1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/code/cut.c__openbsd.2008-06-27	Tue May 12 06:46:59 2015 +0200
     1.3 @@ -0,0 +1,290 @@
     1.4 +/*	$OpenBSD: cut.c,v 1.13 2008/06/27 08:02:13 sobrado Exp $	*/
     1.5 +/*	$NetBSD: cut.c,v 1.9 1995/09/02 05:59:23 jtc Exp $	*/
     1.6 +
     1.7 +/*
     1.8 + * Copyright (c) 1989, 1993
     1.9 + *	The Regents of the University of California.  All rights reserved.
    1.10 + *
    1.11 + * This code is derived from software contributed to Berkeley by
    1.12 + * Adam S. Moskowitz of Menlo Consulting and Marciano Pitargue.
    1.13 + *
    1.14 + * Redistribution and use in source and binary forms, with or without
    1.15 + * modification, are permitted provided that the following conditions
    1.16 + * are met:
    1.17 + * 1. Redistributions of source code must retain the above copyright
    1.18 + *    notice, this list of conditions and the following disclaimer.
    1.19 + * 2. Redistributions in binary form must reproduce the above copyright
    1.20 + *    notice, this list of conditions and the following disclaimer in the
    1.21 + *    documentation and/or other materials provided with the distribution.
    1.22 + * 3. Neither the name of the University nor the names of its contributors
    1.23 + *    may be used to endorse or promote products derived from this software
    1.24 + *    without specific prior written permission.
    1.25 + *
    1.26 + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
    1.27 + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
    1.28 + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
    1.29 + * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
    1.30 + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
    1.31 + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
    1.32 + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
    1.33 + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
    1.34 + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
    1.35 + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
    1.36 + * SUCH DAMAGE.
    1.37 + */
    1.38 +
    1.39 +#ifndef lint
    1.40 +static char copyright[] =
    1.41 +"@(#) Copyright (c) 1989, 1993\n\
    1.42 +	The Regents of the University of California.  All rights reserved.\n";
    1.43 +#endif /* not lint */
    1.44 +
    1.45 +#ifndef lint
    1.46 +#if 0
    1.47 +static char sccsid[] = "@(#)cut.c	8.3 (Berkeley) 5/4/95";
    1.48 +#endif
    1.49 +static char rcsid[] = "$OpenBSD: cut.c,v 1.13 2008/06/27 08:02:13 sobrado Exp $";
    1.50 +#endif /* not lint */
    1.51 +
    1.52 +#include <ctype.h>
    1.53 +#include <err.h>
    1.54 +#include <errno.h>
    1.55 +#include <limits.h>
    1.56 +#include <locale.h>
    1.57 +#include <stdio.h>
    1.58 +#include <stdlib.h>
    1.59 +#include <string.h>
    1.60 +#include <unistd.h>
    1.61 +
    1.62 +int	cflag;
    1.63 +char	dchar;
    1.64 +int	dflag;
    1.65 +int	fflag;
    1.66 +int	sflag;
    1.67 +
    1.68 +void	c_cut(FILE *, char *);
    1.69 +void	f_cut(FILE *, char *);
    1.70 +void	get_list(char *);
    1.71 +void	usage(void);
    1.72 +
    1.73 +int
    1.74 +main(int argc, char *argv[])
    1.75 +{
    1.76 +	FILE *fp;
    1.77 +	void (*fcn)(FILE *, char *);
    1.78 +	int ch;
    1.79 +
    1.80 +	setlocale (LC_ALL, "");
    1.81 +
    1.82 +	dchar = '\t';			/* default delimiter is \t */
    1.83 +
    1.84 +	/* Since we don't support multi-byte characters, the -c and -b 
    1.85 +	   options are equivalent, and the -n option is meaningless. */
    1.86 +	while ((ch = getopt(argc, argv, "b:c:d:f:sn")) != -1)
    1.87 +		switch(ch) {
    1.88 +		case 'b':
    1.89 +		case 'c':
    1.90 +			fcn = c_cut;
    1.91 +			get_list(optarg);
    1.92 +			cflag = 1;
    1.93 +			break;
    1.94 +		case 'd':
    1.95 +			dchar = *optarg;
    1.96 +			dflag = 1;
    1.97 +			break;
    1.98 +		case 'f':
    1.99 +			get_list(optarg);
   1.100 +			fcn = f_cut;
   1.101 +			fflag = 1;
   1.102 +			break;
   1.103 +		case 's':
   1.104 +			sflag = 1;
   1.105 +			break;
   1.106 +		case 'n':
   1.107 +			break;
   1.108 +		case '?':
   1.109 +		default:
   1.110 +			usage();
   1.111 +		}
   1.112 +	argc -= optind;
   1.113 +	argv += optind;
   1.114 +
   1.115 +	if (fflag) {
   1.116 +		if (cflag)
   1.117 +			usage();
   1.118 +	} else if (!cflag || dflag || sflag)
   1.119 +		usage();
   1.120 +
   1.121 +	if (*argv)
   1.122 +		for (; *argv; ++argv) {
   1.123 +			if (!(fp = fopen(*argv, "r")))
   1.124 +				err(1, "%s", *argv);
   1.125 +			fcn(fp, *argv);
   1.126 +			(void)fclose(fp);
   1.127 +		}
   1.128 +	else
   1.129 +		fcn(stdin, "stdin");
   1.130 +	exit(0);
   1.131 +}
   1.132 +
   1.133 +int autostart, autostop, maxval;
   1.134 +
   1.135 +char positions[_POSIX2_LINE_MAX + 1];
   1.136 +
   1.137 +void
   1.138 +get_list(char *list)
   1.139 +{
   1.140 +	int setautostart, start, stop;
   1.141 +	char *pos;
   1.142 +	char *p;
   1.143 +
   1.144 +	/*
   1.145 +	 * set a byte in the positions array to indicate if a field or
   1.146 +	 * column is to be selected; use +1, it's 1-based, not 0-based.
   1.147 +	 * This parser is less restrictive than the Draft 9 POSIX spec.
   1.148 +	 * POSIX doesn't allow lists that aren't in increasing order or
   1.149 +	 * overlapping lists.  We also handle "-3-5" although there's no
   1.150 +	 * real reason too.
   1.151 +	 */
   1.152 +	while ((p = strsep(&list, ", \t"))) {
   1.153 +		setautostart = start = stop = 0;
   1.154 +		if (*p == '-') {
   1.155 +			++p;
   1.156 +			setautostart = 1;
   1.157 +		}
   1.158 +		if (isdigit(*p)) {
   1.159 +			start = stop = strtol(p, &p, 10);
   1.160 +			if (setautostart && start > autostart)
   1.161 +				autostart = start;
   1.162 +		}
   1.163 +		if (*p == '-') {
   1.164 +			if (isdigit(p[1]))
   1.165 +				stop = strtol(p + 1, &p, 10);
   1.166 +			if (*p == '-') {
   1.167 +				++p;
   1.168 +				if (!autostop || autostop > stop)
   1.169 +					autostop = stop;
   1.170 +			}
   1.171 +		}
   1.172 +		if (*p)
   1.173 +			errx(1, "[-cf] list: illegal list value");
   1.174 +		if (!stop || !start)
   1.175 +			errx(1, "[-cf] list: values may not include zero");
   1.176 +		if (stop > _POSIX2_LINE_MAX)
   1.177 +			errx(1, "[-cf] list: %d too large (max %d)",
   1.178 +			    stop, _POSIX2_LINE_MAX);
   1.179 +		if (maxval < stop)
   1.180 +			maxval = stop;
   1.181 +		for (pos = positions + start; start++ <= stop; *pos++ = 1)
   1.182 +			;
   1.183 +	}
   1.184 +
   1.185 +	/* overlapping ranges */
   1.186 +	if (autostop && maxval > autostop)
   1.187 +		maxval = autostop;
   1.188 +
   1.189 +	/* set autostart */
   1.190 +	if (autostart)
   1.191 +		memset(positions + 1, '1', autostart);
   1.192 +}
   1.193 +
   1.194 +/* ARGSUSED */
   1.195 +void
   1.196 +c_cut(FILE *fp, char *fname)
   1.197 +{
   1.198 +	int ch, col;
   1.199 +	char *pos;
   1.200 +
   1.201 +	for (;;) {
   1.202 +		pos = positions + 1;
   1.203 +		for (col = maxval; col; --col) {
   1.204 +			if ((ch = getc(fp)) == EOF)
   1.205 +				return;
   1.206 +			if (ch == '\n')
   1.207 +				break;
   1.208 +			if (*pos++)
   1.209 +				(void)putchar(ch);
   1.210 +		}
   1.211 +		if (ch != '\n') {
   1.212 +			if (autostop)
   1.213 +				while ((ch = getc(fp)) != EOF && ch != '\n')
   1.214 +					(void)putchar(ch);
   1.215 +			else
   1.216 +				while ((ch = getc(fp)) != EOF && ch != '\n')
   1.217 +					;
   1.218 +		}
   1.219 +		(void)putchar('\n');
   1.220 +	}
   1.221 +}
   1.222 +
   1.223 +void
   1.224 +f_cut(FILE *fp, char *fname)
   1.225 +{
   1.226 +	int ch, field, isdelim;
   1.227 +	char *pos, *p, sep;
   1.228 +	int output;
   1.229 +	size_t len;
   1.230 +	char *lbuf, *tbuf;
   1.231 +
   1.232 +	for (sep = dchar, tbuf = NULL; (lbuf = fgetln(fp, &len));) {
   1.233 +		output = 0;
   1.234 +		if (lbuf[len - 1] != '\n') {
   1.235 +			/* no newline at the end of the last line so add one */
   1.236 +			if ((tbuf = (char *)malloc(len + 1)) == NULL)
   1.237 +				err(1, NULL);
   1.238 +			memcpy(tbuf, lbuf, len);
   1.239 +			tbuf[len] = '\n';
   1.240 +			lbuf = tbuf;
   1.241 +		}
   1.242 +		for (isdelim = 0, p = lbuf;; ++p) {
   1.243 +			ch = *p;
   1.244 +			/* this should work if newline is delimiter */
   1.245 +			if (ch == sep)
   1.246 +				isdelim = 1;
   1.247 +			if (ch == '\n') {
   1.248 +				if (!isdelim && !sflag)
   1.249 +					(void)fwrite(lbuf, len, 1, stdout);
   1.250 +				break;
   1.251 +			}
   1.252 +		}
   1.253 +		if (!isdelim)
   1.254 +			continue;
   1.255 +
   1.256 +		pos = positions + 1;
   1.257 +		for (field = maxval, p = lbuf; field; --field, ++pos) {
   1.258 +			if (*pos) {
   1.259 +				if (output++)
   1.260 +					(void)putchar(sep);
   1.261 +				while ((ch = *p++) != '\n' && ch != sep)
   1.262 +					(void)putchar(ch);
   1.263 +			} else
   1.264 +				while ((ch = *p++) != '\n' && ch != sep)
   1.265 +					;
   1.266 +			if (ch == '\n')
   1.267 +				break;
   1.268 +		}
   1.269 +		if (ch != '\n') {
   1.270 +			if (autostop) {
   1.271 +				if (output)
   1.272 +					(void)putchar(sep);
   1.273 +				for (; (ch = *p) != '\n'; ++p)
   1.274 +					(void)putchar(ch);
   1.275 +			} else
   1.276 +				for (; (ch = *p) != '\n'; ++p)
   1.277 +					;
   1.278 +		}
   1.279 +		(void)putchar('\n');
   1.280 +	}
   1.281 +	if (tbuf)
   1.282 +		free(tbuf);
   1.283 +}
   1.284 +
   1.285 +void
   1.286 +usage(void)
   1.287 +{
   1.288 +	(void)fprintf(stderr,
   1.289 +	    "usage: cut -b list [-n] [file ...]\n"
   1.290 +	    "       cut -c list [file ...]\n"
   1.291 +	    "       cut -f list [-s] [-d delim] [file ...]\n");
   1.292 +	exit(1);
   1.293 +}