xref: /freebsd/bin/cat/cat.c (revision 41059135ce931c0f1014a999ffabc6bc470ce856)
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  * Kevin Fall.
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 #if 0
34 #ifndef lint
35 static char const copyright[] =
36 "@(#) Copyright (c) 1989, 1993\n\
37 	The Regents of the University of California.  All rights reserved.\n";
38 #endif /* not lint */
39 #endif
40 
41 #ifndef lint
42 #if 0
43 static char sccsid[] = "@(#)cat.c	8.2 (Berkeley) 4/27/95";
44 #endif
45 #endif /* not lint */
46 #include <sys/cdefs.h>
47 __FBSDID("$FreeBSD$");
48 
49 #include <sys/param.h>
50 #include <sys/stat.h>
51 #ifndef NO_UDOM_SUPPORT
52 #include <sys/socket.h>
53 #include <sys/un.h>
54 #include <errno.h>
55 #include <netdb.h>
56 #endif
57 
58 #include <ctype.h>
59 #include <err.h>
60 #include <fcntl.h>
61 #include <locale.h>
62 #include <stddef.h>
63 #include <stdio.h>
64 #include <stdlib.h>
65 #include <string.h>
66 #include <unistd.h>
67 #include <wchar.h>
68 #include <wctype.h>
69 
70 static int bflag, eflag, lflag, nflag, sflag, tflag, vflag;
71 static int rval;
72 static const char *filename;
73 
74 static void usage(void) __dead2;
75 static void scanfiles(char *argv[], int cooked);
76 static void cook_cat(FILE *);
77 static void raw_cat(int);
78 
79 #ifndef NO_UDOM_SUPPORT
80 static int udom_open(const char *path, int flags);
81 #endif
82 
83 /*
84  * Memory strategy threshold, in pages: if physmem is larger than this,
85  * use a large buffer.
86  */
87 #define	PHYSPAGES_THRESHOLD (32 * 1024)
88 
89 /* Maximum buffer size in bytes - do not allow it to grow larger than this. */
90 #define	BUFSIZE_MAX (2 * 1024 * 1024)
91 
92 /*
93  * Small (default) buffer size in bytes. It's inefficient for this to be
94  * smaller than MAXPHYS.
95  */
96 #define	BUFSIZE_SMALL (MAXPHYS)
97 
98 int
99 main(int argc, char *argv[])
100 {
101 	int ch;
102 	struct flock stdout_lock;
103 
104 	setlocale(LC_CTYPE, "");
105 
106 	while ((ch = getopt(argc, argv, "belnstuv")) != -1)
107 		switch (ch) {
108 		case 'b':
109 			bflag = nflag = 1;	/* -b implies -n */
110 			break;
111 		case 'e':
112 			eflag = vflag = 1;	/* -e implies -v */
113 			break;
114 		case 'l':
115 			lflag = 1;
116 			break;
117 		case 'n':
118 			nflag = 1;
119 			break;
120 		case 's':
121 			sflag = 1;
122 			break;
123 		case 't':
124 			tflag = vflag = 1;	/* -t implies -v */
125 			break;
126 		case 'u':
127 			setbuf(stdout, NULL);
128 			break;
129 		case 'v':
130 			vflag = 1;
131 			break;
132 		default:
133 			usage();
134 		}
135 	argv += optind;
136 
137 	if (lflag) {
138 		stdout_lock.l_len = 0;
139 		stdout_lock.l_start = 0;
140 		stdout_lock.l_type = F_WRLCK;
141 		stdout_lock.l_whence = SEEK_SET;
142 		if (fcntl(STDOUT_FILENO, F_SETLKW, &stdout_lock) == -1)
143 			err(EXIT_FAILURE, "stdout");
144 	}
145 
146 	if (bflag || eflag || nflag || sflag || tflag || vflag)
147 		scanfiles(argv, 1);
148 	else
149 		scanfiles(argv, 0);
150 	if (fclose(stdout))
151 		err(1, "stdout");
152 	exit(rval);
153 	/* NOTREACHED */
154 }
155 
156 static void
157 usage(void)
158 {
159 
160 	fprintf(stderr, "usage: cat [-belnstuv] [file ...]\n");
161 	exit(1);
162 	/* NOTREACHED */
163 }
164 
165 static void
166 scanfiles(char *argv[], int cooked)
167 {
168 	int fd, i;
169 	char *path;
170 	FILE *fp;
171 
172 	i = 0;
173 	fd = -1;
174 	while ((path = argv[i]) != NULL || i == 0) {
175 		if (path == NULL || strcmp(path, "-") == 0) {
176 			filename = "stdin";
177 			fd = STDIN_FILENO;
178 		} else {
179 			filename = path;
180 			fd = open(path, O_RDONLY);
181 #ifndef NO_UDOM_SUPPORT
182 			if (fd < 0 && errno == EOPNOTSUPP)
183 				fd = udom_open(path, O_RDONLY);
184 #endif
185 		}
186 		if (fd < 0) {
187 			warn("%s", path);
188 			rval = 1;
189 		} else if (cooked) {
190 			if (fd == STDIN_FILENO)
191 				cook_cat(stdin);
192 			else {
193 				fp = fdopen(fd, "r");
194 				cook_cat(fp);
195 				fclose(fp);
196 			}
197 		} else {
198 			raw_cat(fd);
199 			if (fd != STDIN_FILENO)
200 				close(fd);
201 		}
202 		if (path == NULL)
203 			break;
204 		++i;
205 	}
206 }
207 
208 static void
209 cook_cat(FILE *fp)
210 {
211 	int ch, gobble, line, prev;
212 	wint_t wch;
213 
214 	/* Reset EOF condition on stdin. */
215 	if (fp == stdin && feof(stdin))
216 		clearerr(stdin);
217 
218 	line = gobble = 0;
219 	for (prev = '\n'; (ch = getc(fp)) != EOF; prev = ch) {
220 		if (prev == '\n') {
221 			if (sflag) {
222 				if (ch == '\n') {
223 					if (gobble)
224 						continue;
225 					gobble = 1;
226 				} else
227 					gobble = 0;
228 			}
229 			if (nflag && (!bflag || ch != '\n')) {
230 				(void)fprintf(stdout, "%6d\t", ++line);
231 				if (ferror(stdout))
232 					break;
233 			}
234 		}
235 		if (ch == '\n') {
236 			if (eflag && putchar('$') == EOF)
237 				break;
238 		} else if (ch == '\t') {
239 			if (tflag) {
240 				if (putchar('^') == EOF || putchar('I') == EOF)
241 					break;
242 				continue;
243 			}
244 		} else if (vflag) {
245 			(void)ungetc(ch, fp);
246 			/*
247 			 * Our getwc(3) doesn't change file position
248 			 * on error.
249 			 */
250 			if ((wch = getwc(fp)) == WEOF) {
251 				if (ferror(fp) && errno == EILSEQ) {
252 					clearerr(fp);
253 					/* Resync attempt. */
254 					memset(&fp->_mbstate, 0, sizeof(mbstate_t));
255 					if ((ch = getc(fp)) == EOF)
256 						break;
257 					wch = ch;
258 					goto ilseq;
259 				} else
260 					break;
261 			}
262 			if (!iswascii(wch) && !iswprint(wch)) {
263 ilseq:
264 				if (putchar('M') == EOF || putchar('-') == EOF)
265 					break;
266 				wch = toascii(wch);
267 			}
268 			if (iswcntrl(wch)) {
269 				ch = toascii(wch);
270 				ch = (ch == '\177') ? '?' : (ch | 0100);
271 				if (putchar('^') == EOF || putchar(ch) == EOF)
272 					break;
273 				continue;
274 			}
275 			if (putwchar(wch) == WEOF)
276 				break;
277 			ch = -1;
278 			continue;
279 		}
280 		if (putchar(ch) == EOF)
281 			break;
282 	}
283 	if (ferror(fp)) {
284 		warn("%s", filename);
285 		rval = 1;
286 		clearerr(fp);
287 	}
288 	if (ferror(stdout))
289 		err(1, "stdout");
290 }
291 
292 static void
293 raw_cat(int rfd)
294 {
295 	int off, wfd;
296 	ssize_t nr, nw;
297 	static size_t bsize;
298 	static char *buf = NULL;
299 	struct stat sbuf;
300 
301 	wfd = fileno(stdout);
302 	if (buf == NULL) {
303 		if (fstat(wfd, &sbuf))
304 			err(1, "stdout");
305 		if (S_ISREG(sbuf.st_mode)) {
306 			/* If there's plenty of RAM, use a large copy buffer */
307 			if (sysconf(_SC_PHYS_PAGES) > PHYSPAGES_THRESHOLD)
308 				bsize = MIN(BUFSIZE_MAX, MAXPHYS * 8);
309 			else
310 				bsize = BUFSIZE_SMALL;
311 		} else
312 			bsize = MAX(sbuf.st_blksize,
313 			    (blksize_t)sysconf(_SC_PAGESIZE));
314 		if ((buf = malloc(bsize)) == NULL)
315 			err(1, "malloc() failure of IO buffer");
316 	}
317 	while ((nr = read(rfd, buf, bsize)) > 0)
318 		for (off = 0; nr; nr -= nw, off += nw)
319 			if ((nw = write(wfd, buf + off, (size_t)nr)) < 0)
320 				err(1, "stdout");
321 	if (nr < 0) {
322 		warn("%s", filename);
323 		rval = 1;
324 	}
325 }
326 
327 #ifndef NO_UDOM_SUPPORT
328 
329 static int
330 udom_open(const char *path, int flags)
331 {
332 	struct addrinfo hints, *res, *res0;
333 	char rpath[PATH_MAX];
334 	int fd = -1;
335 	int error;
336 
337 	/*
338 	 * Construct the unix domain socket address and attempt to connect.
339 	 */
340 	bzero(&hints, sizeof(hints));
341 	hints.ai_family = AF_LOCAL;
342 	if (realpath(path, rpath) == NULL)
343 		return (-1);
344 	error = getaddrinfo(rpath, NULL, &hints, &res0);
345 	if (error) {
346 		warn("%s", gai_strerror(error));
347 		errno = EINVAL;
348 		return (-1);
349 	}
350 	for (res = res0; res != NULL; res = res->ai_next) {
351 		fd = socket(res->ai_family, res->ai_socktype,
352 		    res->ai_protocol);
353 		if (fd < 0) {
354 			freeaddrinfo(res0);
355 			return (-1);
356 		}
357 		error = connect(fd, res->ai_addr, res->ai_addrlen);
358 		if (error == 0)
359 			break;
360 		else {
361 			close(fd);
362 			fd = -1;
363 		}
364 	}
365 	freeaddrinfo(res0);
366 
367 	/*
368 	 * handle the open flags by shutting down appropriate directions
369 	 */
370 	if (fd >= 0) {
371 		switch(flags & O_ACCMODE) {
372 		case O_RDONLY:
373 			if (shutdown(fd, SHUT_WR) == -1)
374 				warn(NULL);
375 			break;
376 		case O_WRONLY:
377 			if (shutdown(fd, SHUT_RD) == -1)
378 				warn(NULL);
379 			break;
380 		default:
381 			break;
382 		}
383 	}
384 	return (fd);
385 }
386 
387 #endif
388