1.271 +				break;
   1.272 +		}
   1.273 +		if (ch != '\n') {
   1.274 +			if (autostop) {
   1.275 +				if (output)
   1.276 +					(void)putchar(sep);
   1.277 +				for (; (ch = *p) != '\n'; ++p)
   1.278 +					(void)putchar(ch);
   1.279 +			} else
   1.280 +				for (; (ch = *p) != '\n'; ++p);
   1.281 +		}
   1.282 +		(void)putchar('\n');
   1.283 +		if (tbuf) {
   1.284 +			free(tbuf);
   1.285 +			tbuf = NULL;
   1.286 +		}
   1.287 +	}
   1.288 +	if (tbuf)
   1.289 +		free(tbuf);
   1.290 +}
   1.291 +
   1.292 +static void
   1.293 +usage(void)
   1.294 +{
   1.295 +	(void)fprintf(stderr, "usage:\tcut -b list [-n] [file ...]\n"
   1.296 +	    "\tcut -c list [file ...]\n"
   1.297 +	    "\tcut -f list [-d string] [-s] [file ...]\n");
   1.298 +	exit(1);
   1.299 +}
   1.300 +
   1.301 +/* make b_put(): */
   1.302 +#define CUT_BYTE 1
   1.303 +#include "x_cut.c"
   1.304 +#undef CUT_BYTE
   1.305 +
   1.306 +/* make c_put(): */
   1.307 +#define CUT_BYTE 0
   1.308 +#include "x_cut.c"
   1.309 +#undef CUT_BYTE