xref: /openbsd/usr.bin/split/split.c (revision 76d0caae)
1 /*	$OpenBSD: split.c,v 1.23 2021/11/28 19:28:42 deraadt Exp $	*/
2 /*	$NetBSD: split.c,v 1.5 1995/08/31 22:22:05 jtc Exp $	*/
3 
4 /*
5  * Copyright (c) 1987, 1993, 1994
6  *	The Regents of the University of California.  All rights reserved.
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. Neither the name of the University nor the names of its contributors
17  *    may be used to endorse or promote products derived from this software
18  *    without specific prior written permission.
19  *
20  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30  * SUCH DAMAGE.
31  */
32 
33 #include <sys/types.h>
34 
35 #include <ctype.h>
36 #include <err.h>
37 #include <fcntl.h>
38 #include <limits.h>
39 #include <stdio.h>
40 #include <stdlib.h>
41 #include <string.h>
42 #include <unistd.h>
43 #include <regex.h>
44 
45 #define _MAXBSIZE (64 * 1024)
46 
47 #define DEFLINE	1000			/* Default num lines per file. */
48 
49 ssize_t	 bytecnt;			/* Byte count to split on. */
50 long	 numlines;			/* Line count to split on. */
51 int	 file_open;			/* If a file open. */
52 int	 ifd = -1, ofd = -1;		/* Input/output file descriptors. */
53 char	 bfr[_MAXBSIZE];		/* I/O buffer. */
54 char	 fname[PATH_MAX];		/* File name prefix. */
55 regex_t	 rgx;
56 int	 pflag;
57 int	 sufflen = 2;			/* File name suffix length. */
58 
59 void newfile(void);
60 void split1(void);
61 void split2(void);
62 __dead void usage(void);
63 
64 int
65 main(int argc, char *argv[])
66 {
67 	int ch, scale;
68 	char *ep, *p;
69 	const char *errstr;
70 
71 	if (pledge("stdio rpath wpath cpath", NULL) == -1)
72 		err(1, "pledge");
73 
74 	while ((ch = getopt(argc, argv, "0123456789a:b:l:p:-")) != -1)
75 		switch (ch) {
76 		case '0': case '1': case '2': case '3': case '4':
77 		case '5': case '6': case '7': case '8': case '9':
78 			/*
79 			 * Undocumented kludge: split was originally designed
80 			 * to take a number after a dash.
81 			 */
82 			if (numlines == 0) {
83 				p = argv[optind - 1];
84 				if (p[0] == '-' && p[1] == ch && !p[2])
85 					numlines = strtol(++p, &ep, 10);
86 				else
87 					numlines =
88 					    strtol(argv[optind] + 1, &ep, 10);
89 				if (numlines <= 0 || *ep)
90 					errx(1, "%s: illegal line count",
91 					    optarg);
92 			}
93 			break;
94 		case '-':		/* Undocumented: historic stdin flag. */
95 			if (ifd != -1)
96 				usage();
97 			ifd = 0;
98 			break;
99 		case 'a':		/* suffix length. */
100 			sufflen = strtonum(optarg, 1, NAME_MAX, &errstr);
101 			if (errstr)
102 				errx(1, "%s: %s", optarg, errstr);
103 			break;
104 		case 'b':		/* Byte count. */
105 			if ((bytecnt = strtol(optarg, &ep, 10)) <= 0 ||
106 			    (*ep != '\0' && *ep != 'k' && *ep != 'm'))
107 				errx(1, "%s: illegal byte count", optarg);
108 			if (*ep == 'k')
109 				scale = 1024;
110 			else if (*ep == 'm')
111 				scale = 1048576;
112 			else
113 				scale = 1;
114 			if (bytecnt > SSIZE_MAX / scale)
115 				errx(1, "%s: byte count too large", optarg);
116 			bytecnt *= scale;
117 			break;
118 		case 'p' :      /* pattern matching. */
119 			if (regcomp(&rgx, optarg, REG_EXTENDED|REG_NOSUB) != 0)
120 				errx(1, "%s: illegal regexp", optarg);
121 			pflag = 1;
122 			break;
123 		case 'l':		/* Line count. */
124 			if (numlines != 0)
125 				usage();
126 			if ((numlines = strtol(optarg, &ep, 10)) <= 0 || *ep)
127 				errx(1, "%s: illegal line count", optarg);
128 			break;
129 		default:
130 			usage();
131 		}
132 	argv += optind;
133 	argc -= optind;
134 
135 	if (*argv != NULL)
136 		if (ifd == -1) {		/* Input file. */
137 			if ((ifd = open(*argv, O_RDONLY)) < 0)
138 				err(1, "%s", *argv);
139 			++argv;
140 		}
141 	if (*argv != NULL)			/* File name prefix. */
142 		(void)strlcpy(fname, *argv++, sizeof(fname));
143 	if (*argv != NULL)
144 		usage();
145 
146 	if (strlen(fname) + sufflen >= sizeof(fname))
147 		errx(1, "suffix is too long");
148 	if (pflag && (numlines != 0 || bytecnt != 0))
149 		usage();
150 
151 	if (numlines == 0)
152 		numlines = DEFLINE;
153 	else if (bytecnt != 0)
154 		usage();
155 
156 	if (ifd == -1)				/* Stdin by default. */
157 		ifd = 0;
158 
159 	if (bytecnt) {
160 		split1();
161 		exit (0);
162 	}
163 	split2();
164 	if (pflag)
165 		regfree(&rgx);
166 	exit(0);
167 }
168 
169 /*
170  * split1 --
171  *	Split the input by bytes.
172  */
173 void
174 split1(void)
175 {
176 	ssize_t bcnt, dist, len;
177 	char *C;
178 
179 	for (bcnt = 0;;)
180 		switch ((len = read(ifd, bfr, sizeof(bfr)))) {
181 		case 0:
182 			exit(0);
183 		case -1:
184 			err(1, "read");
185 			/* NOTREACHED */
186 		default:
187 			if (!file_open)
188 				newfile();
189 			if (bcnt + len >= bytecnt) {
190 				dist = bytecnt - bcnt;
191 				if (write(ofd, bfr, dist) != dist)
192 					err(1, "write");
193 				len -= dist;
194 				for (C = bfr + dist; len >= bytecnt;
195 				    len -= bytecnt, C += bytecnt) {
196 					newfile();
197 					if (write(ofd, C, bytecnt) != bytecnt)
198 						err(1, "write");
199 				}
200 				if (len != 0) {
201 					newfile();
202 					if (write(ofd, C, len) != len)
203 						err(1, "write");
204 				} else
205 					file_open = 0;
206 				bcnt = len;
207 			} else {
208 				bcnt += len;
209 				if (write(ofd, bfr, len) != len)
210 					err(1, "write");
211 			}
212 		}
213 }
214 
215 /*
216  * split2 --
217  *	Split the input by lines.
218  */
219 void
220 split2(void)
221 {
222 	long lcnt = 0;
223 	FILE *infp;
224 
225 	/* Stick a stream on top of input file descriptor */
226 	if ((infp = fdopen(ifd, "r")) == NULL)
227 		err(1, "fdopen");
228 
229 	/* Process input one line at a time */
230 	while (fgets(bfr, sizeof(bfr), infp) != NULL) {
231 		const int len = strlen(bfr);
232 
233 		if (len == 0)
234 			continue;
235 
236 		/* If line is too long to deal with, just write it out */
237 		if (bfr[len - 1] != '\n')
238 			goto writeit;
239 
240 		/* Check if we need to start a new file */
241 		if (pflag) {
242 			regmatch_t pmatch;
243 
244 			pmatch.rm_so = 0;
245 			pmatch.rm_eo = len - 1;
246 			if (regexec(&rgx, bfr, 0, &pmatch, REG_STARTEND) == 0)
247 				newfile();
248 		} else if (lcnt++ == numlines) {
249 			newfile();
250 			lcnt = 1;
251 		}
252 
253 writeit:
254 		/* Open output file if needed */
255 		if (!file_open)
256 			newfile();
257 
258 		/* Write out line */
259 		if (write(ofd, bfr, len) != len)
260 			err(1, "write");
261 	}
262 
263 	/* EOF or error? */
264 	if (ferror(infp))
265 		err(1, "read");
266 	else
267 		exit(0);
268 }
269 
270 /*
271  * newfile --
272  *	Open a new output file.
273  */
274 void
275 newfile(void)
276 {
277 	static char *suffix, *sufftail;
278 	char *sptr;
279 
280 	if (ofd == -1) {
281 		ofd = fileno(stdout);
282 		if (*fname == '\0') {
283 			*fname = 'x';	/* no name specified, use 'x' */
284 			memset(fname + 1, 'a', sufflen);
285 			suffix = fname;
286 			sufflen++;	/* treat 'x' as part of suffix */
287 		} else {
288 			suffix = fname + strlen(fname);
289 			memset(suffix, 'a', sufflen);
290 		}
291 		suffix[sufflen] = '\0';
292 		sufftail = suffix + sufflen - 1;
293 	} else {
294 		for (sptr = sufftail; sptr >= suffix; sptr--) {
295 			if (*sptr != 'z') {
296 				(*sptr)++;
297 				break;
298 			} else
299 				*sptr = 'a';
300 		}
301 		if (sptr < suffix)
302 			errx(1, "too many files");
303 	}
304 
305 	if (!freopen(fname, "w", stdout))
306 		err(1, "%s", fname);
307 	file_open = 1;
308 }
309 
310 __dead void
311 usage(void)
312 {
313 	extern char *__progname;
314 
315 	(void)fprintf(stderr, "usage: %s [-a suffix_length]\n"
316 	    "             [-b byte_count[k|m] | -l line_count | -p pattern] "
317 	    "[file [name]]\n", __progname);
318 	exit(1);
319 }
320